Move code into different folder for packaging
This commit is contained in:
@@ -0,0 +1,167 @@
|
||||
import torch
|
||||
import cv2
|
||||
import onnxruntime
|
||||
import numpy as np
|
||||
import pandas as pd
|
||||
import albumentations as A
|
||||
|
||||
from torchvision import transforms, ops
|
||||
from albumentations.pytorch import ToTensorV2
|
||||
|
||||
from utils.conversions import scale_bboxes
|
||||
from utils.manipulations import get_cutout
|
||||
|
||||
def detect(img_path: str, yolo_path: str, resnet_path: str):
|
||||
"""Load an image, detect individual plants and label them as
|
||||
healthy or wilted.
|
||||
|
||||
:param str img_path: path to image
|
||||
:param str yolo_path: path to yolo weights
|
||||
:param str resnet_path: path to resnet weights
|
||||
:returns: tuple of recent image and dict of bounding boxes and
|
||||
their predictions
|
||||
|
||||
"""
|
||||
img = cv2.imread(img_path)
|
||||
|
||||
# Get bounding boxes from object detection model
|
||||
box_coords = get_boxes(yolo_path, img.copy())
|
||||
|
||||
box_coords.sort_values(by=['xmin'], ignore_index=True, inplace=True)
|
||||
|
||||
predictions = []
|
||||
for _, row in box_coords.iterrows():
|
||||
xmin, xmax = int(row['xmin']), int(row['xmax'])
|
||||
ymin, ymax = int(row['ymin']), int(row['ymax'])
|
||||
|
||||
# Get tensor of ROI in BGR
|
||||
cropped_image = get_cutout(img.copy(), xmin, xmax, ymin, ymax)
|
||||
|
||||
# Classify ROI in RGB
|
||||
predictions.append(classify(resnet_path, cropped_image[..., ::-1]))
|
||||
|
||||
# Gather top class and confidence values
|
||||
cls = []
|
||||
cls_conf = []
|
||||
for pred in predictions:
|
||||
ans, index = torch.topk(pred, 1)
|
||||
cls.append(index.int().item())
|
||||
cls_conf.append(round(ans.double().item(), 6))
|
||||
|
||||
# Add predicted classes and confidence values to pandas dataframe
|
||||
box_coords['cls'] = cls
|
||||
box_coords['cls_conf'] = cls_conf
|
||||
|
||||
return box_coords
|
||||
|
||||
|
||||
def classify(resnet_path, img):
|
||||
"""Classify img with object classification model.
|
||||
|
||||
:param model: object classification model
|
||||
:param img: opencv2 image object in RGB
|
||||
:returns: tensor of class predictions
|
||||
"""
|
||||
|
||||
# Transform image for ResNet
|
||||
transform = transforms.Compose([
|
||||
transforms.ToTensor(),
|
||||
transforms.Normalize(mean=[0.485, 0.456, 0.406],
|
||||
std=[0.229, 0.224, 0.225]),
|
||||
transforms.Resize((224, 224))
|
||||
])
|
||||
|
||||
img = transform(img.copy())
|
||||
batch = img.unsqueeze(0)
|
||||
|
||||
# Do inference
|
||||
session = onnxruntime.InferenceSession(resnet_path)
|
||||
outname = [i.name for i in session.get_outputs()]
|
||||
inname = [i.name for i in session.get_inputs()]
|
||||
inp = {inname[0]: batch.numpy()}
|
||||
out = torch.tensor(np.array(session.run(outname, inp)))[0]
|
||||
|
||||
# Apply softmax to get percentage confidence of classes
|
||||
out = torch.nn.functional.softmax(out, dim=1)[0] * 100
|
||||
return out
|
||||
|
||||
|
||||
def apply_nms(predictions,
|
||||
confidence_threshold: float = 0.3,
|
||||
nms_threshold: float = 0.65):
|
||||
"""Apply Non Maximum Suppression to a list of bboxes.
|
||||
|
||||
:param predictions List[Tensor[N, 7]]: predicted bboxes
|
||||
:param confidence_threshold float: discard all bboxes with lower
|
||||
confidence
|
||||
:param nms_threshold float: discard all overlapping bboxes with
|
||||
higher IoU
|
||||
:returns List[Tensor[N, 7]]: filtered bboxes
|
||||
"""
|
||||
preds_nms = []
|
||||
for pred in predictions:
|
||||
pred = pred[pred[:, 6] > confidence_threshold]
|
||||
|
||||
nms_idx = ops.batched_nms(
|
||||
boxes=pred[:, 1:5],
|
||||
scores=pred[:, 6],
|
||||
idxs=pred[:, 5],
|
||||
iou_threshold=nms_threshold,
|
||||
)
|
||||
preds_nms.append(pred[nms_idx])
|
||||
|
||||
return preds_nms
|
||||
|
||||
|
||||
def get_boxes(yolo_path, image):
|
||||
"""Run object detection model on an image and get the bounding box
|
||||
coordinates of all matches.
|
||||
|
||||
:param model: path to onnx object detection model (YOLO)
|
||||
:param img: opencv2 image object
|
||||
:returns: pandas dataframe of matches
|
||||
"""
|
||||
# Convert from BGR to RGB
|
||||
img = image[..., ::-1].copy()
|
||||
|
||||
resized_hw = (640, 640)
|
||||
original_hw = (image.shape[0], image.shape[1])
|
||||
|
||||
transform = [
|
||||
A.LongestMaxSize(max(resized_hw)),
|
||||
A.PadIfNeeded(
|
||||
resized_hw[0],
|
||||
resized_hw[1],
|
||||
border_mode=0,
|
||||
value=(114, 114, 114),
|
||||
),
|
||||
A.ToFloat(max_value=255),
|
||||
ToTensorV2(transpose_mask=True),
|
||||
]
|
||||
|
||||
# Pad (letterbox) and transform image to correct dims
|
||||
transform = A.Compose(transform)
|
||||
img = transform(image=img)
|
||||
|
||||
# Add batch dimension
|
||||
img['image'] = img['image'].unsqueeze(0)
|
||||
|
||||
# Do inference
|
||||
session = onnxruntime.InferenceSession(yolo_path)
|
||||
outname = [i.name for i in session.get_outputs()]
|
||||
inname = [i.name for i in session.get_inputs()]
|
||||
inp = {inname[0]: img['image'].numpy()}
|
||||
out = torch.tensor(np.array(session.run(outname, inp)))[0]
|
||||
|
||||
# Apply NMS to results
|
||||
preds_nms = apply_nms([out])[0]
|
||||
|
||||
# Convert boxes from resized img to original img
|
||||
xyxy_boxes = preds_nms[:, [1, 2, 3, 4]] # xmin, ymin, xmax, ymax
|
||||
bboxes = scale_bboxes(xyxy_boxes, resized_hw, original_hw).int().numpy()
|
||||
|
||||
# Construct DataFrame with bboxes and their confidence
|
||||
box_coords = pd.DataFrame(np.c_[bboxes, preds_nms[:, 6]])
|
||||
box_coords.columns = ['xmin', 'ymin', 'xmax', 'ymax', 'box_conf']
|
||||
|
||||
return box_coords
|
||||
@@ -0,0 +1,500 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 1,
|
||||
"id": "3fe8177c",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"import fiftyone as fo\n",
|
||||
"from PIL import Image\n",
|
||||
"from detection import detect"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 2,
|
||||
"id": "32f0f8ec",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"name = \"dataset\"\n",
|
||||
"dataset_dir = \"dataset\""
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 3,
|
||||
"id": "6343aa55",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
" 100% |█████████████████| 640/640 [716.7ms elapsed, 0s remaining, 894.6 samples/s] \n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"# The splits to load\n",
|
||||
"splits = [\"val\"]\n",
|
||||
"\n",
|
||||
"# Load the dataset, using tags to mark the samples in each split\n",
|
||||
"dataset = fo.Dataset(name)\n",
|
||||
"for split in splits:\n",
|
||||
" dataset.add_dir(\n",
|
||||
" dataset_dir=dataset_dir,\n",
|
||||
" dataset_type=fo.types.YOLOv5Dataset,\n",
|
||||
" split=split,\n",
|
||||
" tags=split,\n",
|
||||
" )\n",
|
||||
"\n",
|
||||
"classes = dataset.default_classes"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 4,
|
||||
"id": "29827e3f",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
" 100% |█████████████████| 640/640 [8.8m elapsed, 0s remaining, 1.5 samples/s] \n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"# Do detections with model and save bounding boxes\n",
|
||||
"with fo.ProgressBar() as pb:\n",
|
||||
" for sample in pb(dataset.view()):\n",
|
||||
" image = Image.open(sample.filepath)\n",
|
||||
" w, h = image.size\n",
|
||||
" pred = detect(sample.filepath, '../weights/yolo.onnx', '../weights/resnet.onnx')\n",
|
||||
"\n",
|
||||
" detections = []\n",
|
||||
" for _, row in pred.iterrows():\n",
|
||||
" xmin, xmax = int(row['xmin']), int(row['xmax'])\n",
|
||||
" ymin, ymax = int(row['ymin']), int(row['ymax'])\n",
|
||||
" rel_box = [\n",
|
||||
" xmin / w, ymin / h, (xmax - xmin) / w, (ymax - ymin) / h\n",
|
||||
" ]\n",
|
||||
" detections.append(\n",
|
||||
" fo.Detection(label=classes[int(row['cls'])],\n",
|
||||
" bounding_box=rel_box,\n",
|
||||
" confidence=int(row['cls_conf'])))\n",
|
||||
"\n",
|
||||
" sample[\"predictions\"] = fo.Detections(detections=detections)\n",
|
||||
" sample.save()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 5,
|
||||
"id": "06e1b4c0",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"dataset.persistent = True"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 6,
|
||||
"id": "8ad67806",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"Evaluating detections...\n",
|
||||
" 100% |█████████████████| 640/640 [2.0s elapsed, 0s remaining, 319.5 samples/s] \n",
|
||||
"Performing IoU sweep...\n",
|
||||
" 100% |█████████████████| 640/640 [2.1s elapsed, 0s remaining, 297.3 samples/s] \n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"results = dataset.view().evaluate_detections(\n",
|
||||
" \"predictions\",\n",
|
||||
" gt_field=\"ground_truth\",\n",
|
||||
" eval_key=\"eval\",\n",
|
||||
" compute_mAP=True,\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 7,
|
||||
"id": "b180420b",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
" precision recall f1-score support\n",
|
||||
"\n",
|
||||
" Healthy 0.82 0.74 0.78 662\n",
|
||||
" Stressed 0.71 0.78 0.74 488\n",
|
||||
"\n",
|
||||
" micro avg 0.77 0.76 0.76 1150\n",
|
||||
" macro avg 0.77 0.76 0.76 1150\n",
|
||||
"weighted avg 0.77 0.76 0.77 1150\n",
|
||||
"\n",
|
||||
"0.6225828327927432\n"
|
||||
]
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"text/plain": []
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"text/plain": []
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"application/vnd.jupyter.widget-view+json": {
|
||||
"model_id": "0e0a5d23d3f148419bd62ecf4a07dce9",
|
||||
"version_major": 2,
|
||||
"version_minor": 0
|
||||
},
|
||||
"text/plain": [
|
||||
"FigureWidget({\n",
|
||||
" 'data': [{'mode': 'markers',\n",
|
||||
" 'opacity': 0.1,\n",
|
||||
" 'type': 'scatter',\n",
|
||||
" 'uid': 'c432eebe-8bf5-4b15-834c-abdc3af7c18b',\n",
|
||||
" 'x': array([0, 1, 2, 0, 1, 2, 0, 1, 2]),\n",
|
||||
" 'y': array([0, 0, 0, 1, 1, 1, 2, 2, 2])},\n",
|
||||
" {'colorscale': [[0.0, 'rgb(255,245,235)'], [0.125,\n",
|
||||
" 'rgb(254,230,206)'], [0.25, 'rgb(253,208,162)'],\n",
|
||||
" [0.375, 'rgb(253,174,107)'], [0.5, 'rgb(253,141,60)'],\n",
|
||||
" [0.625, 'rgb(241,105,19)'], [0.75, 'rgb(217,72,1)'],\n",
|
||||
" [0.875, 'rgb(166,54,3)'], [1.0, 'rgb(127,39,4)']],\n",
|
||||
" 'hoverinfo': 'skip',\n",
|
||||
" 'showscale': False,\n",
|
||||
" 'type': 'heatmap',\n",
|
||||
" 'uid': '01498ed3-dbe3-4c9e-baf9-49d814522d20',\n",
|
||||
" 'z': array([[105, 158, 0],\n",
|
||||
" [ 0, 382, 106],\n",
|
||||
" [493, 0, 169]]),\n",
|
||||
" 'zmax': 493,\n",
|
||||
" 'zmin': 0},\n",
|
||||
" {'colorbar': {'len': 1, 'lenmode': 'fraction'},\n",
|
||||
" 'colorscale': [[0.0, 'rgb(255,245,235)'], [0.125,\n",
|
||||
" 'rgb(254,230,206)'], [0.25, 'rgb(253,208,162)'],\n",
|
||||
" [0.375, 'rgb(253,174,107)'], [0.5, 'rgb(253,141,60)'],\n",
|
||||
" [0.625, 'rgb(241,105,19)'], [0.75, 'rgb(217,72,1)'],\n",
|
||||
" [0.875, 'rgb(166,54,3)'], [1.0, 'rgb(127,39,4)']],\n",
|
||||
" 'hovertemplate': '<b>count: %{z}</b><br>truth: %{y}<br>predicted: %{x}<extra></extra>',\n",
|
||||
" 'opacity': 0.25,\n",
|
||||
" 'type': 'heatmap',\n",
|
||||
" 'uid': '10618ef9-35c5-451a-b915-32b70d7bfb5e',\n",
|
||||
" 'z': array([[105, 158, 0],\n",
|
||||
" [ 0, 382, 106],\n",
|
||||
" [493, 0, 169]]),\n",
|
||||
" 'zmax': 493,\n",
|
||||
" 'zmin': 0}],\n",
|
||||
" 'layout': {'clickmode': 'event',\n",
|
||||
" 'margin': {'b': 0, 'l': 0, 'r': 0, 't': 30},\n",
|
||||
" 'template': '...',\n",
|
||||
" 'title': {},\n",
|
||||
" 'xaxis': {'constrain': 'domain',\n",
|
||||
" 'range': [-0.5, 2.5],\n",
|
||||
" 'tickmode': 'array',\n",
|
||||
" 'ticktext': [Healthy, Stressed, (none)],\n",
|
||||
" 'tickvals': array([0, 1, 2])},\n",
|
||||
" 'yaxis': {'constrain': 'domain',\n",
|
||||
" 'range': [-0.5, 2.5],\n",
|
||||
" 'scaleanchor': 'x',\n",
|
||||
" 'scaleratio': 1,\n",
|
||||
" 'tickmode': 'array',\n",
|
||||
" 'ticktext': array(['(none)', 'Stressed', 'Healthy'], dtype=object),\n",
|
||||
" 'tickvals': array([0, 1, 2])}}\n",
|
||||
"})"
|
||||
]
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"text/plain": []
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"text/plain": []
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"application/vnd.jupyter.widget-view+json": {
|
||||
"model_id": "915261c4504943ec9b07813fbe75b8c9",
|
||||
"version_major": 2,
|
||||
"version_minor": 0
|
||||
},
|
||||
"text/plain": [
|
||||
"FigureWidget({\n",
|
||||
" 'data': [{'customdata': array([99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 98.8,\n",
|
||||
" 98. , 98. , 97.7, 97. , 97. , 97. , 96.9, 96.4, 96. , 95.7, 94.8, 94.3,\n",
|
||||
" 93.8, 92.8, 92.2, 91.5, 90.9, 90.7, 89.3, 88.3, 87.8, 87.1, 86.8, 86.3,\n",
|
||||
" 85.8, 85.1, 84.6, 84.1, 83.2, 82. , 81.7, 80.8, 79.7, 79.4, 78.6, 77.8,\n",
|
||||
" 77.4, 76.3, 75.5, 74.6, 73.6, 72.3, 71. , 69.9, 68.6, 67.5, 66.4, 65.4,\n",
|
||||
" 64.3, 63.2, 62.2, 61. , 60.2, 59.1, 58.1, 56.9, 50.4, 49. , 47.3, 46.4,\n",
|
||||
" 40.8, 25.3, 10. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. ]),\n",
|
||||
" 'hovertemplate': ('<b>class: %{text}</b><br>recal' ... 'customdata:.3f}<extra></extra>'),\n",
|
||||
" 'line': {'color': '#3366CC'},\n",
|
||||
" 'mode': 'lines',\n",
|
||||
" 'name': 'Healthy (AP = 0.631)',\n",
|
||||
" 'text': array(['Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy',\n",
|
||||
" 'Healthy', 'Healthy', 'Healthy', 'Healthy', 'Healthy'], dtype='<U7'),\n",
|
||||
" 'type': 'scatter',\n",
|
||||
" 'uid': '7d365acd-11ac-471f-95c1-0c45b4af1e64',\n",
|
||||
" 'x': array([0. , 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.1 , 0.11,\n",
|
||||
" 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.2 , 0.21, 0.22, 0.23,\n",
|
||||
" 0.24, 0.25, 0.26, 0.27, 0.28, 0.29, 0.3 , 0.31, 0.32, 0.33, 0.34, 0.35,\n",
|
||||
" 0.36, 0.37, 0.38, 0.39, 0.4 , 0.41, 0.42, 0.43, 0.44, 0.45, 0.46, 0.47,\n",
|
||||
" 0.48, 0.49, 0.5 , 0.51, 0.52, 0.53, 0.54, 0.55, 0.56, 0.57, 0.58, 0.59,\n",
|
||||
" 0.6 , 0.61, 0.62, 0.63, 0.64, 0.65, 0.66, 0.67, 0.68, 0.69, 0.7 , 0.71,\n",
|
||||
" 0.72, 0.73, 0.74, 0.75, 0.76, 0.77, 0.78, 0.79, 0.8 , 0.81, 0.82, 0.83,\n",
|
||||
" 0.84, 0.85, 0.86, 0.87, 0.88, 0.89, 0.9 , 0.91, 0.92, 0.93, 0.94, 0.95,\n",
|
||||
" 0.96, 0.97, 0.98, 0.99, 1. ]),\n",
|
||||
" 'y': array([1. , 1. , 1. , 1. , 1. , 1. ,\n",
|
||||
" 1. , 1. , 1. , 1. , 1. , 0.97197452,\n",
|
||||
" 0.88968238, 0.88968238, 0.88968238, 0.88968238, 0.88968238, 0.88968238,\n",
|
||||
" 0.88968238, 0.88962774, 0.88906313, 0.88406656, 0.87159671, 0.86864529,\n",
|
||||
" 0.86658584, 0.86482334, 0.86414826, 0.86302011, 0.86087578, 0.8598177 ,\n",
|
||||
" 0.85442276, 0.84805184, 0.84805184, 0.84805184, 0.84805184, 0.84805184,\n",
|
||||
" 0.84805184, 0.84805184, 0.84805184, 0.84805184, 0.84805184, 0.84805184,\n",
|
||||
" 0.84805184, 0.84805184, 0.84805184, 0.84805184, 0.84805184, 0.84727236,\n",
|
||||
" 0.84615682, 0.84387252, 0.84024854, 0.83876528, 0.8378955 , 0.83308277,\n",
|
||||
" 0.82724578, 0.82424648, 0.82113648, 0.8193439 , 0.81764747, 0.81764747,\n",
|
||||
" 0.81764747, 0.81716972, 0.81710833, 0.8164769 , 0.81554506, 0.81418573,\n",
|
||||
" 0.81247877, 0.81064826, 0.73478697, 0.73331137, 0.73195383, 0.73122457,\n",
|
||||
" 0.65219773, 0.41090604, 0.16526846, 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. ])},\n",
|
||||
" {'customdata': array([99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. , 99. ,\n",
|
||||
" 98. , 98. , 98. , 98. , 98. , 98. , 97.7, 97. , 97. , 96.7, 96.5, 96. ,\n",
|
||||
" 95.7, 95.7, 95.4, 94.7, 94.6, 93.6, 93.3, 92.4, 91.6, 91.1, 90.5, 90. ,\n",
|
||||
" 89.3, 88.6, 88.2, 87.4, 87.1, 86.2, 85.4, 84.9, 84. , 83.4, 82. , 81.5,\n",
|
||||
" 80.5, 79.4, 78.5, 77.5, 76.4, 75.6, 74.6, 73.7, 72.9, 71.8, 70.9, 70.2,\n",
|
||||
" 69.5, 68.5, 67.1, 66.6, 65.5, 63.8, 62.3, 61.1, 60.1, 53.5, 52.6, 51.2,\n",
|
||||
" 45. , 43.9, 38.2, 37.4, 31.3, 25.5, 10. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. ]),\n",
|
||||
" 'hovertemplate': ('<b>class: %{text}</b><br>recal' ... 'customdata:.3f}<extra></extra>'),\n",
|
||||
" 'line': {'color': '#DC3912'},\n",
|
||||
" 'mode': 'lines',\n",
|
||||
" 'name': 'Stressed (AP = 0.614)',\n",
|
||||
" 'text': array(['Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed',\n",
|
||||
" 'Stressed', 'Stressed', 'Stressed', 'Stressed', 'Stressed'], dtype='<U8'),\n",
|
||||
" 'type': 'scatter',\n",
|
||||
" 'uid': 'c00a4fd2-8937-4000-bfbe-6429b1528d2d',\n",
|
||||
" 'x': array([0. , 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.1 , 0.11,\n",
|
||||
" 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.2 , 0.21, 0.22, 0.23,\n",
|
||||
" 0.24, 0.25, 0.26, 0.27, 0.28, 0.29, 0.3 , 0.31, 0.32, 0.33, 0.34, 0.35,\n",
|
||||
" 0.36, 0.37, 0.38, 0.39, 0.4 , 0.41, 0.42, 0.43, 0.44, 0.45, 0.46, 0.47,\n",
|
||||
" 0.48, 0.49, 0.5 , 0.51, 0.52, 0.53, 0.54, 0.55, 0.56, 0.57, 0.58, 0.59,\n",
|
||||
" 0.6 , 0.61, 0.62, 0.63, 0.64, 0.65, 0.66, 0.67, 0.68, 0.69, 0.7 , 0.71,\n",
|
||||
" 0.72, 0.73, 0.74, 0.75, 0.76, 0.77, 0.78, 0.79, 0.8 , 0.81, 0.82, 0.83,\n",
|
||||
" 0.84, 0.85, 0.86, 0.87, 0.88, 0.89, 0.9 , 0.91, 0.92, 0.93, 0.94, 0.95,\n",
|
||||
" 0.96, 0.97, 0.98, 0.99, 1. ]),\n",
|
||||
" 'y': array([1. , 1. , 1. , 1. , 1. , 1. ,\n",
|
||||
" 1. , 1. , 1. , 1. , 1. , 1. ,\n",
|
||||
" 0.91667218, 0.91667218, 0.91667218, 0.91667218, 0.91667218, 0.91667218,\n",
|
||||
" 0.89886256, 0.8706589 , 0.8706589 , 0.86176273, 0.8566061 , 0.8426063 ,\n",
|
||||
" 0.8414663 , 0.8414663 , 0.8414663 , 0.83779842, 0.83730636, 0.82590162,\n",
|
||||
" 0.82226179, 0.81571494, 0.80136981, 0.79562115, 0.79030661, 0.78516763,\n",
|
||||
" 0.77779393, 0.77779393, 0.7766316 , 0.77013304, 0.76981741, 0.7687736 ,\n",
|
||||
" 0.76706332, 0.76615004, 0.76345747, 0.76120645, 0.75633043, 0.75442522,\n",
|
||||
" 0.75216493, 0.74937749, 0.74462707, 0.73753607, 0.73544297, 0.73534102,\n",
|
||||
" 0.73534102, 0.73491603, 0.73330167, 0.73209432, 0.73192653, 0.73073596,\n",
|
||||
" 0.7298439 , 0.72819654, 0.72615935, 0.72491458, 0.72298246, 0.71613728,\n",
|
||||
" 0.71363144, 0.71260932, 0.71076601, 0.64219247, 0.63885976, 0.636554 ,\n",
|
||||
" 0.57072089, 0.56905625, 0.49989485, 0.4972456 , 0.42550943, 0.35394558,\n",
|
||||
" 0.14235075, 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. , 0. ,\n",
|
||||
" 0. , 0. , 0. , 0. , 0. ])}],\n",
|
||||
" 'layout': {'margin': {'b': 0, 'l': 0, 'r': 0, 't': 30},\n",
|
||||
" 'shapes': [{'line': {'dash': 'dash'}, 'type': 'line', 'x0': 0, 'x1': 1, 'y0': 1, 'y1': 0}],\n",
|
||||
" 'template': '...',\n",
|
||||
" 'xaxis': {'constrain': 'domain', 'range': [0, 1], 'title': {'text': 'Recall'}},\n",
|
||||
" 'yaxis': {'constrain': 'domain',\n",
|
||||
" 'range': [0, 1],\n",
|
||||
" 'scaleanchor': 'x',\n",
|
||||
" 'scaleratio': 1,\n",
|
||||
" 'title': {'text': 'Precision'}}}\n",
|
||||
"})"
|
||||
]
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"# Print a classification report for all classes\n",
|
||||
"results.print_report()\n",
|
||||
"\n",
|
||||
"print(results.mAP())\n",
|
||||
"\n",
|
||||
"# Plot confusion matrix\n",
|
||||
"matrix = results.plot_confusion_matrix(classes=classes)\n",
|
||||
"matrix.show()\n",
|
||||
"\n",
|
||||
"pr_curves = results.plot_pr_curves(classes=classes)\n",
|
||||
"pr_curves.show()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 8,
|
||||
"id": "d1137788",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"Session launched. Run `session.show()` to open the App in a cell output.\n"
|
||||
]
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"application/javascript": [
|
||||
"window.open('http://localhost:5151/');"
|
||||
],
|
||||
"text/plain": [
|
||||
"<IPython.core.display.Javascript object>"
|
||||
]
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"session = fo.launch_app(dataset, auto=False)\n",
|
||||
"session.view = dataset.view()\n",
|
||||
"session.plots.attach(matrix)\n",
|
||||
"session.open_tab()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 9,
|
||||
"id": "6eed9a86",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"def export_dataset(dataset, export_dir):\n",
|
||||
" label_field = \"ground_truth\"\n",
|
||||
"\n",
|
||||
" # The splits to export\n",
|
||||
" splits = [\"val\"]\n",
|
||||
"\n",
|
||||
" classes = [\"Healthy\", \"Stressed\"]\n",
|
||||
"\n",
|
||||
" # Export the splits\n",
|
||||
" for split in splits:\n",
|
||||
" split_view = dataset.match_tags(split)\n",
|
||||
" split_view.export(\n",
|
||||
" export_dir=export_dir,\n",
|
||||
" dataset_type=fo.types.YOLOv5Dataset,\n",
|
||||
" label_field=label_field,\n",
|
||||
" split=split,\n",
|
||||
" classes=classes,\n",
|
||||
" )"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "19c5b271",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "ebdde519",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3 (ipykernel)",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 3
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.7.15"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 5
|
||||
}
|
||||
@@ -0,0 +1,64 @@
|
||||
import fiftyone as fo
|
||||
from PIL import Image
|
||||
from detection import detect
|
||||
|
||||
name = "dataset-small"
|
||||
dataset_dir = "/home/zenon/Documents/master-thesis/classification/evaluation/dataset-small"
|
||||
|
||||
# The splits to load
|
||||
splits = ["val"]
|
||||
|
||||
# Load the dataset, using tags to mark the samples in each split
|
||||
dataset = fo.Dataset(name)
|
||||
for split in splits:
|
||||
dataset.add_dir(
|
||||
dataset_dir=dataset_dir,
|
||||
dataset_type=fo.types.YOLOv5Dataset,
|
||||
split=split,
|
||||
tags=split,
|
||||
)
|
||||
|
||||
classes = dataset.default_classes
|
||||
predictions_view = dataset.view()
|
||||
|
||||
with fo.ProgressBar() as pb:
|
||||
for sample in pb(predictions_view):
|
||||
image = Image.open(sample.filepath)
|
||||
w, h = image.size
|
||||
pred = detect(sample.filepath, '../weights/yolo.onnx', '../weights/resnet.onnx')
|
||||
|
||||
detections = []
|
||||
for _, row in pred.iterrows():
|
||||
xmin, xmax = int(row['xmin']), int(row['xmax'])
|
||||
ymin, ymax = int(row['ymin']), int(row['ymax'])
|
||||
rel_box = [
|
||||
xmin / w, ymin / h, (xmax - xmin) / w, (ymax - ymin) / h
|
||||
]
|
||||
detections.append(
|
||||
fo.Detection(label=classes[int(row['cls'])],
|
||||
bounding_box=rel_box,
|
||||
confidence=int(row['cls_conf'])))
|
||||
|
||||
sample["yolo_resnet"] = fo.Detections(detections=detections)
|
||||
sample.save()
|
||||
|
||||
results = predictions_view.evaluate_detections(
|
||||
"yolo_resnet",
|
||||
gt_field="ground_truth",
|
||||
eval_key="eval",
|
||||
compute_mAP=True,
|
||||
)
|
||||
|
||||
# Get the 10 most common classes in the dataset
|
||||
counts = dataset.count_values("ground_truth.detections.label")
|
||||
classes_top10 = sorted(counts, key=counts.get, reverse=True)[:10]
|
||||
|
||||
# Print a classification report for the top-10 classes
|
||||
results.print_report(classes=classes_top10)
|
||||
|
||||
plot = results.plot_pr_curves(classes=["Healthy", "Stressed"])
|
||||
plot.show()
|
||||
|
||||
session = fo.launch_app(dataset)
|
||||
session.view = predictions_view
|
||||
session.wait()
|
||||
@@ -0,0 +1,135 @@
|
||||
import logging
|
||||
import argparse
|
||||
import cv2
|
||||
import json
|
||||
import os
|
||||
|
||||
from utils.conversions import convert_to_yolo
|
||||
from detection import detect
|
||||
|
||||
template = [{
|
||||
"data": {
|
||||
"image": "/data/local-files/?d=evaluation/images/0.jpg"
|
||||
},
|
||||
"predictions": [{
|
||||
"model_version":
|
||||
"one",
|
||||
"score":
|
||||
0.0,
|
||||
"result": [{
|
||||
"id": "result1",
|
||||
"type": "rectanglelabels",
|
||||
"from_name": "label",
|
||||
"to_name": "image",
|
||||
"original_width": 474,
|
||||
"original_height": 266,
|
||||
"image_rotation": 0,
|
||||
"confidence": 0,
|
||||
"value": {
|
||||
"rotation": 0,
|
||||
"x": 19.62,
|
||||
"y": 15.04,
|
||||
"width": 55.06,
|
||||
"height": 78.2,
|
||||
"rectanglelabels": ["Stressed"]
|
||||
}
|
||||
}]
|
||||
}]
|
||||
}]
|
||||
|
||||
|
||||
def create_json_annotation(image_path, yolo_path, resnet_path):
|
||||
"""Create a JSON representation of identified bounding boxes.
|
||||
|
||||
:param image_path str: path to image
|
||||
:param yolo_path str: path to YOLO model in ONNX format
|
||||
:param resnet_path str: path to ResNet model in ONNX format
|
||||
:returns Dict: bounding boxes in labelstudio JSON format
|
||||
"""
|
||||
template[0]['data'][
|
||||
'image'] = "/data/local-files/?d=evaluation/" + image_path
|
||||
img = cv2.imread(image_path)
|
||||
(height, width) = img.shape[0], img.shape[1]
|
||||
bboxes = detect(image_path, yolo_path, resnet_path)
|
||||
result = template[0]['predictions'][0]['result']
|
||||
|
||||
results = []
|
||||
for idx, row in bboxes.iterrows():
|
||||
modified = convert_to_yolo(row, width, height)
|
||||
json_result = {}
|
||||
json_result['id'] = 'result' + str(idx + 1)
|
||||
json_result['type'] = 'rectanglelabels'
|
||||
json_result['from_name'] = 'label'
|
||||
json_result['to_name'] = 'image'
|
||||
json_result['original_width'] = width
|
||||
json_result['original_height'] = height
|
||||
json_result['image_rotation'] = 0
|
||||
json_result['value'] = {}
|
||||
json_result['value']['rotation'] = 0
|
||||
json_result['value']['x'] = modified['xmin%']
|
||||
json_result['value']['y'] = modified['ymin%']
|
||||
json_result['value']['width'] = modified['width%']
|
||||
json_result['value']['height'] = modified['height%']
|
||||
if modified['cls'] == 0:
|
||||
json_result['value']['rectanglelabels'] = ['Healthy']
|
||||
else:
|
||||
json_result['value']['rectanglelabels'] = ['Stressed']
|
||||
results.append(json_result)
|
||||
|
||||
template[0]['predictions'][0]['result'] = results
|
||||
return template
|
||||
|
||||
|
||||
def write_labels_to_disk(image_dir, output_dir, yolo_path, resnet_path):
|
||||
"""Read images from disk, classify them and output bounding boxes
|
||||
in labelstudio JSON format.
|
||||
|
||||
:param image_dir str: directory containing images to label
|
||||
:param output_dir str: directory to save JSON files to
|
||||
:param yolo_path str: path to YOLO model in ONNX format
|
||||
:param resnet_path str: path to ResNet model in ONNX format
|
||||
:returns: None
|
||||
"""
|
||||
image_dir = os.path.join(image_dir, '')
|
||||
for file in os.listdir(image_dir):
|
||||
filename = os.fsdecode(file)
|
||||
filename_wo_ext = os.path.splitext(filename)[0]
|
||||
rel_output_path = os.path.join(output_dir, filename_wo_ext + '.json')
|
||||
json_data = create_json_annotation(image_dir + filename, yolo_path,
|
||||
resnet_path)
|
||||
os.makedirs(os.path.dirname(os.path.join(output_dir, filename)),
|
||||
exist_ok=True)
|
||||
logging.info('Writing json file for %s', filename)
|
||||
with open(rel_output_path, 'w') as f:
|
||||
json.dump(json_data, f)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--source',
|
||||
type=str,
|
||||
help='source folder with images',
|
||||
required=True)
|
||||
parser.add_argument('--output',
|
||||
type=str,
|
||||
help='output folder for json files',
|
||||
required=True)
|
||||
parser.add_argument('--yolo',
|
||||
type=str,
|
||||
help='path to YOLO model in ONNX format',
|
||||
required=True)
|
||||
parser.add_argument('--resnet',
|
||||
type=str,
|
||||
help='path to ResNet model in ONNX format',
|
||||
required=True)
|
||||
parser.add_argument(
|
||||
'--log',
|
||||
type=str,
|
||||
help='log level (debug, info, warning, error, critical)',
|
||||
default='warning')
|
||||
opt = parser.parse_args()
|
||||
numeric_level = getattr(logging, opt.log.upper(), None)
|
||||
logging.basicConfig(format='%(levelname)s::%(asctime)s::%(message)s',
|
||||
datefmt='%Y-%m-%dT%H:%M:%S',
|
||||
level=numeric_level)
|
||||
write_labels_to_disk(opt.source, opt.output, opt.yolo, opt.resnet)
|
||||
@@ -0,0 +1,13 @@
|
||||
CC=gcc
|
||||
CXX=g++
|
||||
SRCS=capture.cc
|
||||
OBJS=$(subst .cc,.o,$(SRCS))
|
||||
|
||||
all: $(SRCS)
|
||||
$(CXX) -std=c++11 -Wall -I/usr/lib/opencv -I/usr/include/opencv4 $< -L/usr/lib -lopencv_core -lopencv_videoio -lopencv_imgcodecs -o capture
|
||||
|
||||
clean:
|
||||
rm -f $(OBJS)
|
||||
|
||||
distclean: clean
|
||||
rm -f capture
|
||||
@@ -0,0 +1,58 @@
|
||||
#include <cstdio>
|
||||
#include <opencv2/opencv.hpp>
|
||||
|
||||
using namespace std;
|
||||
using namespace cv;
|
||||
|
||||
string video_pipeline(int capture_width, int capture_height, int framerate,
|
||||
int flip_method) {
|
||||
return "nvarguscamerasrc ! "
|
||||
"video/x-raw(memory:NVMM),width=(int)" +
|
||||
to_string(capture_width) + ",height=(int)" +
|
||||
to_string(capture_height) + ",framerate=(fraction)" +
|
||||
to_string(framerate) +
|
||||
"/1 ! "
|
||||
"nvvidconv flip-method=" +
|
||||
to_string(flip_method) +
|
||||
" ! "
|
||||
"video/x-raw,format=(string)BGRx ! "
|
||||
"videoconvert ! "
|
||||
"video/x-raw,format=(string)BGR,width=(int)3264,height=(int)2464 ! "
|
||||
"appsink";
|
||||
}
|
||||
|
||||
int main() {
|
||||
int capture_width = 3264;
|
||||
int capture_height = 2464;
|
||||
int framerate = 21;
|
||||
int flip_method = 2;
|
||||
|
||||
string pipeline =
|
||||
video_pipeline(capture_width, capture_height, framerate, flip_method);
|
||||
cout << "Using pipeline: \n\t" << pipeline << endl;
|
||||
|
||||
VideoCapture cap(pipeline, CAP_GSTREAMER);
|
||||
if (!cap.isOpened()) {
|
||||
cout << "Failed to open camera." << endl;
|
||||
return -1;
|
||||
}
|
||||
|
||||
Mat img;
|
||||
|
||||
// Take 21 images (1s) to get automatic ISO
|
||||
for (int i = 0; i < 21; i++) {
|
||||
cap.read(img);
|
||||
}
|
||||
|
||||
cap.read(img);
|
||||
|
||||
try {
|
||||
imwrite("current.jpg", img);
|
||||
} catch (const Exception &ex) {
|
||||
fprintf(stderr, "Exception converting image to JPG format: %s\n",
|
||||
ex.what());
|
||||
}
|
||||
|
||||
cap.release();
|
||||
return 0;
|
||||
}
|
||||
@@ -0,0 +1,153 @@
|
||||
import argparse
|
||||
import logging
|
||||
import atexit
|
||||
import os
|
||||
|
||||
import cv2
|
||||
import json
|
||||
import base64
|
||||
import datetime
|
||||
from dateutil.parser import parse
|
||||
from copy import deepcopy
|
||||
|
||||
from flask import Flask
|
||||
from apscheduler.schedulers.background import BackgroundScheduler
|
||||
|
||||
from multiprocessing import Manager
|
||||
|
||||
from evaluation.detection import detect
|
||||
from utils.manipulations import draw_boxes
|
||||
|
||||
app = Flask(__name__)
|
||||
scheduler = BackgroundScheduler(daemon=True)
|
||||
manager = Manager()
|
||||
pred = manager.dict()
|
||||
|
||||
|
||||
def get_pred():
|
||||
tmp = deepcopy(pred)
|
||||
take_image()
|
||||
logging.debug('Starting image classification')
|
||||
preds = detect('current.jpg', '../weights/yolo.onnx',
|
||||
'../weights/resnet.onnx')
|
||||
logging.debug('Finished image classification: %s', preds)
|
||||
logging.debug('Reading current.jpg for drawing bounding boxes')
|
||||
current = cv2.imread('current.jpg')
|
||||
logging.debug('Drawing bounding boxes on current.jpg')
|
||||
bbox_img = draw_boxes(
|
||||
current, preds[['xmin', 'ymin', 'xmax',
|
||||
'ymax']].itertuples(index=False, name=None))
|
||||
logging.debug(
|
||||
'Finished drawing bounding boxes. Saving to current_bbox.jpg ...')
|
||||
cv2.imwrite('current_bbox.jpg', bbox_img)
|
||||
|
||||
# Clear superfluous bboxes if less detected
|
||||
# if len(preds) < len(pred):
|
||||
# logging.debug(
|
||||
# 'Current round contains less bboxes than previous round: old: %s\nnew: %s',
|
||||
# json.dumps(preds.copy()), json.dumps(pred.copy()))
|
||||
# for key in pred:
|
||||
# if key not in preds:
|
||||
# pred.pop(key)
|
||||
|
||||
pred.clear()
|
||||
for idx, row in preds.iterrows():
|
||||
new = []
|
||||
state = int(round(float(row['cls_conf']) / 10, 0))
|
||||
new.append(state)
|
||||
new.append(datetime.datetime.now(datetime.timezone.utc).isoformat())
|
||||
try:
|
||||
pred[str(idx)] = tmp[str(idx)]
|
||||
except KeyError:
|
||||
pass
|
||||
|
||||
try:
|
||||
if tmp[idx][2] == -1 and state > 3:
|
||||
logging.debug(
|
||||
'State is worse than 3 for the first time: %s ... populating third field',
|
||||
str(state))
|
||||
new.append(
|
||||
datetime.datetime.now(datetime.timezone.utc).isoformat())
|
||||
elif tmp[idx][2] != -1 and state <= 3:
|
||||
logging.debug(
|
||||
'State changed from worse than 3 to better than 3')
|
||||
new.append(-1)
|
||||
elif tmp[idx][2] != -1 and state > 3:
|
||||
logging.debug('State is still worse than 3')
|
||||
new.append(tmp[idx][2])
|
||||
except:
|
||||
logging.debug('Third key does not exist')
|
||||
if state > 3:
|
||||
logging.debug(
|
||||
'State is worse than 3. Populating third field with timestamp'
|
||||
)
|
||||
new.append(
|
||||
datetime.datetime.now(datetime.timezone.utc).isoformat())
|
||||
else:
|
||||
logging.debug(
|
||||
'State is better than 3. Populating third field with -1')
|
||||
new.append(-1)
|
||||
|
||||
pred[idx] = new
|
||||
|
||||
bbox_img_b64 = base64.b64encode(cv2.imencode('.jpg', bbox_img)[1]).decode()
|
||||
pred['image'] = bbox_img_b64
|
||||
logging.debug('Saved bbox_img to json')
|
||||
|
||||
|
||||
def take_image():
|
||||
"""Take an image with the webcam and save it to the specified
|
||||
path.
|
||||
|
||||
:param str img_path: path image should be saved to
|
||||
:returns: captured image
|
||||
"""
|
||||
capture_path = os.path.join('.', 'image-capture', 'capture')
|
||||
if os.path.isfile(capture_path) and os.access(capture_path, os.X_OK):
|
||||
logging.debug('Starting image capture')
|
||||
os.system('./image-capture/capture')
|
||||
logging.debug('Finished image capture')
|
||||
else:
|
||||
logging.critical(
|
||||
'Image capture binary is not at path %s. Shutting down server...',
|
||||
capture_path)
|
||||
app.terminate()
|
||||
|
||||
|
||||
@app.route('/')
|
||||
def index():
|
||||
copy = pred.copy()
|
||||
for key, value in copy.items():
|
||||
if key == 'image':
|
||||
continue
|
||||
if value[2] != -1:
|
||||
logging.debug('value: %s', value)
|
||||
# Calc difference to now
|
||||
time_below_thresh = parse(value[2])
|
||||
now = datetime.datetime.now(datetime.timezone.utc)
|
||||
delta = now - time_below_thresh
|
||||
value[2] = str(delta)
|
||||
return json.dumps(copy)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument(
|
||||
'--log',
|
||||
type=str,
|
||||
help='log level (debug, info, warning, error, critical)',
|
||||
default='warning')
|
||||
opt = parser.parse_args()
|
||||
numeric_level = getattr(logging, opt.log.upper(), None)
|
||||
logging.basicConfig(format='%(levelname)s::%(asctime)s::%(message)s',
|
||||
datefmt='%Y-%m-%dT%H:%M:%S',
|
||||
level=numeric_level)
|
||||
|
||||
scheduler.add_job(func=get_pred,
|
||||
trigger='interval',
|
||||
minutes=2,
|
||||
next_run_time=datetime.datetime.now())
|
||||
scheduler.start()
|
||||
atexit.register(scheduler.shutdown)
|
||||
|
||||
app.run()
|
||||
@@ -0,0 +1,24 @@
|
||||
flask==2.0.3
|
||||
apscheduler==3.10.0
|
||||
albumentations==1.3.0
|
||||
pandas==1.1.5
|
||||
onnxruntime==1.8.0
|
||||
opencv-python==4.7.0
|
||||
torch==1.8.1
|
||||
torchvision==0.9.1
|
||||
numpy==1.18.0
|
||||
scipy==1.1.0
|
||||
scikit-learn==0.21.3
|
||||
PIL==8.0.4
|
||||
argparse==1.1
|
||||
json==2.0.9
|
||||
matplotlib==3.3.4
|
||||
|
||||
# For labeling/conversion
|
||||
# label-studio==1.7.0
|
||||
|
||||
# For interactive evaluation
|
||||
# fiftyone
|
||||
|
||||
# Visualize gradients
|
||||
# pytorch-gradcam
|
||||
@@ -0,0 +1,8 @@
|
||||
from setuptools import setup, find_packages
|
||||
|
||||
setup(name='thesis',
|
||||
version='0.1.0',
|
||||
author='Tobias Eidelpes',
|
||||
author_email='e1527193@student.tuwien.ac.at',
|
||||
packages=find_packages(),
|
||||
description='Flower State Classification')
|
||||
File diff suppressed because one or more lines are too long
@@ -0,0 +1,169 @@
|
||||
from __future__ import print_function, division
|
||||
|
||||
import torch
|
||||
import torch.nn as nn
|
||||
import torch.optim as optim
|
||||
from torch.optim import lr_scheduler
|
||||
import torch.backends.cudnn as cudnn
|
||||
import numpy as np
|
||||
import torchvision
|
||||
from torchvision import datasets, models, transforms
|
||||
import matplotlib.pyplot as plt
|
||||
import time
|
||||
import os
|
||||
import copy
|
||||
|
||||
cudnn.benchmark = True
|
||||
plt.ion() # interactive mode
|
||||
|
||||
data_transforms = {
|
||||
'train': transforms.Compose([
|
||||
transforms.RandomResizedCrop(224),
|
||||
transforms.RandomHorizontalFlip(),
|
||||
transforms.ToTensor(),
|
||||
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
|
||||
]),
|
||||
'val': transforms.Compose([
|
||||
transforms.Resize(256),
|
||||
transforms.CenterCrop(224),
|
||||
transforms.ToTensor(),
|
||||
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
|
||||
]),
|
||||
}
|
||||
|
||||
data_dir = 'hymenoptera_data'
|
||||
image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),
|
||||
data_transforms[x])
|
||||
for x in ['train', 'val']}
|
||||
dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4,
|
||||
shuffle=True, num_workers=4)
|
||||
for x in ['train', 'val']}
|
||||
dataset_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}
|
||||
class_names = image_datasets['train'].classes
|
||||
|
||||
device = torch.device("cpu")
|
||||
|
||||
def imshow(inp, title=None):
|
||||
"""Imshow for Tensor."""
|
||||
inp = inp.numpy().transpose((1, 2, 0))
|
||||
mean = np.array([0.485, 0.456, 0.406])
|
||||
std = np.array([0.229, 0.224, 0.225])
|
||||
inp = std * inp + mean
|
||||
inp = np.clip(inp, 0, 1)
|
||||
plt.imshow(inp)
|
||||
if title is not None:
|
||||
plt.title(title)
|
||||
plt.pause(0.001) # pause a bit so that plots are updated
|
||||
|
||||
def train_model(model, criterion, optimizer, scheduler, num_epochs=25):
|
||||
since = time.time()
|
||||
|
||||
best_model_wts = copy.deepcopy(model.state_dict())
|
||||
best_acc = 0.0
|
||||
|
||||
for epoch in range(num_epochs):
|
||||
print(f'Epoch {epoch}/{num_epochs - 1}')
|
||||
print('-' * 10)
|
||||
|
||||
# Each epoch has a training and validation phase
|
||||
for phase in ['train', 'val']:
|
||||
if phase == 'train':
|
||||
model.train() # Set model to training mode
|
||||
else:
|
||||
model.eval() # Set model to evaluate mode
|
||||
|
||||
running_loss = 0.0
|
||||
running_corrects = 0
|
||||
|
||||
# Iterate over data.
|
||||
for inputs, labels in dataloaders[phase]:
|
||||
inputs = inputs.to(device)
|
||||
labels = labels.to(device)
|
||||
|
||||
# zero the parameter gradients
|
||||
optimizer.zero_grad()
|
||||
|
||||
# forward
|
||||
# track history if only in train
|
||||
with torch.set_grad_enabled(phase == 'train'):
|
||||
outputs = model(inputs)
|
||||
_, preds = torch.max(outputs, 1)
|
||||
loss = criterion(outputs, labels)
|
||||
|
||||
# backward + optimize only if in training phase
|
||||
if phase == 'train':
|
||||
loss.backward()
|
||||
optimizer.step()
|
||||
|
||||
# statistics
|
||||
running_loss += loss.item() * inputs.size(0)
|
||||
running_corrects += torch.sum(preds == labels.data)
|
||||
if phase == 'train':
|
||||
scheduler.step()
|
||||
|
||||
epoch_loss = running_loss / dataset_sizes[phase]
|
||||
epoch_acc = running_corrects.double() / dataset_sizes[phase]
|
||||
|
||||
print(f'{phase} Loss: {epoch_loss:.4f} Acc: {epoch_acc:.4f}')
|
||||
|
||||
# deep copy the model
|
||||
if phase == 'val' and epoch_acc > best_acc:
|
||||
best_acc = epoch_acc
|
||||
best_model_wts = copy.deepcopy(model.state_dict())
|
||||
|
||||
print()
|
||||
|
||||
time_elapsed = time.time() - since
|
||||
print(f'Training complete in {time_elapsed // 60:.0f}m {time_elapsed % 60:.0f}s')
|
||||
print(f'Best val Acc: {best_acc:4f}')
|
||||
|
||||
# load best model weights
|
||||
model.load_state_dict(best_model_wts)
|
||||
return model
|
||||
|
||||
def visualize_model(model, num_images=6):
|
||||
was_training = model.training
|
||||
model.eval()
|
||||
images_so_far = 0
|
||||
fig = plt.figure()
|
||||
|
||||
with torch.no_grad():
|
||||
for i, (inputs, labels) in enumerate(dataloaders['val']):
|
||||
inputs = inputs.to(device)
|
||||
labels = labels.to(device)
|
||||
|
||||
outputs = model(inputs)
|
||||
_, preds = torch.max(outputs, 1)
|
||||
|
||||
for j in range(inputs.size()[0]):
|
||||
images_so_far += 1
|
||||
ax = plt.subplot(num_images//2, 2, images_so_far)
|
||||
ax.axis('off')
|
||||
ax.set_title(f'predicted: {class_names[preds[j]]}')
|
||||
imshow(inputs.cpu().data[j])
|
||||
|
||||
if images_so_far == num_images:
|
||||
model.train(mode=was_training)
|
||||
return
|
||||
model.train(mode=was_training)
|
||||
|
||||
model_ft = models.resnet18(pretrained=True)
|
||||
num_ftrs = model_ft.fc.in_features
|
||||
# Here the size of each output sample is set to 2.
|
||||
# Alternatively, it can be generalized to nn.Linear(num_ftrs, len(class_names)).
|
||||
model_ft.fc = nn.Linear(num_ftrs, 2)
|
||||
|
||||
model_ft = model_ft.to(device)
|
||||
|
||||
criterion = nn.CrossEntropyLoss()
|
||||
|
||||
# Observe that all parameters are being optimized
|
||||
optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)
|
||||
|
||||
# Decay LR by a factor of 0.1 every 7 epochs
|
||||
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)
|
||||
|
||||
model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler,
|
||||
num_epochs=25)
|
||||
|
||||
visualize_model(model_ft)
|
||||
@@ -0,0 +1,42 @@
|
||||
def convert_to_yolo(bbox, width, height):
|
||||
modified = bbox.copy()
|
||||
modified['xmin%'] = round(bbox['xmin'] / width * 100, 2)
|
||||
modified['ymin%'] = round(bbox['ymin'] / height * 100, 2)
|
||||
modified['width%'] = round((bbox['xmax'] - bbox['xmin']) / width * 100, 2)
|
||||
modified['height%'] = round((bbox['ymax'] - bbox['ymin']) / height * 100,
|
||||
2)
|
||||
return modified
|
||||
|
||||
|
||||
def scale_bboxes(bboxes, resized_hw, original_hw):
|
||||
"""Scale bounding boxes from a padded and resized image to fit on
|
||||
original image.
|
||||
|
||||
:param xyxy_boxes Tensor[N, 4]: tensor of xmin, ymin, xmax, ymax
|
||||
per bounding box
|
||||
:param resized_hw Tuple: height and width of the resized image
|
||||
:param original_hw Tuple: height and width of the original image
|
||||
:returns Tensor[N, 4]: tensor of xmin, ymin, xmax, ymax per
|
||||
bounding box
|
||||
"""
|
||||
scaled_boxes = bboxes.clone()
|
||||
scale_ratio = resized_hw[0] / original_hw[0], resized_hw[1] / original_hw[1]
|
||||
|
||||
# Remove padding
|
||||
pad_scale = min(scale_ratio)
|
||||
padding = (resized_hw[1] - original_hw[1] * pad_scale) / 2, (
|
||||
resized_hw[0] - original_hw[0] * pad_scale) / 2
|
||||
scaled_boxes[:, [0, 2]] -= padding[0] # x padding
|
||||
scaled_boxes[:, [1, 3]] -= padding[1] # y padding
|
||||
scale_ratio = (pad_scale, pad_scale)
|
||||
|
||||
scaled_boxes[:, [0, 2]] /= scale_ratio[1]
|
||||
scaled_boxes[:, [1, 3]] /= scale_ratio[0]
|
||||
|
||||
# Clip bounding xyxy bounding boxes to image shape (height, width)
|
||||
scaled_boxes[:, 0].clamp_(0, original_hw[1]) # xmin
|
||||
scaled_boxes[:, 1].clamp_(0, original_hw[0]) # ymin
|
||||
scaled_boxes[:, 2].clamp_(0, original_hw[1]) # xmax
|
||||
scaled_boxes[:, 3].clamp_(0, original_hw[0]) # ymax
|
||||
|
||||
return scaled_boxes
|
||||
@@ -0,0 +1,32 @@
|
||||
import cv2
|
||||
|
||||
|
||||
def draw_boxes(image, bboxes):
|
||||
img = image.copy()
|
||||
for idx, bbox in enumerate(bboxes):
|
||||
xmin, ymin, xmax, ymax = bbox
|
||||
xmin, ymin, xmax, ymax = int(xmin), int(ymin), int(xmax), int(ymax)
|
||||
# Draw bounding box and number on original image
|
||||
img = cv2.rectangle(img, (xmin, ymin), (xmax, ymax), (0, 255, 0), 2)
|
||||
img = cv2.putText(img, str(idx), (xmin + 5, ymin + 25),
|
||||
cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 0, 0), 4,
|
||||
cv2.LINE_AA)
|
||||
img = cv2.putText(img, str(idx), (xmin + 5, ymin + 25),
|
||||
cv2.FONT_HERSHEY_SIMPLEX, 0.9, (255, 255, 255), 2,
|
||||
cv2.LINE_AA)
|
||||
return img
|
||||
|
||||
|
||||
def get_cutout(img, xmin, xmax, ymin, ymax):
|
||||
"""Cut out a bounding box from an image and transform it for
|
||||
object classification model.
|
||||
|
||||
:param img: opencv2 image object in BGR
|
||||
:param int xmin: start of bounding box on x axis
|
||||
:param int xmax: end of bounding box on x axis
|
||||
:param int ymin: start of bounding box on y axis
|
||||
:param int ymax: end of bounding box on y axis
|
||||
:returns: tensor of cropped image in BGR
|
||||
"""
|
||||
cropped_image = img[ymin:ymax, xmin:xmax]
|
||||
return cropped_image
|
||||
@@ -0,0 +1,140 @@
|
||||
# parameters
|
||||
nc: 1 # number of classes
|
||||
depth_multiple: 1.0 # model depth multiple
|
||||
width_multiple: 1.0 # layer channel multiple
|
||||
|
||||
# anchors
|
||||
anchors:
|
||||
- [12,16, 19,36, 40,28] # P3/8
|
||||
- [36,75, 76,55, 72,146] # P4/16
|
||||
- [142,110, 192,243, 459,401] # P5/32
|
||||
|
||||
# yolov7 backbone
|
||||
backbone:
|
||||
# [from, number, module, args]
|
||||
[[-1, 1, Conv, [32, 3, 1]], # 0
|
||||
|
||||
[-1, 1, Conv, [64, 3, 2]], # 1-P1/2
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
|
||||
[-1, 1, Conv, [128, 3, 2]], # 3-P2/4
|
||||
[-1, 1, Conv, [64, 1, 1]],
|
||||
[-2, 1, Conv, [64, 1, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[[-1, -3, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [256, 1, 1]], # 11
|
||||
|
||||
[-1, 1, MP, []],
|
||||
[-1, 1, Conv, [128, 1, 1]],
|
||||
[-3, 1, Conv, [128, 1, 1]],
|
||||
[-1, 1, Conv, [128, 3, 2]],
|
||||
[[-1, -3], 1, Concat, [1]], # 16-P3/8
|
||||
[-1, 1, Conv, [128, 1, 1]],
|
||||
[-2, 1, Conv, [128, 1, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[[-1, -3, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [512, 1, 1]], # 24
|
||||
|
||||
[-1, 1, MP, []],
|
||||
[-1, 1, Conv, [256, 1, 1]],
|
||||
[-3, 1, Conv, [256, 1, 1]],
|
||||
[-1, 1, Conv, [256, 3, 2]],
|
||||
[[-1, -3], 1, Concat, [1]], # 29-P4/16
|
||||
[-1, 1, Conv, [256, 1, 1]],
|
||||
[-2, 1, Conv, [256, 1, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[[-1, -3, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [1024, 1, 1]], # 37
|
||||
|
||||
[-1, 1, MP, []],
|
||||
[-1, 1, Conv, [512, 1, 1]],
|
||||
[-3, 1, Conv, [512, 1, 1]],
|
||||
[-1, 1, Conv, [512, 3, 2]],
|
||||
[[-1, -3], 1, Concat, [1]], # 42-P5/32
|
||||
[-1, 1, Conv, [256, 1, 1]],
|
||||
[-2, 1, Conv, [256, 1, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[[-1, -3, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [1024, 1, 1]], # 50
|
||||
]
|
||||
|
||||
# yolov7 head
|
||||
head:
|
||||
[[-1, 1, SPPCSPC, [512]], # 51
|
||||
|
||||
[-1, 1, Conv, [256, 1, 1]],
|
||||
[-1, 1, nn.Upsample, [None, 2, 'nearest']],
|
||||
[37, 1, Conv, [256, 1, 1]], # route backbone P4
|
||||
[[-1, -2], 1, Concat, [1]],
|
||||
|
||||
[-1, 1, Conv, [256, 1, 1]],
|
||||
[-2, 1, Conv, [256, 1, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[[-1, -2, -3, -4, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [256, 1, 1]], # 63
|
||||
|
||||
[-1, 1, Conv, [128, 1, 1]],
|
||||
[-1, 1, nn.Upsample, [None, 2, 'nearest']],
|
||||
[24, 1, Conv, [128, 1, 1]], # route backbone P3
|
||||
[[-1, -2], 1, Concat, [1]],
|
||||
|
||||
[-1, 1, Conv, [128, 1, 1]],
|
||||
[-2, 1, Conv, [128, 1, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[-1, 1, Conv, [64, 3, 1]],
|
||||
[[-1, -2, -3, -4, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [128, 1, 1]], # 75
|
||||
|
||||
[-1, 1, MP, []],
|
||||
[-1, 1, Conv, [128, 1, 1]],
|
||||
[-3, 1, Conv, [128, 1, 1]],
|
||||
[-1, 1, Conv, [128, 3, 2]],
|
||||
[[-1, -3, 63], 1, Concat, [1]],
|
||||
|
||||
[-1, 1, Conv, [256, 1, 1]],
|
||||
[-2, 1, Conv, [256, 1, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[-1, 1, Conv, [128, 3, 1]],
|
||||
[[-1, -2, -3, -4, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [256, 1, 1]], # 88
|
||||
|
||||
[-1, 1, MP, []],
|
||||
[-1, 1, Conv, [256, 1, 1]],
|
||||
[-3, 1, Conv, [256, 1, 1]],
|
||||
[-1, 1, Conv, [256, 3, 2]],
|
||||
[[-1, -3, 51], 1, Concat, [1]],
|
||||
|
||||
[-1, 1, Conv, [512, 1, 1]],
|
||||
[-2, 1, Conv, [512, 1, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[-1, 1, Conv, [256, 3, 1]],
|
||||
[[-1, -2, -3, -4, -5, -6], 1, Concat, [1]],
|
||||
[-1, 1, Conv, [512, 1, 1]], # 101
|
||||
|
||||
[75, 1, RepConv, [256, 3, 1]],
|
||||
[88, 1, RepConv, [512, 3, 1]],
|
||||
[101, 1, RepConv, [1024, 3, 1]],
|
||||
|
||||
[[102,103,104], 1, IDetect, [nc, anchors]], # Detect(P3, P4, P5)
|
||||
]
|
||||
@@ -0,0 +1,196 @@
|
||||
import argparse
|
||||
import time
|
||||
from pathlib import Path
|
||||
|
||||
import cv2
|
||||
import torch
|
||||
import torch.backends.cudnn as cudnn
|
||||
from numpy import random
|
||||
|
||||
from models.experimental import attempt_load
|
||||
from utils.datasets import LoadStreams, LoadImages
|
||||
from utils.general import check_img_size, check_requirements, check_imshow, non_max_suppression, apply_classifier, \
|
||||
scale_coords, xyxy2xywh, strip_optimizer, set_logging, increment_path
|
||||
from utils.plots import plot_one_box
|
||||
from utils.torch_utils import select_device, load_classifier, time_synchronized, TracedModel
|
||||
|
||||
|
||||
def detect(save_img=False):
|
||||
source, weights, view_img, save_txt, imgsz, trace = opt.source, opt.weights, opt.view_img, opt.save_txt, opt.img_size, not opt.no_trace
|
||||
save_img = not opt.nosave and not source.endswith('.txt') # save inference images
|
||||
webcam = source.isnumeric() or source.endswith('.txt') or source.lower().startswith(
|
||||
('rtsp://', 'rtmp://', 'http://', 'https://'))
|
||||
|
||||
# Directories
|
||||
save_dir = Path(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok)) # increment run
|
||||
(save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir
|
||||
|
||||
# Initialize
|
||||
set_logging()
|
||||
device = select_device(opt.device)
|
||||
half = device.type != 'cpu' # half precision only supported on CUDA
|
||||
|
||||
# Load model
|
||||
model = attempt_load(weights, map_location=device) # load FP32 model
|
||||
stride = int(model.stride.max()) # model stride
|
||||
imgsz = check_img_size(imgsz, s=stride) # check img_size
|
||||
|
||||
if trace:
|
||||
model = TracedModel(model, device, opt.img_size)
|
||||
|
||||
if half:
|
||||
model.half() # to FP16
|
||||
|
||||
# Second-stage classifier
|
||||
classify = False
|
||||
if classify:
|
||||
modelc = load_classifier(name='resnet101', n=2) # initialize
|
||||
modelc.load_state_dict(torch.load('weights/resnet101.pt', map_location=device)['model']).to(device).eval()
|
||||
|
||||
# Set Dataloader
|
||||
vid_path, vid_writer = None, None
|
||||
if webcam:
|
||||
view_img = check_imshow()
|
||||
cudnn.benchmark = True # set True to speed up constant image size inference
|
||||
dataset = LoadStreams(source, img_size=imgsz, stride=stride)
|
||||
else:
|
||||
dataset = LoadImages(source, img_size=imgsz, stride=stride)
|
||||
|
||||
# Get names and colors
|
||||
names = model.module.names if hasattr(model, 'module') else model.names
|
||||
colors = [[random.randint(0, 255) for _ in range(3)] for _ in names]
|
||||
|
||||
# Run inference
|
||||
if device.type != 'cpu':
|
||||
model(torch.zeros(1, 3, imgsz, imgsz).to(device).type_as(next(model.parameters()))) # run once
|
||||
old_img_w = old_img_h = imgsz
|
||||
old_img_b = 1
|
||||
|
||||
t0 = time.time()
|
||||
for path, img, im0s, vid_cap in dataset:
|
||||
img = torch.from_numpy(img).to(device)
|
||||
img = img.half() if half else img.float() # uint8 to fp16/32
|
||||
img /= 255.0 # 0 - 255 to 0.0 - 1.0
|
||||
if img.ndimension() == 3:
|
||||
img = img.unsqueeze(0)
|
||||
|
||||
# Warmup
|
||||
if device.type != 'cpu' and (old_img_b != img.shape[0] or old_img_h != img.shape[2] or old_img_w != img.shape[3]):
|
||||
old_img_b = img.shape[0]
|
||||
old_img_h = img.shape[2]
|
||||
old_img_w = img.shape[3]
|
||||
for i in range(3):
|
||||
model(img, augment=opt.augment)[0]
|
||||
|
||||
# Inference
|
||||
t1 = time_synchronized()
|
||||
with torch.no_grad(): # Calculating gradients would cause a GPU memory leak
|
||||
pred = model(img, augment=opt.augment)[0]
|
||||
t2 = time_synchronized()
|
||||
|
||||
# Apply NMS
|
||||
pred = non_max_suppression(pred, opt.conf_thres, opt.iou_thres, classes=opt.classes, agnostic=opt.agnostic_nms)
|
||||
t3 = time_synchronized()
|
||||
|
||||
# Apply Classifier
|
||||
if classify:
|
||||
pred = apply_classifier(pred, modelc, img, im0s)
|
||||
|
||||
# Process detections
|
||||
for i, det in enumerate(pred): # detections per image
|
||||
if webcam: # batch_size >= 1
|
||||
p, s, im0, frame = path[i], '%g: ' % i, im0s[i].copy(), dataset.count
|
||||
else:
|
||||
p, s, im0, frame = path, '', im0s, getattr(dataset, 'frame', 0)
|
||||
|
||||
p = Path(p) # to Path
|
||||
save_path = str(save_dir / p.name) # img.jpg
|
||||
txt_path = str(save_dir / 'labels' / p.stem) + ('' if dataset.mode == 'image' else f'_{frame}') # img.txt
|
||||
gn = torch.tensor(im0.shape)[[1, 0, 1, 0]] # normalization gain whwh
|
||||
if len(det):
|
||||
# Rescale boxes from img_size to im0 size
|
||||
det[:, :4] = scale_coords(img.shape[2:], det[:, :4], im0.shape).round()
|
||||
|
||||
# Print results
|
||||
for c in det[:, -1].unique():
|
||||
n = (det[:, -1] == c).sum() # detections per class
|
||||
s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string
|
||||
|
||||
# Write results
|
||||
for *xyxy, conf, cls in reversed(det):
|
||||
if save_txt: # Write to file
|
||||
xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh
|
||||
line = (cls, *xywh, conf) if opt.save_conf else (cls, *xywh) # label format
|
||||
with open(txt_path + '.txt', 'a') as f:
|
||||
f.write(('%g ' * len(line)).rstrip() % line + '\n')
|
||||
|
||||
if save_img or view_img: # Add bbox to image
|
||||
label = f'{names[int(cls)]} {conf:.2f}'
|
||||
plot_one_box(xyxy, im0, label=label, color=colors[int(cls)], line_thickness=1)
|
||||
|
||||
# Print time (inference + NMS)
|
||||
print(f'{s}Done. ({(1E3 * (t2 - t1)):.1f}ms) Inference, ({(1E3 * (t3 - t2)):.1f}ms) NMS')
|
||||
|
||||
# Stream results
|
||||
if view_img:
|
||||
cv2.imshow(str(p), im0)
|
||||
cv2.waitKey(1) # 1 millisecond
|
||||
|
||||
# Save results (image with detections)
|
||||
if save_img:
|
||||
if dataset.mode == 'image':
|
||||
cv2.imwrite(save_path, im0)
|
||||
print(f" The image with the result is saved in: {save_path}")
|
||||
else: # 'video' or 'stream'
|
||||
if vid_path != save_path: # new video
|
||||
vid_path = save_path
|
||||
if isinstance(vid_writer, cv2.VideoWriter):
|
||||
vid_writer.release() # release previous video writer
|
||||
if vid_cap: # video
|
||||
fps = vid_cap.get(cv2.CAP_PROP_FPS)
|
||||
w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH))
|
||||
h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
|
||||
else: # stream
|
||||
fps, w, h = 30, im0.shape[1], im0.shape[0]
|
||||
save_path += '.mp4'
|
||||
vid_writer = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h))
|
||||
vid_writer.write(im0)
|
||||
|
||||
if save_txt or save_img:
|
||||
s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else ''
|
||||
#print(f"Results saved to {save_dir}{s}")
|
||||
|
||||
print(f'Done. ({time.time() - t0:.3f}s)')
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--weights', nargs='+', type=str, default='yolov7.pt', help='model.pt path(s)')
|
||||
parser.add_argument('--source', type=str, default='inference/images', help='source') # file/folder, 0 for webcam
|
||||
parser.add_argument('--img-size', type=int, default=640, help='inference size (pixels)')
|
||||
parser.add_argument('--conf-thres', type=float, default=0.25, help='object confidence threshold')
|
||||
parser.add_argument('--iou-thres', type=float, default=0.45, help='IOU threshold for NMS')
|
||||
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
|
||||
parser.add_argument('--view-img', action='store_true', help='display results')
|
||||
parser.add_argument('--save-txt', action='store_true', help='save results to *.txt')
|
||||
parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels')
|
||||
parser.add_argument('--nosave', action='store_true', help='do not save images/videos')
|
||||
parser.add_argument('--classes', nargs='+', type=int, help='filter by class: --class 0, or --class 0 2 3')
|
||||
parser.add_argument('--agnostic-nms', action='store_true', help='class-agnostic NMS')
|
||||
parser.add_argument('--augment', action='store_true', help='augmented inference')
|
||||
parser.add_argument('--update', action='store_true', help='update all models')
|
||||
parser.add_argument('--project', default='runs/detect', help='save results to project/name')
|
||||
parser.add_argument('--name', default='exp', help='save results to project/name')
|
||||
parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
|
||||
parser.add_argument('--no-trace', action='store_true', help='don`t trace model')
|
||||
opt = parser.parse_args()
|
||||
print(opt)
|
||||
#check_requirements(exclude=('pycocotools', 'thop'))
|
||||
|
||||
with torch.no_grad():
|
||||
if opt.update: # update all models (to fix SourceChangeWarning)
|
||||
for opt.weights in ['yolov7.pt']:
|
||||
detect()
|
||||
strip_optimizer(opt.weights)
|
||||
else:
|
||||
detect()
|
||||
@@ -0,0 +1,205 @@
|
||||
import argparse
|
||||
import sys
|
||||
import time
|
||||
import warnings
|
||||
|
||||
sys.path.append('./') # to run '$ python *.py' files in subdirectories
|
||||
|
||||
import torch
|
||||
import torch.nn as nn
|
||||
from torch.utils.mobile_optimizer import optimize_for_mobile
|
||||
|
||||
import models
|
||||
from models.experimental import attempt_load, End2End
|
||||
from utils.activations import Hardswish, SiLU
|
||||
from utils.general import set_logging, check_img_size
|
||||
from utils.torch_utils import select_device
|
||||
from utils.add_nms import RegisterNMS
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--weights', type=str, default='./yolor-csp-c.pt', help='weights path')
|
||||
parser.add_argument('--img-size', nargs='+', type=int, default=[640, 640], help='image size') # height, width
|
||||
parser.add_argument('--batch-size', type=int, default=1, help='batch size')
|
||||
parser.add_argument('--dynamic', action='store_true', help='dynamic ONNX axes')
|
||||
parser.add_argument('--dynamic-batch', action='store_true', help='dynamic batch onnx for tensorrt and onnx-runtime')
|
||||
parser.add_argument('--grid', action='store_true', help='export Detect() layer grid')
|
||||
parser.add_argument('--end2end', action='store_true', help='export end2end onnx')
|
||||
parser.add_argument('--max-wh', type=int, default=None, help='None for tensorrt nms, int value for onnx-runtime nms')
|
||||
parser.add_argument('--topk-all', type=int, default=100, help='topk objects for every images')
|
||||
parser.add_argument('--iou-thres', type=float, default=0.45, help='iou threshold for NMS')
|
||||
parser.add_argument('--conf-thres', type=float, default=0.25, help='conf threshold for NMS')
|
||||
parser.add_argument('--device', default='cpu', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
|
||||
parser.add_argument('--simplify', action='store_true', help='simplify onnx model')
|
||||
parser.add_argument('--include-nms', action='store_true', help='export end2end onnx')
|
||||
parser.add_argument('--fp16', action='store_true', help='CoreML FP16 half-precision export')
|
||||
parser.add_argument('--int8', action='store_true', help='CoreML INT8 quantization')
|
||||
opt = parser.parse_args()
|
||||
opt.img_size *= 2 if len(opt.img_size) == 1 else 1 # expand
|
||||
opt.dynamic = opt.dynamic and not opt.end2end
|
||||
opt.dynamic = False if opt.dynamic_batch else opt.dynamic
|
||||
print(opt)
|
||||
set_logging()
|
||||
t = time.time()
|
||||
|
||||
# Load PyTorch model
|
||||
device = select_device(opt.device)
|
||||
model = attempt_load(opt.weights, map_location=device) # load FP32 model
|
||||
labels = model.names
|
||||
|
||||
# Checks
|
||||
gs = int(max(model.stride)) # grid size (max stride)
|
||||
opt.img_size = [check_img_size(x, gs) for x in opt.img_size] # verify img_size are gs-multiples
|
||||
|
||||
# Input
|
||||
img = torch.zeros(opt.batch_size, 3, *opt.img_size).to(device) # image size(1,3,320,192) iDetection
|
||||
|
||||
# Update model
|
||||
for k, m in model.named_modules():
|
||||
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility
|
||||
if isinstance(m, models.common.Conv): # assign export-friendly activations
|
||||
if isinstance(m.act, nn.Hardswish):
|
||||
m.act = Hardswish()
|
||||
elif isinstance(m.act, nn.SiLU):
|
||||
m.act = SiLU()
|
||||
# elif isinstance(m, models.yolo.Detect):
|
||||
# m.forward = m.forward_export # assign forward (optional)
|
||||
model.model[-1].export = not opt.grid # set Detect() layer grid export
|
||||
y = model(img) # dry run
|
||||
if opt.include_nms:
|
||||
model.model[-1].include_nms = True
|
||||
y = None
|
||||
|
||||
# TorchScript export
|
||||
try:
|
||||
print('\nStarting TorchScript export with torch %s...' % torch.__version__)
|
||||
f = opt.weights.replace('.pt', '.torchscript.pt') # filename
|
||||
ts = torch.jit.trace(model, img, strict=False)
|
||||
ts.save(f)
|
||||
print('TorchScript export success, saved as %s' % f)
|
||||
except Exception as e:
|
||||
print('TorchScript export failure: %s' % e)
|
||||
|
||||
# CoreML export
|
||||
try:
|
||||
import coremltools as ct
|
||||
|
||||
print('\nStarting CoreML export with coremltools %s...' % ct.__version__)
|
||||
# convert model from torchscript and apply pixel scaling as per detect.py
|
||||
ct_model = ct.convert(ts, inputs=[ct.ImageType('image', shape=img.shape, scale=1 / 255.0, bias=[0, 0, 0])])
|
||||
bits, mode = (8, 'kmeans_lut') if opt.int8 else (16, 'linear') if opt.fp16 else (32, None)
|
||||
if bits < 32:
|
||||
if sys.platform.lower() == 'darwin': # quantization only supported on macOS
|
||||
with warnings.catch_warnings():
|
||||
warnings.filterwarnings("ignore", category=DeprecationWarning) # suppress numpy==1.20 float warning
|
||||
ct_model = ct.models.neural_network.quantization_utils.quantize_weights(ct_model, bits, mode)
|
||||
else:
|
||||
print('quantization only supported on macOS, skipping...')
|
||||
|
||||
f = opt.weights.replace('.pt', '.mlmodel') # filename
|
||||
ct_model.save(f)
|
||||
print('CoreML export success, saved as %s' % f)
|
||||
except Exception as e:
|
||||
print('CoreML export failure: %s' % e)
|
||||
|
||||
# TorchScript-Lite export
|
||||
try:
|
||||
print('\nStarting TorchScript-Lite export with torch %s...' % torch.__version__)
|
||||
f = opt.weights.replace('.pt', '.torchscript.ptl') # filename
|
||||
tsl = torch.jit.trace(model, img, strict=False)
|
||||
tsl = optimize_for_mobile(tsl)
|
||||
tsl._save_for_lite_interpreter(f)
|
||||
print('TorchScript-Lite export success, saved as %s' % f)
|
||||
except Exception as e:
|
||||
print('TorchScript-Lite export failure: %s' % e)
|
||||
|
||||
# ONNX export
|
||||
try:
|
||||
import onnx
|
||||
|
||||
print('\nStarting ONNX export with onnx %s...' % onnx.__version__)
|
||||
f = opt.weights.replace('.pt', '.onnx') # filename
|
||||
model.eval()
|
||||
output_names = ['classes', 'boxes'] if y is None else ['output']
|
||||
dynamic_axes = None
|
||||
if opt.dynamic:
|
||||
dynamic_axes = {'images': {0: 'batch', 2: 'height', 3: 'width'}, # size(1,3,640,640)
|
||||
'output': {0: 'batch', 2: 'y', 3: 'x'}}
|
||||
if opt.dynamic_batch:
|
||||
opt.batch_size = 'batch'
|
||||
dynamic_axes = {
|
||||
'images': {
|
||||
0: 'batch',
|
||||
}, }
|
||||
if opt.end2end and opt.max_wh is None:
|
||||
output_axes = {
|
||||
'num_dets': {0: 'batch'},
|
||||
'det_boxes': {0: 'batch'},
|
||||
'det_scores': {0: 'batch'},
|
||||
'det_classes': {0: 'batch'},
|
||||
}
|
||||
else:
|
||||
output_axes = {
|
||||
'output': {0: 'batch'},
|
||||
}
|
||||
dynamic_axes.update(output_axes)
|
||||
if opt.grid:
|
||||
if opt.end2end:
|
||||
print('\nStarting export end2end onnx model for %s...' % 'TensorRT' if opt.max_wh is None else 'onnxruntime')
|
||||
model = End2End(model,opt.topk_all,opt.iou_thres,opt.conf_thres,opt.max_wh,device,len(labels))
|
||||
if opt.end2end and opt.max_wh is None:
|
||||
output_names = ['num_dets', 'det_boxes', 'det_scores', 'det_classes']
|
||||
shapes = [opt.batch_size, 1, opt.batch_size, opt.topk_all, 4,
|
||||
opt.batch_size, opt.topk_all, opt.batch_size, opt.topk_all]
|
||||
else:
|
||||
output_names = ['output']
|
||||
else:
|
||||
model.model[-1].concat = True
|
||||
|
||||
torch.onnx.export(model, img, f, verbose=False, opset_version=12, input_names=['images'],
|
||||
output_names=output_names,
|
||||
dynamic_axes=dynamic_axes)
|
||||
|
||||
# Checks
|
||||
onnx_model = onnx.load(f) # load onnx model
|
||||
onnx.checker.check_model(onnx_model) # check onnx model
|
||||
|
||||
if opt.end2end and opt.max_wh is None:
|
||||
for i in onnx_model.graph.output:
|
||||
for j in i.type.tensor_type.shape.dim:
|
||||
j.dim_param = str(shapes.pop(0))
|
||||
|
||||
# print(onnx.helper.printable_graph(onnx_model.graph)) # print a human readable model
|
||||
|
||||
# # Metadata
|
||||
# d = {'stride': int(max(model.stride))}
|
||||
# for k, v in d.items():
|
||||
# meta = onnx_model.metadata_props.add()
|
||||
# meta.key, meta.value = k, str(v)
|
||||
# onnx.save(onnx_model, f)
|
||||
|
||||
if opt.simplify:
|
||||
try:
|
||||
import onnxsim
|
||||
|
||||
print('\nStarting to simplify ONNX...')
|
||||
onnx_model, check = onnxsim.simplify(onnx_model)
|
||||
assert check, 'assert check failed'
|
||||
except Exception as e:
|
||||
print(f'Simplifier failure: {e}')
|
||||
|
||||
# print(onnx.helper.printable_graph(onnx_model.graph)) # print a human readable model
|
||||
onnx.save(onnx_model,f)
|
||||
print('ONNX export success, saved as %s' % f)
|
||||
|
||||
if opt.include_nms:
|
||||
print('Registering NMS plugin for ONNX...')
|
||||
mo = RegisterNMS(f)
|
||||
mo.register_nms()
|
||||
mo.save(f)
|
||||
|
||||
except Exception as e:
|
||||
print('ONNX export failure: %s' % e)
|
||||
|
||||
# Finish
|
||||
print('\nExport complete (%.2fs). Visualize with https://github.com/lutzroeder/netron.' % (time.time() - t))
|
||||
@@ -0,0 +1,31 @@
|
||||
lr0: 0.01 # initial learning rate (SGD=1E-2, Adam=1E-3)
|
||||
lrf: 0.1 # final OneCycleLR learning rate (lr0 * lrf)
|
||||
momentum: 0.937 # SGD momentum/Adam beta1
|
||||
weight_decay: 0.0005 # optimizer weight decay 5e-4
|
||||
warmup_epochs: 3.0 # warmup epochs (fractions ok)
|
||||
warmup_momentum: 0.8 # warmup initial momentum
|
||||
warmup_bias_lr: 0.1 # warmup initial bias lr
|
||||
box: 0.05 # box loss gain
|
||||
cls: 0.3 # cls loss gain
|
||||
cls_pw: 1.0 # cls BCELoss positive_weight
|
||||
obj: 0.7 # obj loss gain (scale with pixels)
|
||||
obj_pw: 1.0 # obj BCELoss positive_weight
|
||||
iou_t: 0.20 # IoU training threshold
|
||||
anchor_t: 4.0 # anchor-multiple threshold
|
||||
# anchors: 3 # anchors per output layer (0 to ignore)
|
||||
fl_gamma: 0.0 # focal loss gamma (efficientDet default gamma=1.5)
|
||||
hsv_h: 0.015 # image HSV-Hue augmentation (fraction)
|
||||
hsv_s: 0.7 # image HSV-Saturation augmentation (fraction)
|
||||
hsv_v: 0.4 # image HSV-Value augmentation (fraction)
|
||||
degrees: 0.0 # image rotation (+/- deg)
|
||||
translate: 0.2 # image translation (+/- fraction)
|
||||
scale: 0.5 # image scale (+/- gain)
|
||||
shear: 0.0 # image shear (+/- deg)
|
||||
perspective: 0.0 # image perspective (+/- fraction), range 0-0.001
|
||||
flipud: 0.0 # image flip up-down (probability)
|
||||
fliplr: 0.5 # image flip left-right (probability)
|
||||
mosaic: 1.0 # image mosaic (probability)
|
||||
mixup: 0.0 # image mixup (probability)
|
||||
copy_paste: 0.0 # image copy paste (probability)
|
||||
paste_in: 0.0 # image copy paste (probability), use 0 for faster training
|
||||
loss_ota: 1 # use ComputeLossOTA, use 0 for faster training
|
||||
@@ -0,0 +1 @@
|
||||
# init
|
||||
File diff suppressed because it is too large
Load Diff
@@ -0,0 +1,272 @@
|
||||
import numpy as np
|
||||
import random
|
||||
import torch
|
||||
import torch.nn as nn
|
||||
|
||||
from models.common import Conv, DWConv
|
||||
from utils.google_utils import attempt_download
|
||||
|
||||
|
||||
class CrossConv(nn.Module):
|
||||
# Cross Convolution Downsample
|
||||
def __init__(self, c1, c2, k=3, s=1, g=1, e=1.0, shortcut=False):
|
||||
# ch_in, ch_out, kernel, stride, groups, expansion, shortcut
|
||||
super(CrossConv, self).__init__()
|
||||
c_ = int(c2 * e) # hidden channels
|
||||
self.cv1 = Conv(c1, c_, (1, k), (1, s))
|
||||
self.cv2 = Conv(c_, c2, (k, 1), (s, 1), g=g)
|
||||
self.add = shortcut and c1 == c2
|
||||
|
||||
def forward(self, x):
|
||||
return x + self.cv2(self.cv1(x)) if self.add else self.cv2(self.cv1(x))
|
||||
|
||||
|
||||
class Sum(nn.Module):
|
||||
# Weighted sum of 2 or more layers https://arxiv.org/abs/1911.09070
|
||||
def __init__(self, n, weight=False): # n: number of inputs
|
||||
super(Sum, self).__init__()
|
||||
self.weight = weight # apply weights boolean
|
||||
self.iter = range(n - 1) # iter object
|
||||
if weight:
|
||||
self.w = nn.Parameter(-torch.arange(1., n) / 2, requires_grad=True) # layer weights
|
||||
|
||||
def forward(self, x):
|
||||
y = x[0] # no weight
|
||||
if self.weight:
|
||||
w = torch.sigmoid(self.w) * 2
|
||||
for i in self.iter:
|
||||
y = y + x[i + 1] * w[i]
|
||||
else:
|
||||
for i in self.iter:
|
||||
y = y + x[i + 1]
|
||||
return y
|
||||
|
||||
|
||||
class MixConv2d(nn.Module):
|
||||
# Mixed Depthwise Conv https://arxiv.org/abs/1907.09595
|
||||
def __init__(self, c1, c2, k=(1, 3), s=1, equal_ch=True):
|
||||
super(MixConv2d, self).__init__()
|
||||
groups = len(k)
|
||||
if equal_ch: # equal c_ per group
|
||||
i = torch.linspace(0, groups - 1E-6, c2).floor() # c2 indices
|
||||
c_ = [(i == g).sum() for g in range(groups)] # intermediate channels
|
||||
else: # equal weight.numel() per group
|
||||
b = [c2] + [0] * groups
|
||||
a = np.eye(groups + 1, groups, k=-1)
|
||||
a -= np.roll(a, 1, axis=1)
|
||||
a *= np.array(k) ** 2
|
||||
a[0] = 1
|
||||
c_ = np.linalg.lstsq(a, b, rcond=None)[0].round() # solve for equal weight indices, ax = b
|
||||
|
||||
self.m = nn.ModuleList([nn.Conv2d(c1, int(c_[g]), k[g], s, k[g] // 2, bias=False) for g in range(groups)])
|
||||
self.bn = nn.BatchNorm2d(c2)
|
||||
self.act = nn.LeakyReLU(0.1, inplace=True)
|
||||
|
||||
def forward(self, x):
|
||||
return x + self.act(self.bn(torch.cat([m(x) for m in self.m], 1)))
|
||||
|
||||
|
||||
class Ensemble(nn.ModuleList):
|
||||
# Ensemble of models
|
||||
def __init__(self):
|
||||
super(Ensemble, self).__init__()
|
||||
|
||||
def forward(self, x, augment=False):
|
||||
y = []
|
||||
for module in self:
|
||||
y.append(module(x, augment)[0])
|
||||
# y = torch.stack(y).max(0)[0] # max ensemble
|
||||
# y = torch.stack(y).mean(0) # mean ensemble
|
||||
y = torch.cat(y, 1) # nms ensemble
|
||||
return y, None # inference, train output
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
class ORT_NMS(torch.autograd.Function):
|
||||
'''ONNX-Runtime NMS operation'''
|
||||
@staticmethod
|
||||
def forward(ctx,
|
||||
boxes,
|
||||
scores,
|
||||
max_output_boxes_per_class=torch.tensor([100]),
|
||||
iou_threshold=torch.tensor([0.45]),
|
||||
score_threshold=torch.tensor([0.25])):
|
||||
device = boxes.device
|
||||
batch = scores.shape[0]
|
||||
num_det = random.randint(0, 100)
|
||||
batches = torch.randint(0, batch, (num_det,)).sort()[0].to(device)
|
||||
idxs = torch.arange(100, 100 + num_det).to(device)
|
||||
zeros = torch.zeros((num_det,), dtype=torch.int64).to(device)
|
||||
selected_indices = torch.cat([batches[None], zeros[None], idxs[None]], 0).T.contiguous()
|
||||
selected_indices = selected_indices.to(torch.int64)
|
||||
return selected_indices
|
||||
|
||||
@staticmethod
|
||||
def symbolic(g, boxes, scores, max_output_boxes_per_class, iou_threshold, score_threshold):
|
||||
return g.op("NonMaxSuppression", boxes, scores, max_output_boxes_per_class, iou_threshold, score_threshold)
|
||||
|
||||
|
||||
class TRT_NMS(torch.autograd.Function):
|
||||
'''TensorRT NMS operation'''
|
||||
@staticmethod
|
||||
def forward(
|
||||
ctx,
|
||||
boxes,
|
||||
scores,
|
||||
background_class=-1,
|
||||
box_coding=1,
|
||||
iou_threshold=0.45,
|
||||
max_output_boxes=100,
|
||||
plugin_version="1",
|
||||
score_activation=0,
|
||||
score_threshold=0.25,
|
||||
):
|
||||
batch_size, num_boxes, num_classes = scores.shape
|
||||
num_det = torch.randint(0, max_output_boxes, (batch_size, 1), dtype=torch.int32)
|
||||
det_boxes = torch.randn(batch_size, max_output_boxes, 4)
|
||||
det_scores = torch.randn(batch_size, max_output_boxes)
|
||||
det_classes = torch.randint(0, num_classes, (batch_size, max_output_boxes), dtype=torch.int32)
|
||||
return num_det, det_boxes, det_scores, det_classes
|
||||
|
||||
@staticmethod
|
||||
def symbolic(g,
|
||||
boxes,
|
||||
scores,
|
||||
background_class=-1,
|
||||
box_coding=1,
|
||||
iou_threshold=0.45,
|
||||
max_output_boxes=100,
|
||||
plugin_version="1",
|
||||
score_activation=0,
|
||||
score_threshold=0.25):
|
||||
out = g.op("TRT::EfficientNMS_TRT",
|
||||
boxes,
|
||||
scores,
|
||||
background_class_i=background_class,
|
||||
box_coding_i=box_coding,
|
||||
iou_threshold_f=iou_threshold,
|
||||
max_output_boxes_i=max_output_boxes,
|
||||
plugin_version_s=plugin_version,
|
||||
score_activation_i=score_activation,
|
||||
score_threshold_f=score_threshold,
|
||||
outputs=4)
|
||||
nums, boxes, scores, classes = out
|
||||
return nums, boxes, scores, classes
|
||||
|
||||
|
||||
class ONNX_ORT(nn.Module):
|
||||
'''onnx module with ONNX-Runtime NMS operation.'''
|
||||
def __init__(self, max_obj=100, iou_thres=0.45, score_thres=0.25, max_wh=640, device=None, n_classes=80):
|
||||
super().__init__()
|
||||
self.device = device if device else torch.device("cpu")
|
||||
self.max_obj = torch.tensor([max_obj]).to(device)
|
||||
self.iou_threshold = torch.tensor([iou_thres]).to(device)
|
||||
self.score_threshold = torch.tensor([score_thres]).to(device)
|
||||
self.max_wh = max_wh # if max_wh != 0 : non-agnostic else : agnostic
|
||||
self.convert_matrix = torch.tensor([[1, 0, 1, 0], [0, 1, 0, 1], [-0.5, 0, 0.5, 0], [0, -0.5, 0, 0.5]],
|
||||
dtype=torch.float32,
|
||||
device=self.device)
|
||||
self.n_classes=n_classes
|
||||
|
||||
def forward(self, x):
|
||||
boxes = x[:, :, :4]
|
||||
conf = x[:, :, 4:5]
|
||||
scores = x[:, :, 5:]
|
||||
if self.n_classes == 1:
|
||||
scores = conf # for models with one class, cls_loss is 0 and cls_conf is always 0.5,
|
||||
# so there is no need to multiplicate.
|
||||
else:
|
||||
scores *= conf # conf = obj_conf * cls_conf
|
||||
boxes @= self.convert_matrix
|
||||
max_score, category_id = scores.max(2, keepdim=True)
|
||||
dis = category_id.float() * self.max_wh
|
||||
nmsbox = boxes + dis
|
||||
max_score_tp = max_score.transpose(1, 2).contiguous()
|
||||
selected_indices = ORT_NMS.apply(nmsbox, max_score_tp, self.max_obj, self.iou_threshold, self.score_threshold)
|
||||
X, Y = selected_indices[:, 0], selected_indices[:, 2]
|
||||
selected_boxes = boxes[X, Y, :]
|
||||
selected_categories = category_id[X, Y, :].float()
|
||||
selected_scores = max_score[X, Y, :]
|
||||
X = X.unsqueeze(1).float()
|
||||
return torch.cat([X, selected_boxes, selected_categories, selected_scores], 1)
|
||||
|
||||
class ONNX_TRT(nn.Module):
|
||||
'''onnx module with TensorRT NMS operation.'''
|
||||
def __init__(self, max_obj=100, iou_thres=0.45, score_thres=0.25, max_wh=None ,device=None, n_classes=80):
|
||||
super().__init__()
|
||||
assert max_wh is None
|
||||
self.device = device if device else torch.device('cpu')
|
||||
self.background_class = -1,
|
||||
self.box_coding = 1,
|
||||
self.iou_threshold = iou_thres
|
||||
self.max_obj = max_obj
|
||||
self.plugin_version = '1'
|
||||
self.score_activation = 0
|
||||
self.score_threshold = score_thres
|
||||
self.n_classes=n_classes
|
||||
|
||||
def forward(self, x):
|
||||
boxes = x[:, :, :4]
|
||||
conf = x[:, :, 4:5]
|
||||
scores = x[:, :, 5:]
|
||||
if self.n_classes == 1:
|
||||
scores = conf # for models with one class, cls_loss is 0 and cls_conf is always 0.5,
|
||||
# so there is no need to multiplicate.
|
||||
else:
|
||||
scores *= conf # conf = obj_conf * cls_conf
|
||||
num_det, det_boxes, det_scores, det_classes = TRT_NMS.apply(boxes, scores, self.background_class, self.box_coding,
|
||||
self.iou_threshold, self.max_obj,
|
||||
self.plugin_version, self.score_activation,
|
||||
self.score_threshold)
|
||||
return num_det, det_boxes, det_scores, det_classes
|
||||
|
||||
|
||||
class End2End(nn.Module):
|
||||
'''export onnx or tensorrt model with NMS operation.'''
|
||||
def __init__(self, model, max_obj=100, iou_thres=0.45, score_thres=0.25, max_wh=None, device=None, n_classes=80):
|
||||
super().__init__()
|
||||
device = device if device else torch.device('cpu')
|
||||
assert isinstance(max_wh,(int)) or max_wh is None
|
||||
self.model = model.to(device)
|
||||
self.model.model[-1].end2end = True
|
||||
self.patch_model = ONNX_TRT if max_wh is None else ONNX_ORT
|
||||
self.end2end = self.patch_model(max_obj, iou_thres, score_thres, max_wh, device, n_classes)
|
||||
self.end2end.eval()
|
||||
|
||||
def forward(self, x):
|
||||
x = self.model(x)
|
||||
x = self.end2end(x)
|
||||
return x
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
def attempt_load(weights, map_location=None):
|
||||
# Loads an ensemble of models weights=[a,b,c] or a single model weights=[a] or weights=a
|
||||
model = Ensemble()
|
||||
for w in weights if isinstance(weights, list) else [weights]:
|
||||
attempt_download(w)
|
||||
ckpt = torch.load(w, map_location=map_location) # load
|
||||
model.append(ckpt['ema' if ckpt.get('ema') else 'model'].float().fuse().eval()) # FP32 model
|
||||
|
||||
# Compatibility updates
|
||||
for m in model.modules():
|
||||
if type(m) in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU]:
|
||||
m.inplace = True # pytorch 1.7.0 compatibility
|
||||
elif type(m) is nn.Upsample:
|
||||
m.recompute_scale_factor = None # torch 1.11.0 compatibility
|
||||
elif type(m) is Conv:
|
||||
m._non_persistent_buffers_set = set() # pytorch 1.6.0 compatibility
|
||||
|
||||
if len(model) == 1:
|
||||
return model[-1] # return model
|
||||
else:
|
||||
print('Ensemble created with %s\n' % weights)
|
||||
for k in ['names', 'stride']:
|
||||
setattr(model, k, getattr(model[-1], k))
|
||||
return model # return ensemble
|
||||
|
||||
|
||||
@@ -0,0 +1,843 @@
|
||||
import argparse
|
||||
import logging
|
||||
import sys
|
||||
from copy import deepcopy
|
||||
|
||||
sys.path.append('./') # to run '$ python *.py' files in subdirectories
|
||||
logger = logging.getLogger(__name__)
|
||||
import torch
|
||||
from models.common import *
|
||||
from models.experimental import *
|
||||
from utils.autoanchor import check_anchor_order
|
||||
from utils.general import make_divisible, check_file, set_logging
|
||||
from utils.torch_utils import time_synchronized, fuse_conv_and_bn, model_info, scale_img, initialize_weights, \
|
||||
select_device, copy_attr
|
||||
from utils.loss import SigmoidBin
|
||||
|
||||
try:
|
||||
import thop # for FLOPS computation
|
||||
except ImportError:
|
||||
thop = None
|
||||
|
||||
|
||||
class Detect(nn.Module):
|
||||
stride = None # strides computed during build
|
||||
export = False # onnx export
|
||||
end2end = False
|
||||
include_nms = False
|
||||
concat = False
|
||||
|
||||
def __init__(self, nc=80, anchors=(), ch=()): # detection layer
|
||||
super(Detect, self).__init__()
|
||||
self.nc = nc # number of classes
|
||||
self.no = nc + 5 # number of outputs per anchor
|
||||
self.nl = len(anchors) # number of detection layers
|
||||
self.na = len(anchors[0]) // 2 # number of anchors
|
||||
self.grid = [torch.zeros(1)] * self.nl # init grid
|
||||
a = torch.tensor(anchors).float().view(self.nl, -1, 2)
|
||||
self.register_buffer('anchors', a) # shape(nl,na,2)
|
||||
self.register_buffer('anchor_grid', a.clone().view(self.nl, 1, -1, 1, 1, 2)) # shape(nl,1,na,1,1,2)
|
||||
self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv
|
||||
|
||||
def forward(self, x):
|
||||
# x = x.copy() # for profiling
|
||||
z = [] # inference output
|
||||
self.training |= self.export
|
||||
for i in range(self.nl):
|
||||
x[i] = self.m[i](x[i]) # conv
|
||||
bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85)
|
||||
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
|
||||
if not self.training: # inference
|
||||
if self.grid[i].shape[2:4] != x[i].shape[2:4]:
|
||||
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
|
||||
y = x[i].sigmoid()
|
||||
if not torch.onnx.is_in_onnx_export():
|
||||
y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
|
||||
else:
|
||||
xy, wh, conf = y.split((2, 2, self.nc + 1), 4) # y.tensor_split((2, 4, 5), 4) # torch 1.8.0
|
||||
xy = xy * (2. * self.stride[i]) + (self.stride[i] * (self.grid[i] - 0.5)) # new xy
|
||||
wh = wh ** 2 * (4 * self.anchor_grid[i].data) # new wh
|
||||
y = torch.cat((xy, wh, conf), 4)
|
||||
z.append(y.view(bs, -1, self.no))
|
||||
|
||||
if self.training:
|
||||
out = x
|
||||
elif self.end2end:
|
||||
out = torch.cat(z, 1)
|
||||
elif self.include_nms:
|
||||
z = self.convert(z)
|
||||
out = (z, )
|
||||
elif self.concat:
|
||||
out = torch.cat(z, 1)
|
||||
else:
|
||||
out = (torch.cat(z, 1), x)
|
||||
|
||||
return out
|
||||
|
||||
@staticmethod
|
||||
def _make_grid(nx=20, ny=20):
|
||||
yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
|
||||
return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
|
||||
|
||||
def convert(self, z):
|
||||
z = torch.cat(z, 1)
|
||||
box = z[:, :, :4]
|
||||
conf = z[:, :, 4:5]
|
||||
score = z[:, :, 5:]
|
||||
score *= conf
|
||||
convert_matrix = torch.tensor([[1, 0, 1, 0], [0, 1, 0, 1], [-0.5, 0, 0.5, 0], [0, -0.5, 0, 0.5]],
|
||||
dtype=torch.float32,
|
||||
device=z.device)
|
||||
box @= convert_matrix
|
||||
return (box, score)
|
||||
|
||||
|
||||
class IDetect(nn.Module):
|
||||
stride = None # strides computed during build
|
||||
export = False # onnx export
|
||||
end2end = False
|
||||
include_nms = False
|
||||
concat = False
|
||||
|
||||
def __init__(self, nc=80, anchors=(), ch=()): # detection layer
|
||||
super(IDetect, self).__init__()
|
||||
self.nc = nc # number of classes
|
||||
self.no = nc + 5 # number of outputs per anchor
|
||||
self.nl = len(anchors) # number of detection layers
|
||||
self.na = len(anchors[0]) // 2 # number of anchors
|
||||
self.grid = [torch.zeros(1)] * self.nl # init grid
|
||||
a = torch.tensor(anchors).float().view(self.nl, -1, 2)
|
||||
self.register_buffer('anchors', a) # shape(nl,na,2)
|
||||
self.register_buffer('anchor_grid', a.clone().view(self.nl, 1, -1, 1, 1, 2)) # shape(nl,1,na,1,1,2)
|
||||
self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv
|
||||
|
||||
self.ia = nn.ModuleList(ImplicitA(x) for x in ch)
|
||||
self.im = nn.ModuleList(ImplicitM(self.no * self.na) for _ in ch)
|
||||
|
||||
def forward(self, x):
|
||||
# x = x.copy() # for profiling
|
||||
z = [] # inference output
|
||||
self.training |= self.export
|
||||
for i in range(self.nl):
|
||||
x[i] = self.m[i](self.ia[i](x[i])) # conv
|
||||
x[i] = self.im[i](x[i])
|
||||
bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85)
|
||||
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
|
||||
if not self.training: # inference
|
||||
if self.grid[i].shape[2:4] != x[i].shape[2:4]:
|
||||
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
|
||||
|
||||
y = x[i].sigmoid()
|
||||
y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
|
||||
z.append(y.view(bs, -1, self.no))
|
||||
|
||||
return x if self.training else (torch.cat(z, 1), x)
|
||||
|
||||
def fuseforward(self, x):
|
||||
# x = x.copy() # for profiling
|
||||
z = [] # inference output
|
||||
self.training |= self.export
|
||||
for i in range(self.nl):
|
||||
x[i] = self.m[i](x[i]) # conv
|
||||
bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85)
|
||||
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
|
||||
if not self.training: # inference
|
||||
if self.grid[i].shape[2:4] != x[i].shape[2:4]:
|
||||
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
|
||||
|
||||
y = x[i].sigmoid()
|
||||
if not torch.onnx.is_in_onnx_export():
|
||||
y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
|
||||
else:
|
||||
xy, wh, conf = y.split((2, 2, self.nc + 1), 4) # y.tensor_split((2, 4, 5), 4) # torch 1.8.0
|
||||
xy = xy * (2. * self.stride[i]) + (self.stride[i] * (self.grid[i] - 0.5)) # new xy
|
||||
wh = wh ** 2 * (4 * self.anchor_grid[i].data) # new wh
|
||||
y = torch.cat((xy, wh, conf), 4)
|
||||
z.append(y.view(bs, -1, self.no))
|
||||
|
||||
if self.training:
|
||||
out = x
|
||||
elif self.end2end:
|
||||
out = torch.cat(z, 1)
|
||||
elif self.include_nms:
|
||||
z = self.convert(z)
|
||||
out = (z, )
|
||||
elif self.concat:
|
||||
out = torch.cat(z, 1)
|
||||
else:
|
||||
out = (torch.cat(z, 1), x)
|
||||
|
||||
return out
|
||||
|
||||
def fuse(self):
|
||||
print("IDetect.fuse")
|
||||
# fuse ImplicitA and Convolution
|
||||
for i in range(len(self.m)):
|
||||
c1,c2,_,_ = self.m[i].weight.shape
|
||||
c1_,c2_, _,_ = self.ia[i].implicit.shape
|
||||
self.m[i].bias += torch.matmul(self.m[i].weight.reshape(c1,c2),self.ia[i].implicit.reshape(c2_,c1_)).squeeze(1)
|
||||
|
||||
# fuse ImplicitM and Convolution
|
||||
for i in range(len(self.m)):
|
||||
c1,c2, _,_ = self.im[i].implicit.shape
|
||||
self.m[i].bias *= self.im[i].implicit.reshape(c2)
|
||||
self.m[i].weight *= self.im[i].implicit.transpose(0,1)
|
||||
|
||||
@staticmethod
|
||||
def _make_grid(nx=20, ny=20):
|
||||
yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
|
||||
return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
|
||||
|
||||
def convert(self, z):
|
||||
z = torch.cat(z, 1)
|
||||
box = z[:, :, :4]
|
||||
conf = z[:, :, 4:5]
|
||||
score = z[:, :, 5:]
|
||||
score *= conf
|
||||
convert_matrix = torch.tensor([[1, 0, 1, 0], [0, 1, 0, 1], [-0.5, 0, 0.5, 0], [0, -0.5, 0, 0.5]],
|
||||
dtype=torch.float32,
|
||||
device=z.device)
|
||||
box @= convert_matrix
|
||||
return (box, score)
|
||||
|
||||
|
||||
class IKeypoint(nn.Module):
|
||||
stride = None # strides computed during build
|
||||
export = False # onnx export
|
||||
|
||||
def __init__(self, nc=80, anchors=(), nkpt=17, ch=(), inplace=True, dw_conv_kpt=False): # detection layer
|
||||
super(IKeypoint, self).__init__()
|
||||
self.nc = nc # number of classes
|
||||
self.nkpt = nkpt
|
||||
self.dw_conv_kpt = dw_conv_kpt
|
||||
self.no_det=(nc + 5) # number of outputs per anchor for box and class
|
||||
self.no_kpt = 3*self.nkpt ## number of outputs per anchor for keypoints
|
||||
self.no = self.no_det+self.no_kpt
|
||||
self.nl = len(anchors) # number of detection layers
|
||||
self.na = len(anchors[0]) // 2 # number of anchors
|
||||
self.grid = [torch.zeros(1)] * self.nl # init grid
|
||||
self.flip_test = False
|
||||
a = torch.tensor(anchors).float().view(self.nl, -1, 2)
|
||||
self.register_buffer('anchors', a) # shape(nl,na,2)
|
||||
self.register_buffer('anchor_grid', a.clone().view(self.nl, 1, -1, 1, 1, 2)) # shape(nl,1,na,1,1,2)
|
||||
self.m = nn.ModuleList(nn.Conv2d(x, self.no_det * self.na, 1) for x in ch) # output conv
|
||||
|
||||
self.ia = nn.ModuleList(ImplicitA(x) for x in ch)
|
||||
self.im = nn.ModuleList(ImplicitM(self.no_det * self.na) for _ in ch)
|
||||
|
||||
if self.nkpt is not None:
|
||||
if self.dw_conv_kpt: #keypoint head is slightly more complex
|
||||
self.m_kpt = nn.ModuleList(
|
||||
nn.Sequential(DWConv(x, x, k=3), Conv(x,x),
|
||||
DWConv(x, x, k=3), Conv(x, x),
|
||||
DWConv(x, x, k=3), Conv(x,x),
|
||||
DWConv(x, x, k=3), Conv(x, x),
|
||||
DWConv(x, x, k=3), Conv(x, x),
|
||||
DWConv(x, x, k=3), nn.Conv2d(x, self.no_kpt * self.na, 1)) for x in ch)
|
||||
else: #keypoint head is a single convolution
|
||||
self.m_kpt = nn.ModuleList(nn.Conv2d(x, self.no_kpt * self.na, 1) for x in ch)
|
||||
|
||||
self.inplace = inplace # use in-place ops (e.g. slice assignment)
|
||||
|
||||
def forward(self, x):
|
||||
# x = x.copy() # for profiling
|
||||
z = [] # inference output
|
||||
self.training |= self.export
|
||||
for i in range(self.nl):
|
||||
if self.nkpt is None or self.nkpt==0:
|
||||
x[i] = self.im[i](self.m[i](self.ia[i](x[i]))) # conv
|
||||
else :
|
||||
x[i] = torch.cat((self.im[i](self.m[i](self.ia[i](x[i]))), self.m_kpt[i](x[i])), axis=1)
|
||||
|
||||
bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85)
|
||||
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
x_det = x[i][..., :6]
|
||||
x_kpt = x[i][..., 6:]
|
||||
|
||||
if not self.training: # inference
|
||||
if self.grid[i].shape[2:4] != x[i].shape[2:4]:
|
||||
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
|
||||
kpt_grid_x = self.grid[i][..., 0:1]
|
||||
kpt_grid_y = self.grid[i][..., 1:2]
|
||||
|
||||
if self.nkpt == 0:
|
||||
y = x[i].sigmoid()
|
||||
else:
|
||||
y = x_det.sigmoid()
|
||||
|
||||
if self.inplace:
|
||||
xy = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i].view(1, self.na, 1, 1, 2) # wh
|
||||
if self.nkpt != 0:
|
||||
x_kpt[..., 0::3] = (x_kpt[..., ::3] * 2. - 0.5 + kpt_grid_x.repeat(1,1,1,1,17)) * self.stride[i] # xy
|
||||
x_kpt[..., 1::3] = (x_kpt[..., 1::3] * 2. - 0.5 + kpt_grid_y.repeat(1,1,1,1,17)) * self.stride[i] # xy
|
||||
#x_kpt[..., 0::3] = (x_kpt[..., ::3] + kpt_grid_x.repeat(1,1,1,1,17)) * self.stride[i] # xy
|
||||
#x_kpt[..., 1::3] = (x_kpt[..., 1::3] + kpt_grid_y.repeat(1,1,1,1,17)) * self.stride[i] # xy
|
||||
#print('=============')
|
||||
#print(self.anchor_grid[i].shape)
|
||||
#print(self.anchor_grid[i][...,0].unsqueeze(4).shape)
|
||||
#print(x_kpt[..., 0::3].shape)
|
||||
#x_kpt[..., 0::3] = ((x_kpt[..., 0::3].tanh() * 2.) ** 3 * self.anchor_grid[i][...,0].unsqueeze(4).repeat(1,1,1,1,self.nkpt)) + kpt_grid_x.repeat(1,1,1,1,17) * self.stride[i] # xy
|
||||
#x_kpt[..., 1::3] = ((x_kpt[..., 1::3].tanh() * 2.) ** 3 * self.anchor_grid[i][...,1].unsqueeze(4).repeat(1,1,1,1,self.nkpt)) + kpt_grid_y.repeat(1,1,1,1,17) * self.stride[i] # xy
|
||||
#x_kpt[..., 0::3] = (((x_kpt[..., 0::3].sigmoid() * 4.) ** 2 - 8.) * self.anchor_grid[i][...,0].unsqueeze(4).repeat(1,1,1,1,self.nkpt)) + kpt_grid_x.repeat(1,1,1,1,17) * self.stride[i] # xy
|
||||
#x_kpt[..., 1::3] = (((x_kpt[..., 1::3].sigmoid() * 4.) ** 2 - 8.) * self.anchor_grid[i][...,1].unsqueeze(4).repeat(1,1,1,1,self.nkpt)) + kpt_grid_y.repeat(1,1,1,1,17) * self.stride[i] # xy
|
||||
x_kpt[..., 2::3] = x_kpt[..., 2::3].sigmoid()
|
||||
|
||||
y = torch.cat((xy, wh, y[..., 4:], x_kpt), dim = -1)
|
||||
|
||||
else: # for YOLOv5 on AWS Inferentia https://github.com/ultralytics/yolov5/pull/2953
|
||||
xy = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
|
||||
if self.nkpt != 0:
|
||||
y[..., 6:] = (y[..., 6:] * 2. - 0.5 + self.grid[i].repeat((1,1,1,1,self.nkpt))) * self.stride[i] # xy
|
||||
y = torch.cat((xy, wh, y[..., 4:]), -1)
|
||||
|
||||
z.append(y.view(bs, -1, self.no))
|
||||
|
||||
return x if self.training else (torch.cat(z, 1), x)
|
||||
|
||||
@staticmethod
|
||||
def _make_grid(nx=20, ny=20):
|
||||
yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
|
||||
return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
|
||||
|
||||
|
||||
class IAuxDetect(nn.Module):
|
||||
stride = None # strides computed during build
|
||||
export = False # onnx export
|
||||
end2end = False
|
||||
include_nms = False
|
||||
concat = False
|
||||
|
||||
def __init__(self, nc=80, anchors=(), ch=()): # detection layer
|
||||
super(IAuxDetect, self).__init__()
|
||||
self.nc = nc # number of classes
|
||||
self.no = nc + 5 # number of outputs per anchor
|
||||
self.nl = len(anchors) # number of detection layers
|
||||
self.na = len(anchors[0]) // 2 # number of anchors
|
||||
self.grid = [torch.zeros(1)] * self.nl # init grid
|
||||
a = torch.tensor(anchors).float().view(self.nl, -1, 2)
|
||||
self.register_buffer('anchors', a) # shape(nl,na,2)
|
||||
self.register_buffer('anchor_grid', a.clone().view(self.nl, 1, -1, 1, 1, 2)) # shape(nl,1,na,1,1,2)
|
||||
self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch[:self.nl]) # output conv
|
||||
self.m2 = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch[self.nl:]) # output conv
|
||||
|
||||
self.ia = nn.ModuleList(ImplicitA(x) for x in ch[:self.nl])
|
||||
self.im = nn.ModuleList(ImplicitM(self.no * self.na) for _ in ch[:self.nl])
|
||||
|
||||
def forward(self, x):
|
||||
# x = x.copy() # for profiling
|
||||
z = [] # inference output
|
||||
self.training |= self.export
|
||||
for i in range(self.nl):
|
||||
x[i] = self.m[i](self.ia[i](x[i])) # conv
|
||||
x[i] = self.im[i](x[i])
|
||||
bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85)
|
||||
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
|
||||
x[i+self.nl] = self.m2[i](x[i+self.nl])
|
||||
x[i+self.nl] = x[i+self.nl].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
|
||||
if not self.training: # inference
|
||||
if self.grid[i].shape[2:4] != x[i].shape[2:4]:
|
||||
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
|
||||
|
||||
y = x[i].sigmoid()
|
||||
if not torch.onnx.is_in_onnx_export():
|
||||
y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
|
||||
else:
|
||||
xy, wh, conf = y.split((2, 2, self.nc + 1), 4) # y.tensor_split((2, 4, 5), 4) # torch 1.8.0
|
||||
xy = xy * (2. * self.stride[i]) + (self.stride[i] * (self.grid[i] - 0.5)) # new xy
|
||||
wh = wh ** 2 * (4 * self.anchor_grid[i].data) # new wh
|
||||
y = torch.cat((xy, wh, conf), 4)
|
||||
z.append(y.view(bs, -1, self.no))
|
||||
|
||||
return x if self.training else (torch.cat(z, 1), x[:self.nl])
|
||||
|
||||
def fuseforward(self, x):
|
||||
# x = x.copy() # for profiling
|
||||
z = [] # inference output
|
||||
self.training |= self.export
|
||||
for i in range(self.nl):
|
||||
x[i] = self.m[i](x[i]) # conv
|
||||
bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85)
|
||||
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
|
||||
if not self.training: # inference
|
||||
if self.grid[i].shape[2:4] != x[i].shape[2:4]:
|
||||
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
|
||||
|
||||
y = x[i].sigmoid()
|
||||
if not torch.onnx.is_in_onnx_export():
|
||||
y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
|
||||
else:
|
||||
xy = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
wh = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i].data # wh
|
||||
y = torch.cat((xy, wh, y[..., 4:]), -1)
|
||||
z.append(y.view(bs, -1, self.no))
|
||||
|
||||
if self.training:
|
||||
out = x
|
||||
elif self.end2end:
|
||||
out = torch.cat(z, 1)
|
||||
elif self.include_nms:
|
||||
z = self.convert(z)
|
||||
out = (z, )
|
||||
elif self.concat:
|
||||
out = torch.cat(z, 1)
|
||||
else:
|
||||
out = (torch.cat(z, 1), x)
|
||||
|
||||
return out
|
||||
|
||||
def fuse(self):
|
||||
print("IAuxDetect.fuse")
|
||||
# fuse ImplicitA and Convolution
|
||||
for i in range(len(self.m)):
|
||||
c1,c2,_,_ = self.m[i].weight.shape
|
||||
c1_,c2_, _,_ = self.ia[i].implicit.shape
|
||||
self.m[i].bias += torch.matmul(self.m[i].weight.reshape(c1,c2),self.ia[i].implicit.reshape(c2_,c1_)).squeeze(1)
|
||||
|
||||
# fuse ImplicitM and Convolution
|
||||
for i in range(len(self.m)):
|
||||
c1,c2, _,_ = self.im[i].implicit.shape
|
||||
self.m[i].bias *= self.im[i].implicit.reshape(c2)
|
||||
self.m[i].weight *= self.im[i].implicit.transpose(0,1)
|
||||
|
||||
@staticmethod
|
||||
def _make_grid(nx=20, ny=20):
|
||||
yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
|
||||
return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
|
||||
|
||||
def convert(self, z):
|
||||
z = torch.cat(z, 1)
|
||||
box = z[:, :, :4]
|
||||
conf = z[:, :, 4:5]
|
||||
score = z[:, :, 5:]
|
||||
score *= conf
|
||||
convert_matrix = torch.tensor([[1, 0, 1, 0], [0, 1, 0, 1], [-0.5, 0, 0.5, 0], [0, -0.5, 0, 0.5]],
|
||||
dtype=torch.float32,
|
||||
device=z.device)
|
||||
box @= convert_matrix
|
||||
return (box, score)
|
||||
|
||||
|
||||
class IBin(nn.Module):
|
||||
stride = None # strides computed during build
|
||||
export = False # onnx export
|
||||
|
||||
def __init__(self, nc=80, anchors=(), ch=(), bin_count=21): # detection layer
|
||||
super(IBin, self).__init__()
|
||||
self.nc = nc # number of classes
|
||||
self.bin_count = bin_count
|
||||
|
||||
self.w_bin_sigmoid = SigmoidBin(bin_count=self.bin_count, min=0.0, max=4.0)
|
||||
self.h_bin_sigmoid = SigmoidBin(bin_count=self.bin_count, min=0.0, max=4.0)
|
||||
# classes, x,y,obj
|
||||
self.no = nc + 3 + \
|
||||
self.w_bin_sigmoid.get_length() + self.h_bin_sigmoid.get_length() # w-bce, h-bce
|
||||
# + self.x_bin_sigmoid.get_length() + self.y_bin_sigmoid.get_length()
|
||||
|
||||
self.nl = len(anchors) # number of detection layers
|
||||
self.na = len(anchors[0]) // 2 # number of anchors
|
||||
self.grid = [torch.zeros(1)] * self.nl # init grid
|
||||
a = torch.tensor(anchors).float().view(self.nl, -1, 2)
|
||||
self.register_buffer('anchors', a) # shape(nl,na,2)
|
||||
self.register_buffer('anchor_grid', a.clone().view(self.nl, 1, -1, 1, 1, 2)) # shape(nl,1,na,1,1,2)
|
||||
self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na, 1) for x in ch) # output conv
|
||||
|
||||
self.ia = nn.ModuleList(ImplicitA(x) for x in ch)
|
||||
self.im = nn.ModuleList(ImplicitM(self.no * self.na) for _ in ch)
|
||||
|
||||
def forward(self, x):
|
||||
|
||||
#self.x_bin_sigmoid.use_fw_regression = True
|
||||
#self.y_bin_sigmoid.use_fw_regression = True
|
||||
self.w_bin_sigmoid.use_fw_regression = True
|
||||
self.h_bin_sigmoid.use_fw_regression = True
|
||||
|
||||
# x = x.copy() # for profiling
|
||||
z = [] # inference output
|
||||
self.training |= self.export
|
||||
for i in range(self.nl):
|
||||
x[i] = self.m[i](self.ia[i](x[i])) # conv
|
||||
x[i] = self.im[i](x[i])
|
||||
bs, _, ny, nx = x[i].shape # x(bs,255,20,20) to x(bs,3,20,20,85)
|
||||
x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
|
||||
|
||||
if not self.training: # inference
|
||||
if self.grid[i].shape[2:4] != x[i].shape[2:4]:
|
||||
self.grid[i] = self._make_grid(nx, ny).to(x[i].device)
|
||||
|
||||
y = x[i].sigmoid()
|
||||
y[..., 0:2] = (y[..., 0:2] * 2. - 0.5 + self.grid[i]) * self.stride[i] # xy
|
||||
#y[..., 2:4] = (y[..., 2:4] * 2) ** 2 * self.anchor_grid[i] # wh
|
||||
|
||||
|
||||
#px = (self.x_bin_sigmoid.forward(y[..., 0:12]) + self.grid[i][..., 0]) * self.stride[i]
|
||||
#py = (self.y_bin_sigmoid.forward(y[..., 12:24]) + self.grid[i][..., 1]) * self.stride[i]
|
||||
|
||||
pw = self.w_bin_sigmoid.forward(y[..., 2:24]) * self.anchor_grid[i][..., 0]
|
||||
ph = self.h_bin_sigmoid.forward(y[..., 24:46]) * self.anchor_grid[i][..., 1]
|
||||
|
||||
#y[..., 0] = px
|
||||
#y[..., 1] = py
|
||||
y[..., 2] = pw
|
||||
y[..., 3] = ph
|
||||
|
||||
y = torch.cat((y[..., 0:4], y[..., 46:]), dim=-1)
|
||||
|
||||
z.append(y.view(bs, -1, y.shape[-1]))
|
||||
|
||||
return x if self.training else (torch.cat(z, 1), x)
|
||||
|
||||
@staticmethod
|
||||
def _make_grid(nx=20, ny=20):
|
||||
yv, xv = torch.meshgrid([torch.arange(ny), torch.arange(nx)])
|
||||
return torch.stack((xv, yv), 2).view((1, 1, ny, nx, 2)).float()
|
||||
|
||||
|
||||
class Model(nn.Module):
|
||||
def __init__(self, cfg='yolor-csp-c.yaml', ch=3, nc=None, anchors=None): # model, input channels, number of classes
|
||||
super(Model, self).__init__()
|
||||
self.traced = False
|
||||
if isinstance(cfg, dict):
|
||||
self.yaml = cfg # model dict
|
||||
else: # is *.yaml
|
||||
import yaml # for torch hub
|
||||
self.yaml_file = Path(cfg).name
|
||||
with open(cfg) as f:
|
||||
self.yaml = yaml.load(f, Loader=yaml.SafeLoader) # model dict
|
||||
|
||||
# Define model
|
||||
ch = self.yaml['ch'] = self.yaml.get('ch', ch) # input channels
|
||||
if nc and nc != self.yaml['nc']:
|
||||
logger.info(f"Overriding model.yaml nc={self.yaml['nc']} with nc={nc}")
|
||||
self.yaml['nc'] = nc # override yaml value
|
||||
if anchors:
|
||||
logger.info(f'Overriding model.yaml anchors with anchors={anchors}')
|
||||
self.yaml['anchors'] = round(anchors) # override yaml value
|
||||
self.model, self.save = parse_model(deepcopy(self.yaml), ch=[ch]) # model, savelist
|
||||
self.names = [str(i) for i in range(self.yaml['nc'])] # default names
|
||||
# print([x.shape for x in self.forward(torch.zeros(1, ch, 64, 64))])
|
||||
|
||||
# Build strides, anchors
|
||||
m = self.model[-1] # Detect()
|
||||
if isinstance(m, Detect):
|
||||
s = 256 # 2x min stride
|
||||
m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))]) # forward
|
||||
check_anchor_order(m)
|
||||
m.anchors /= m.stride.view(-1, 1, 1)
|
||||
self.stride = m.stride
|
||||
self._initialize_biases() # only run once
|
||||
# print('Strides: %s' % m.stride.tolist())
|
||||
if isinstance(m, IDetect):
|
||||
s = 256 # 2x min stride
|
||||
m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))]) # forward
|
||||
check_anchor_order(m)
|
||||
m.anchors /= m.stride.view(-1, 1, 1)
|
||||
self.stride = m.stride
|
||||
self._initialize_biases() # only run once
|
||||
# print('Strides: %s' % m.stride.tolist())
|
||||
if isinstance(m, IAuxDetect):
|
||||
s = 256 # 2x min stride
|
||||
m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))[:4]]) # forward
|
||||
#print(m.stride)
|
||||
check_anchor_order(m)
|
||||
m.anchors /= m.stride.view(-1, 1, 1)
|
||||
self.stride = m.stride
|
||||
self._initialize_aux_biases() # only run once
|
||||
# print('Strides: %s' % m.stride.tolist())
|
||||
if isinstance(m, IBin):
|
||||
s = 256 # 2x min stride
|
||||
m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))]) # forward
|
||||
check_anchor_order(m)
|
||||
m.anchors /= m.stride.view(-1, 1, 1)
|
||||
self.stride = m.stride
|
||||
self._initialize_biases_bin() # only run once
|
||||
# print('Strides: %s' % m.stride.tolist())
|
||||
if isinstance(m, IKeypoint):
|
||||
s = 256 # 2x min stride
|
||||
m.stride = torch.tensor([s / x.shape[-2] for x in self.forward(torch.zeros(1, ch, s, s))]) # forward
|
||||
check_anchor_order(m)
|
||||
m.anchors /= m.stride.view(-1, 1, 1)
|
||||
self.stride = m.stride
|
||||
self._initialize_biases_kpt() # only run once
|
||||
# print('Strides: %s' % m.stride.tolist())
|
||||
|
||||
# Init weights, biases
|
||||
initialize_weights(self)
|
||||
self.info()
|
||||
logger.info('')
|
||||
|
||||
def forward(self, x, augment=False, profile=False):
|
||||
if augment:
|
||||
img_size = x.shape[-2:] # height, width
|
||||
s = [1, 0.83, 0.67] # scales
|
||||
f = [None, 3, None] # flips (2-ud, 3-lr)
|
||||
y = [] # outputs
|
||||
for si, fi in zip(s, f):
|
||||
xi = scale_img(x.flip(fi) if fi else x, si, gs=int(self.stride.max()))
|
||||
yi = self.forward_once(xi)[0] # forward
|
||||
# cv2.imwrite(f'img_{si}.jpg', 255 * xi[0].cpu().numpy().transpose((1, 2, 0))[:, :, ::-1]) # save
|
||||
yi[..., :4] /= si # de-scale
|
||||
if fi == 2:
|
||||
yi[..., 1] = img_size[0] - yi[..., 1] # de-flip ud
|
||||
elif fi == 3:
|
||||
yi[..., 0] = img_size[1] - yi[..., 0] # de-flip lr
|
||||
y.append(yi)
|
||||
return torch.cat(y, 1), None # augmented inference, train
|
||||
else:
|
||||
return self.forward_once(x, profile) # single-scale inference, train
|
||||
|
||||
def forward_once(self, x, profile=False):
|
||||
y, dt = [], [] # outputs
|
||||
for m in self.model:
|
||||
if m.f != -1: # if not from previous layer
|
||||
x = y[m.f] if isinstance(m.f, int) else [x if j == -1 else y[j] for j in m.f] # from earlier layers
|
||||
|
||||
if not hasattr(self, 'traced'):
|
||||
self.traced=False
|
||||
|
||||
if self.traced:
|
||||
if isinstance(m, Detect) or isinstance(m, IDetect) or isinstance(m, IAuxDetect) or isinstance(m, IKeypoint):
|
||||
break
|
||||
|
||||
if profile:
|
||||
c = isinstance(m, (Detect, IDetect, IAuxDetect, IBin))
|
||||
o = thop.profile(m, inputs=(x.copy() if c else x,), verbose=False)[0] / 1E9 * 2 if thop else 0 # FLOPS
|
||||
for _ in range(10):
|
||||
m(x.copy() if c else x)
|
||||
t = time_synchronized()
|
||||
for _ in range(10):
|
||||
m(x.copy() if c else x)
|
||||
dt.append((time_synchronized() - t) * 100)
|
||||
print('%10.1f%10.0f%10.1fms %-40s' % (o, m.np, dt[-1], m.type))
|
||||
|
||||
x = m(x) # run
|
||||
|
||||
y.append(x if m.i in self.save else None) # save output
|
||||
|
||||
if profile:
|
||||
print('%.1fms total' % sum(dt))
|
||||
return x
|
||||
|
||||
def _initialize_biases(self, cf=None): # initialize biases into Detect(), cf is class frequency
|
||||
# https://arxiv.org/abs/1708.02002 section 3.3
|
||||
# cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1.
|
||||
m = self.model[-1] # Detect() module
|
||||
for mi, s in zip(m.m, m.stride): # from
|
||||
b = mi.bias.view(m.na, -1) # conv.bias(255) to (3,85)
|
||||
b.data[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image)
|
||||
b.data[:, 5:] += math.log(0.6 / (m.nc - 0.99)) if cf is None else torch.log(cf / cf.sum()) # cls
|
||||
mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True)
|
||||
|
||||
def _initialize_aux_biases(self, cf=None): # initialize biases into Detect(), cf is class frequency
|
||||
# https://arxiv.org/abs/1708.02002 section 3.3
|
||||
# cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1.
|
||||
m = self.model[-1] # Detect() module
|
||||
for mi, mi2, s in zip(m.m, m.m2, m.stride): # from
|
||||
b = mi.bias.view(m.na, -1) # conv.bias(255) to (3,85)
|
||||
b.data[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image)
|
||||
b.data[:, 5:] += math.log(0.6 / (m.nc - 0.99)) if cf is None else torch.log(cf / cf.sum()) # cls
|
||||
mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True)
|
||||
b2 = mi2.bias.view(m.na, -1) # conv.bias(255) to (3,85)
|
||||
b2.data[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image)
|
||||
b2.data[:, 5:] += math.log(0.6 / (m.nc - 0.99)) if cf is None else torch.log(cf / cf.sum()) # cls
|
||||
mi2.bias = torch.nn.Parameter(b2.view(-1), requires_grad=True)
|
||||
|
||||
def _initialize_biases_bin(self, cf=None): # initialize biases into Detect(), cf is class frequency
|
||||
# https://arxiv.org/abs/1708.02002 section 3.3
|
||||
# cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1.
|
||||
m = self.model[-1] # Bin() module
|
||||
bc = m.bin_count
|
||||
for mi, s in zip(m.m, m.stride): # from
|
||||
b = mi.bias.view(m.na, -1) # conv.bias(255) to (3,85)
|
||||
old = b[:, (0,1,2,bc+3)].data
|
||||
obj_idx = 2*bc+4
|
||||
b[:, :obj_idx].data += math.log(0.6 / (bc + 1 - 0.99))
|
||||
b[:, obj_idx].data += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image)
|
||||
b[:, (obj_idx+1):].data += math.log(0.6 / (m.nc - 0.99)) if cf is None else torch.log(cf / cf.sum()) # cls
|
||||
b[:, (0,1,2,bc+3)].data = old
|
||||
mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True)
|
||||
|
||||
def _initialize_biases_kpt(self, cf=None): # initialize biases into Detect(), cf is class frequency
|
||||
# https://arxiv.org/abs/1708.02002 section 3.3
|
||||
# cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1.
|
||||
m = self.model[-1] # Detect() module
|
||||
for mi, s in zip(m.m, m.stride): # from
|
||||
b = mi.bias.view(m.na, -1) # conv.bias(255) to (3,85)
|
||||
b.data[:, 4] += math.log(8 / (640 / s) ** 2) # obj (8 objects per 640 image)
|
||||
b.data[:, 5:] += math.log(0.6 / (m.nc - 0.99)) if cf is None else torch.log(cf / cf.sum()) # cls
|
||||
mi.bias = torch.nn.Parameter(b.view(-1), requires_grad=True)
|
||||
|
||||
def _print_biases(self):
|
||||
m = self.model[-1] # Detect() module
|
||||
for mi in m.m: # from
|
||||
b = mi.bias.detach().view(m.na, -1).T # conv.bias(255) to (3,85)
|
||||
print(('%6g Conv2d.bias:' + '%10.3g' * 6) % (mi.weight.shape[1], *b[:5].mean(1).tolist(), b[5:].mean()))
|
||||
|
||||
# def _print_weights(self):
|
||||
# for m in self.model.modules():
|
||||
# if type(m) is Bottleneck:
|
||||
# print('%10.3g' % (m.w.detach().sigmoid() * 2)) # shortcut weights
|
||||
|
||||
def fuse(self): # fuse model Conv2d() + BatchNorm2d() layers
|
||||
print('Fusing layers... ')
|
||||
for m in self.model.modules():
|
||||
if isinstance(m, RepConv):
|
||||
#print(f" fuse_repvgg_block")
|
||||
m.fuse_repvgg_block()
|
||||
elif isinstance(m, RepConv_OREPA):
|
||||
#print(f" switch_to_deploy")
|
||||
m.switch_to_deploy()
|
||||
elif type(m) is Conv and hasattr(m, 'bn'):
|
||||
m.conv = fuse_conv_and_bn(m.conv, m.bn) # update conv
|
||||
delattr(m, 'bn') # remove batchnorm
|
||||
m.forward = m.fuseforward # update forward
|
||||
elif isinstance(m, (IDetect, IAuxDetect)):
|
||||
m.fuse()
|
||||
m.forward = m.fuseforward
|
||||
self.info()
|
||||
return self
|
||||
|
||||
def nms(self, mode=True): # add or remove NMS module
|
||||
present = type(self.model[-1]) is NMS # last layer is NMS
|
||||
if mode and not present:
|
||||
print('Adding NMS... ')
|
||||
m = NMS() # module
|
||||
m.f = -1 # from
|
||||
m.i = self.model[-1].i + 1 # index
|
||||
self.model.add_module(name='%s' % m.i, module=m) # add
|
||||
self.eval()
|
||||
elif not mode and present:
|
||||
print('Removing NMS... ')
|
||||
self.model = self.model[:-1] # remove
|
||||
return self
|
||||
|
||||
def autoshape(self): # add autoShape module
|
||||
print('Adding autoShape... ')
|
||||
m = autoShape(self) # wrap model
|
||||
copy_attr(m, self, include=('yaml', 'nc', 'hyp', 'names', 'stride'), exclude=()) # copy attributes
|
||||
return m
|
||||
|
||||
def info(self, verbose=False, img_size=640): # print model information
|
||||
model_info(self, verbose, img_size)
|
||||
|
||||
|
||||
def parse_model(d, ch): # model_dict, input_channels(3)
|
||||
logger.info('\n%3s%18s%3s%10s %-40s%-30s' % ('', 'from', 'n', 'params', 'module', 'arguments'))
|
||||
anchors, nc, gd, gw = d['anchors'], d['nc'], d['depth_multiple'], d['width_multiple']
|
||||
na = (len(anchors[0]) // 2) if isinstance(anchors, list) else anchors # number of anchors
|
||||
no = na * (nc + 5) # number of outputs = anchors * (classes + 5)
|
||||
|
||||
layers, save, c2 = [], [], ch[-1] # layers, savelist, ch out
|
||||
for i, (f, n, m, args) in enumerate(d['backbone'] + d['head']): # from, number, module, args
|
||||
m = eval(m) if isinstance(m, str) else m # eval strings
|
||||
for j, a in enumerate(args):
|
||||
try:
|
||||
args[j] = eval(a) if isinstance(a, str) else a # eval strings
|
||||
except:
|
||||
pass
|
||||
|
||||
n = max(round(n * gd), 1) if n > 1 else n # depth gain
|
||||
if m in [nn.Conv2d, Conv, RobustConv, RobustConv2, DWConv, GhostConv, RepConv, RepConv_OREPA, DownC,
|
||||
SPP, SPPF, SPPCSPC, GhostSPPCSPC, MixConv2d, Focus, Stem, GhostStem, CrossConv,
|
||||
Bottleneck, BottleneckCSPA, BottleneckCSPB, BottleneckCSPC,
|
||||
RepBottleneck, RepBottleneckCSPA, RepBottleneckCSPB, RepBottleneckCSPC,
|
||||
Res, ResCSPA, ResCSPB, ResCSPC,
|
||||
RepRes, RepResCSPA, RepResCSPB, RepResCSPC,
|
||||
ResX, ResXCSPA, ResXCSPB, ResXCSPC,
|
||||
RepResX, RepResXCSPA, RepResXCSPB, RepResXCSPC,
|
||||
Ghost, GhostCSPA, GhostCSPB, GhostCSPC,
|
||||
SwinTransformerBlock, STCSPA, STCSPB, STCSPC,
|
||||
SwinTransformer2Block, ST2CSPA, ST2CSPB, ST2CSPC]:
|
||||
c1, c2 = ch[f], args[0]
|
||||
if c2 != no: # if not output
|
||||
c2 = make_divisible(c2 * gw, 8)
|
||||
|
||||
args = [c1, c2, *args[1:]]
|
||||
if m in [DownC, SPPCSPC, GhostSPPCSPC,
|
||||
BottleneckCSPA, BottleneckCSPB, BottleneckCSPC,
|
||||
RepBottleneckCSPA, RepBottleneckCSPB, RepBottleneckCSPC,
|
||||
ResCSPA, ResCSPB, ResCSPC,
|
||||
RepResCSPA, RepResCSPB, RepResCSPC,
|
||||
ResXCSPA, ResXCSPB, ResXCSPC,
|
||||
RepResXCSPA, RepResXCSPB, RepResXCSPC,
|
||||
GhostCSPA, GhostCSPB, GhostCSPC,
|
||||
STCSPA, STCSPB, STCSPC,
|
||||
ST2CSPA, ST2CSPB, ST2CSPC]:
|
||||
args.insert(2, n) # number of repeats
|
||||
n = 1
|
||||
elif m is nn.BatchNorm2d:
|
||||
args = [ch[f]]
|
||||
elif m is Concat:
|
||||
c2 = sum([ch[x] for x in f])
|
||||
elif m is Chuncat:
|
||||
c2 = sum([ch[x] for x in f])
|
||||
elif m is Shortcut:
|
||||
c2 = ch[f[0]]
|
||||
elif m is Foldcut:
|
||||
c2 = ch[f] // 2
|
||||
elif m in [Detect, IDetect, IAuxDetect, IBin, IKeypoint]:
|
||||
args.append([ch[x] for x in f])
|
||||
if isinstance(args[1], int): # number of anchors
|
||||
args[1] = [list(range(args[1] * 2))] * len(f)
|
||||
elif m is ReOrg:
|
||||
c2 = ch[f] * 4
|
||||
elif m is Contract:
|
||||
c2 = ch[f] * args[0] ** 2
|
||||
elif m is Expand:
|
||||
c2 = ch[f] // args[0] ** 2
|
||||
else:
|
||||
c2 = ch[f]
|
||||
|
||||
m_ = nn.Sequential(*[m(*args) for _ in range(n)]) if n > 1 else m(*args) # module
|
||||
t = str(m)[8:-2].replace('__main__.', '') # module type
|
||||
np = sum([x.numel() for x in m_.parameters()]) # number params
|
||||
m_.i, m_.f, m_.type, m_.np = i, f, t, np # attach index, 'from' index, type, number params
|
||||
logger.info('%3s%18s%3s%10.0f %-40s%-30s' % (i, f, n, np, t, args)) # print
|
||||
save.extend(x % i for x in ([f] if isinstance(f, int) else f) if x != -1) # append to savelist
|
||||
layers.append(m_)
|
||||
if i == 0:
|
||||
ch = []
|
||||
ch.append(c2)
|
||||
return nn.Sequential(*layers), sorted(save)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--cfg', type=str, default='yolor-csp-c.yaml', help='model.yaml')
|
||||
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
|
||||
parser.add_argument('--profile', action='store_true', help='profile model speed')
|
||||
opt = parser.parse_args()
|
||||
opt.cfg = check_file(opt.cfg) # check file
|
||||
set_logging()
|
||||
device = select_device(opt.device)
|
||||
|
||||
# Create model
|
||||
model = Model(opt.cfg).to(device)
|
||||
model.train()
|
||||
|
||||
if opt.profile:
|
||||
img = torch.rand(1, 3, 640, 640).to(device)
|
||||
y = model(img, profile=True)
|
||||
|
||||
# Profile
|
||||
# img = torch.rand(8 if torch.cuda.is_available() else 1, 3, 640, 640).to(device)
|
||||
# y = model(img, profile=True)
|
||||
|
||||
# Tensorboard
|
||||
# from torch.utils.tensorboard import SummaryWriter
|
||||
# tb_writer = SummaryWriter()
|
||||
# print("Run 'tensorboard --logdir=models/runs' to view tensorboard at http://localhost:6006/")
|
||||
# tb_writer.add_graph(model.model, img) # add model to tensorboard
|
||||
# tb_writer.add_image('test', img[0], dataformats='CWH') # add model to tensorboard
|
||||
@@ -0,0 +1,39 @@
|
||||
# Usage: pip install -r requirements.txt
|
||||
|
||||
# Base ----------------------------------------
|
||||
matplotlib>=3.2.2
|
||||
numpy>=1.18.5
|
||||
opencv>=4.1.1
|
||||
Pillow>=7.1.2
|
||||
PyYAML>=5.3.1
|
||||
requests>=2.23.0
|
||||
scipy>=1.4.1
|
||||
#torch>=1.7.0,!=1.12.0
|
||||
#torchvision>=0.8.1,!=0.13.0
|
||||
tqdm>=4.41.0
|
||||
protobuf<4.21.3
|
||||
|
||||
# Logging -------------------------------------
|
||||
tensorboard>=2.4.1
|
||||
# wandb
|
||||
|
||||
# Plotting ------------------------------------
|
||||
pandas>=1.1.4
|
||||
seaborn>=0.11.0
|
||||
|
||||
# Export --------------------------------------
|
||||
# coremltools>=4.1 # CoreML export
|
||||
# onnx>=1.9.0 # ONNX export
|
||||
# onnx-simplifier>=0.3.6 # ONNX simplifier
|
||||
# scikit-learn==0.19.2 # CoreML quantization
|
||||
# tensorflow>=2.4.1 # TFLite export
|
||||
# tensorflowjs>=3.9.0 # TF.js export
|
||||
# openvino-dev # OpenVINO export
|
||||
|
||||
# Extras --------------------------------------
|
||||
#ipython # interactive notebook
|
||||
#psutil # system utilization
|
||||
#thop # FLOPs computation
|
||||
# albumentations>=1.0.3
|
||||
# pycocotools>=2.0 # COCO mAP
|
||||
# roboflow
|
||||
Binary file not shown.
|
After Width: | Height: | Size: 83 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 82 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 83 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 83 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 78 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 196 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 204 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 238 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 246 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 243 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 249 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 81 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 83 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 84 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 83 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 78 KiB |
BIN
Binary file not shown.
BIN
Binary file not shown.
@@ -0,0 +1,30 @@
|
||||
lr0: 0.01
|
||||
lrf: 0.1
|
||||
momentum: 0.937
|
||||
weight_decay: 0.0005
|
||||
warmup_epochs: 3.0
|
||||
warmup_momentum: 0.8
|
||||
warmup_bias_lr: 0.1
|
||||
box: 0.05
|
||||
cls: 0.3
|
||||
cls_pw: 1.0
|
||||
obj: 0.7
|
||||
obj_pw: 1.0
|
||||
iou_t: 0.2
|
||||
anchor_t: 4.0
|
||||
fl_gamma: 0.0
|
||||
hsv_h: 0.015
|
||||
hsv_s: 0.7
|
||||
hsv_v: 0.4
|
||||
degrees: 0.0
|
||||
translate: 0.2
|
||||
scale: 0.5
|
||||
shear: 0.0
|
||||
perspective: 0.0
|
||||
flipud: 0.0
|
||||
fliplr: 0.5
|
||||
mosaic: 1.0
|
||||
mixup: 0.0
|
||||
copy_paste: 0.0
|
||||
paste_in: 0.0
|
||||
loss_ota: 1
|
||||
@@ -0,0 +1,43 @@
|
||||
weights: ./runs/train/yolov7-custom7/weights/last.pt
|
||||
cfg: ''
|
||||
data: data/data.yaml
|
||||
hyp: hyp/hyp.scratch.custom.yaml
|
||||
epochs: 300
|
||||
batch_size: 32
|
||||
img_size:
|
||||
- 640
|
||||
- 640
|
||||
rect: false
|
||||
resume: true
|
||||
nosave: false
|
||||
notest: false
|
||||
noautoanchor: false
|
||||
evolve: false
|
||||
bucket: ''
|
||||
cache_images: false
|
||||
image_weights: false
|
||||
device: ''
|
||||
multi_scale: false
|
||||
single_cls: false
|
||||
adam: false
|
||||
sync_bn: false
|
||||
local_rank: -1
|
||||
workers: 24
|
||||
project: runs/train
|
||||
entity: null
|
||||
name: yolov7-custom
|
||||
exist_ok: false
|
||||
quad: false
|
||||
linear_lr: false
|
||||
label_smoothing: 0.0
|
||||
upload_dataset: false
|
||||
bbox_interval: -1
|
||||
save_period: -1
|
||||
artifact_alias: latest
|
||||
freeze:
|
||||
- 0
|
||||
v5_metric: false
|
||||
world_size: 1
|
||||
global_rank: -1
|
||||
save_dir: runs/train/yolov7-custom7
|
||||
total_batch_size: 32
|
||||
Binary file not shown.
|
After Width: | Height: | Size: 255 KiB |
@@ -0,0 +1,300 @@
|
||||
0/299 7.49G 0.04468 0.01796 0 0.06264 87 640 0.7777 0.6004 0.7016 0.5741 0.04719 0.007429 0
|
||||
1/299 5.66G 0.03713 0.01763 0 0.05476 87 640 0.7596 0.5931 0.6851 0.5284 0.04633 0.00784 0
|
||||
2/299 5.9G 0.03728 0.01787 0 0.05515 64 640 0.7899 0.5904 0.6901 0.5618 0.04848 0.007925 0
|
||||
3/299 5.87G 0.03721 0.01785 0 0.05507 128 640 0.7593 0.5991 0.6911 0.5547 0.04522 0.007872 0
|
||||
4/299 5.9G 0.03695 0.01766 0 0.05461 39 640 0.7454 0.6202 0.7018 0.5798 0.04608 0.007888 0
|
||||
5/299 5.93G 0.03676 0.01754 0 0.0543 137 640 0.7464 0.6215 0.705 0.581 0.04514 0.007977 0
|
||||
6/299 5.86G 0.03661 0.01751 0 0.05412 56 640 0.7571 0.6202 0.7054 0.5784 0.04585 0.00789 0
|
||||
7/299 5.86G 0.0364 0.01736 0 0.05376 54 640 0.7173 0.6373 0.7068 0.579 0.0451 0.007987 0
|
||||
8/299 5.85G 0.03634 0.01742 0 0.05376 57 640 0.7662 0.608 0.7047 0.5751 0.04503 0.008045 0
|
||||
9/299 5.86G 0.03623 0.01741 0 0.05363 47 640 0.7784 0.6025 0.7073 0.5809 0.04456 0.008036 0
|
||||
10/299 5.9G 0.03609 0.01736 0 0.05346 71 640 0.7502 0.621 0.7086 0.5802 0.04439 0.008041 0
|
||||
11/299 5.86G 0.03618 0.01724 0 0.05342 53 640 0.7426 0.6317 0.7125 0.5834 0.04437 0.007992 0
|
||||
12/299 5.86G 0.03605 0.0173 0 0.05335 88 640 0.7776 0.6078 0.7135 0.5833 0.04424 0.007987 0
|
||||
13/299 5.86G 0.03607 0.01725 0 0.05332 59 640 0.7796 0.6095 0.7141 0.5833 0.04418 0.00799 0
|
||||
14/299 5.9G 0.03605 0.01732 0 0.05337 60 640 0.7595 0.6237 0.715 0.586 0.04409 0.007978 0
|
||||
15/299 5.9G 0.036 0.01727 0 0.05327 47 640 0.7671 0.6205 0.7158 0.5863 0.04409 0.007974 0
|
||||
16/299 5.87G 0.03607 0.01724 0 0.05332 45 640 0.7687 0.6195 0.7162 0.5854 0.04404 0.00798 0
|
||||
17/299 5.91G 0.03601 0.01725 0 0.05327 47 640 0.7911 0.6043 0.7166 0.5853 0.04405 0.007981 0
|
||||
18/299 5.87G 0.03592 0.01727 0 0.05319 89 640 0.7766 0.6136 0.7173 0.5873 0.04405 0.007966 0
|
||||
19/299 5.91G 0.03595 0.01723 0 0.05318 100 640 0.7518 0.629 0.7177 0.587 0.04406 0.007955 0
|
||||
20/299 5.92G 0.03585 0.01722 0 0.05307 33 640 0.7668 0.6212 0.7187 0.5882 0.04402 0.007938 0
|
||||
21/299 5.92G 0.03583 0.0172 0 0.05304 41 640 0.7607 0.6255 0.7191 0.5882 0.04397 0.007931 0
|
||||
22/299 5.87G 0.03584 0.01725 0 0.05309 63 640 0.7635 0.6244 0.7199 0.5875 0.04394 0.007917 0
|
||||
23/299 5.87G 0.03586 0.01716 0 0.05302 61 640 0.7617 0.6256 0.7203 0.5873 0.04397 0.00791 0
|
||||
24/299 5.87G 0.03579 0.01722 0 0.05301 31 640 0.7636 0.6249 0.7205 0.588 0.04393 0.007901 0
|
||||
25/299 5.86G 0.03574 0.01711 0 0.05286 42 640 0.7632 0.6244 0.7207 0.5876 0.04391 0.007886 0
|
||||
26/299 5.86G 0.03571 0.01722 0 0.05293 116 640 0.7758 0.6163 0.7213 0.5881 0.04382 0.007885 0
|
||||
27/299 5.86G 0.03578 0.01721 0 0.05299 41 640 0.7705 0.6179 0.7213 0.5878 0.04377 0.007879 0
|
||||
28/299 5.92G 0.03573 0.01717 0 0.0529 47 640 0.7665 0.6208 0.7216 0.5881 0.04377 0.007871 0
|
||||
29/299 5.87G 0.03582 0.0171 0 0.05292 51 640 0.7625 0.6251 0.722 0.5887 0.04381 0.007856 0
|
||||
30/299 5.87G 0.03573 0.01714 0 0.05287 71 640 0.7647 0.6226 0.7223 0.5886 0.04376 0.007859 0
|
||||
31/299 5.87G 0.03569 0.0171 0 0.05279 41 640 0.7632 0.6245 0.7229 0.5892 0.04369 0.007852 0
|
||||
32/299 5.87G 0.03571 0.01713 0 0.05284 68 640 0.7562 0.6299 0.7228 0.5911 0.04364 0.007851 0
|
||||
33/299 5.87G 0.03569 0.01715 0 0.05284 68 640 0.7694 0.6207 0.7222 0.5907 0.04356 0.00785 0
|
||||
34/299 5.86G 0.03567 0.01714 0 0.05281 39 640 0.76 0.6259 0.7219 0.5928 0.04348 0.007849 0
|
||||
35/299 5.87G 0.03559 0.01711 0 0.0527 77 640 0.7766 0.6126 0.7211 0.5942 0.04343 0.007854 0
|
||||
36/299 5.87G 0.03573 0.01711 0 0.05284 44 640 0.7425 0.6344 0.7205 0.594 0.04342 0.007862 0
|
||||
37/299 5.86G 0.03564 0.01712 0 0.05276 37 640 0.745 0.6322 0.7202 0.5945 0.04342 0.007857 0
|
||||
38/299 5.92G 0.0356 0.01714 0 0.05274 56 640 0.7534 0.6276 0.7206 0.5942 0.04345 0.007856 0
|
||||
39/299 5.86G 0.03554 0.01706 0 0.0526 60 640 0.7461 0.632 0.7204 0.5949 0.04346 0.007852 0
|
||||
40/299 5.87G 0.03559 0.01705 0 0.05264 37 640 0.8028 0.5948 0.7199 0.595 0.04346 0.007851 0
|
||||
41/299 5.86G 0.03552 0.01714 0 0.05266 40 640 0.7811 0.607 0.7197 0.5951 0.04346 0.007853 0
|
||||
42/299 5.86G 0.03557 0.01707 0 0.05264 56 640 0.7981 0.5982 0.7198 0.5958 0.04342 0.007856 0
|
||||
43/299 5.87G 0.0355 0.01701 0 0.05251 33 640 0.8033 0.5938 0.719 0.5958 0.04343 0.007865 0
|
||||
44/299 5.86G 0.03554 0.01703 0 0.05257 39 640 0.8001 0.5936 0.7184 0.596 0.0434 0.007872 0
|
||||
45/299 5.87G 0.03559 0.01707 0 0.05266 46 640 0.7607 0.6158 0.7174 0.5961 0.04339 0.007877 0
|
||||
46/299 5.9G 0.03559 0.01705 0 0.05264 42 640 0.7626 0.6139 0.7174 0.5959 0.04342 0.007887 0
|
||||
47/299 5.91G 0.03546 0.01701 0 0.05247 34 640 0.7602 0.6156 0.7177 0.5963 0.04337 0.007893 0
|
||||
48/299 5.86G 0.03549 0.01699 0 0.05248 74 640 0.7588 0.6174 0.7178 0.5967 0.04339 0.007901 0
|
||||
49/299 5.87G 0.03542 0.01707 0 0.05249 50 640 0.7597 0.6171 0.7174 0.596 0.04338 0.007909 0
|
||||
50/299 5.86G 0.03551 0.0171 0 0.05261 78 640 0.7609 0.6185 0.7176 0.5965 0.04333 0.007913 0
|
||||
51/299 5.92G 0.03543 0.01698 0 0.05241 65 640 0.7572 0.6222 0.7178 0.5966 0.04325 0.007917 0
|
||||
52/299 5.85G 0.03554 0.01706 0 0.0526 42 640 0.7436 0.6285 0.7171 0.5964 0.04314 0.007932 0
|
||||
53/299 5.9G 0.03548 0.01714 0 0.05262 96 640 0.7469 0.6261 0.7173 0.5963 0.04311 0.007942 0
|
||||
54/299 5.91G 0.03538 0.017 0 0.05238 40 640 0.7536 0.6211 0.7173 0.5965 0.04307 0.007946 0
|
||||
55/299 5.91G 0.03547 0.01704 0 0.05252 41 640 0.7198 0.6459 0.7176 0.5966 0.04304 0.007947 0
|
||||
56/299 5.91G 0.03539 0.01706 0 0.05245 60 640 0.7455 0.6271 0.7182 0.5968 0.043 0.007952 0
|
||||
57/299 5.92G 0.03543 0.01706 0 0.0525 72 640 0.7465 0.6261 0.7181 0.5963 0.04297 0.007965 0
|
||||
58/299 5.92G 0.03543 0.01708 0 0.05251 35 640 0.7415 0.6293 0.7175 0.5961 0.043 0.007974 0
|
||||
59/299 5.92G 0.03544 0.01703 0 0.05246 57 640 0.7573 0.6193 0.718 0.5953 0.04304 0.007976 0
|
||||
60/299 5.91G 0.03537 0.01702 0 0.05239 39 640 0.7636 0.6146 0.7179 0.595 0.04305 0.007972 0
|
||||
61/299 5.86G 0.03541 0.01707 0 0.05248 36 640 0.7552 0.6197 0.7178 0.5954 0.04306 0.007978 0
|
||||
62/299 5.91G 0.03535 0.01706 0 0.05241 61 640 0.7562 0.6205 0.7178 0.5957 0.04304 0.007979 0
|
||||
63/299 5.87G 0.03537 0.01701 0 0.05238 91 640 0.7522 0.6222 0.7174 0.5952 0.04312 0.00797 0
|
||||
64/299 5.92G 0.03531 0.01693 0 0.05224 47 640 0.7446 0.6276 0.7168 0.5952 0.04312 0.007968 0
|
||||
65/299 5.86G 0.03531 0.01698 0 0.05229 112 640 0.7534 0.6215 0.7163 0.5955 0.04311 0.00797 0
|
||||
66/299 5.93G 0.03532 0.01706 0 0.05237 55 640 0.7521 0.6222 0.7165 0.5946 0.04311 0.007974 0
|
||||
67/299 5.92G 0.03527 0.01697 0 0.05225 26 640 0.7548 0.6207 0.7166 0.5945 0.04311 0.007983 0
|
||||
68/299 5.87G 0.03523 0.01697 0 0.0522 52 640 0.7421 0.6293 0.7166 0.5947 0.04306 0.007991 0
|
||||
69/299 5.87G 0.03529 0.01706 0 0.05235 75 640 0.7388 0.6305 0.7152 0.5945 0.043 0.008003 0
|
||||
70/299 5.92G 0.03523 0.01694 0 0.05217 40 640 0.7614 0.6146 0.715 0.5941 0.04296 0.008012 0
|
||||
71/299 5.87G 0.03524 0.01704 0 0.05228 57 640 0.7437 0.627 0.7147 0.5941 0.04296 0.008016 0
|
||||
72/299 5.87G 0.03521 0.01692 0 0.05213 64 640 0.7436 0.6283 0.7142 0.594 0.04294 0.00802 0
|
||||
73/299 5.9G 0.03524 0.01701 0 0.05225 47 640 0.738 0.6322 0.7143 0.5941 0.04293 0.008024 0
|
||||
74/299 5.91G 0.03522 0.0169 0 0.05212 83 640 0.7439 0.6276 0.7146 0.5943 0.04297 0.008029 0
|
||||
75/299 5.86G 0.03524 0.01694 0 0.05218 49 640 0.7319 0.6364 0.7144 0.5942 0.04296 0.008035 0
|
||||
76/299 5.91G 0.03516 0.01697 0 0.05214 45 640 0.735 0.6329 0.7137 0.5938 0.04296 0.008043 0
|
||||
77/299 5.85G 0.03526 0.017 0 0.05226 73 640 0.7388 0.6298 0.7133 0.5936 0.04293 0.008054 0
|
||||
78/299 5.87G 0.03508 0.01696 0 0.05204 72 640 0.7342 0.6344 0.7128 0.5933 0.04292 0.00806 0
|
||||
79/299 5.86G 0.03513 0.01691 0 0.05204 38 640 0.728 0.6381 0.713 0.5934 0.0429 0.008066 0
|
||||
80/299 5.91G 0.03515 0.01699 0 0.05214 43 640 0.7194 0.6466 0.7139 0.5946 0.04287 0.008062 0
|
||||
81/299 5.85G 0.03513 0.01686 0 0.05199 22 640 0.7349 0.6332 0.7142 0.5949 0.04283 0.008058 0
|
||||
82/299 5.91G 0.03513 0.01696 0 0.05209 83 640 0.7341 0.6334 0.714 0.595 0.04282 0.008062 0
|
||||
83/299 5.93G 0.03513 0.01701 0 0.05214 49 640 0.7576 0.6183 0.7135 0.5947 0.04285 0.008067 0
|
||||
84/299 5.86G 0.03519 0.01695 0 0.05214 54 640 0.7491 0.6247 0.713 0.5943 0.04286 0.00808 0
|
||||
85/299 5.91G 0.03511 0.01698 0 0.05209 26 640 0.7257 0.6391 0.7124 0.5939 0.04285 0.008092 0
|
||||
86/299 5.87G 0.03504 0.01686 0 0.0519 40 640 0.7198 0.6427 0.7126 0.5939 0.04278 0.008095 0
|
||||
87/299 5.92G 0.03507 0.01687 0 0.05194 53 640 0.7182 0.6459 0.7137 0.5951 0.04275 0.008092 0
|
||||
88/299 5.92G 0.03506 0.01688 0 0.05193 36 640 0.7207 0.6437 0.7134 0.595 0.04273 0.008099 0
|
||||
89/299 5.86G 0.03507 0.01693 0 0.052 72 640 0.7256 0.6397 0.7141 0.5948 0.04272 0.008099 0
|
||||
90/299 5.86G 0.03503 0.01685 0 0.05188 66 640 0.7218 0.641 0.7142 0.5949 0.04269 0.008104 0
|
||||
91/299 5.87G 0.03503 0.01693 0 0.05196 50 640 0.7211 0.6427 0.7142 0.5952 0.04264 0.008103 0
|
||||
92/299 5.86G 0.03509 0.01691 0 0.052 46 640 0.708 0.653 0.7154 0.5958 0.04258 0.008103 0
|
||||
93/299 5.86G 0.03502 0.01692 0 0.05193 39 640 0.73 0.6364 0.7162 0.5964 0.04255 0.008104 0
|
||||
94/299 5.86G 0.03504 0.01685 0 0.05189 62 640 0.7153 0.6488 0.7167 0.5963 0.04257 0.00811 0
|
||||
95/299 5.91G 0.035 0.01686 0 0.05186 86 640 0.7273 0.6381 0.7154 0.5953 0.04254 0.008118 0
|
||||
96/299 5.87G 0.03494 0.01687 0 0.05181 52 640 0.7147 0.6508 0.7151 0.5955 0.04255 0.008122 0
|
||||
97/299 5.87G 0.03486 0.01688 0 0.05174 51 640 0.7424 0.6282 0.7147 0.5954 0.04251 0.008131 0
|
||||
98/299 5.85G 0.03494 0.01678 0 0.05172 62 640 0.7351 0.6351 0.7155 0.5959 0.04251 0.008132 0
|
||||
99/299 5.91G 0.03492 0.01689 0 0.05181 40 640 0.7285 0.6391 0.7149 0.596 0.04256 0.00814 0
|
||||
100/299 5.87G 0.03487 0.01685 0 0.05172 51 640 0.7239 0.641 0.7145 0.5952 0.04259 0.008144 0
|
||||
101/299 5.92G 0.03492 0.01677 0 0.05169 58 640 0.7355 0.632 0.7138 0.595 0.04263 0.008149 0
|
||||
102/299 5.92G 0.03493 0.01686 0 0.05179 36 640 0.7357 0.632 0.714 0.5947 0.04266 0.008147 0
|
||||
103/299 5.87G 0.03491 0.01678 0 0.05169 66 640 0.733 0.6317 0.7137 0.5953 0.04267 0.008152 0
|
||||
104/299 5.86G 0.03489 0.01692 0 0.05181 65 640 0.7421 0.6259 0.7136 0.5955 0.04268 0.008159 0
|
||||
105/299 5.86G 0.03483 0.01685 0 0.05168 29 640 0.7311 0.6317 0.7136 0.5962 0.04267 0.008158 0
|
||||
106/299 5.86G 0.03476 0.01679 0 0.05154 49 640 0.743 0.6232 0.7143 0.5963 0.0426 0.008163 0
|
||||
107/299 5.87G 0.03481 0.01676 0 0.05157 41 640 0.7346 0.6297 0.7138 0.5966 0.04254 0.008162 0
|
||||
108/299 5.92G 0.03479 0.01679 0 0.05158 47 640 0.7301 0.6342 0.7138 0.597 0.04257 0.008159 0
|
||||
109/299 5.87G 0.03479 0.0168 0 0.0516 47 640 0.7398 0.626 0.7139 0.5967 0.04257 0.008157 0
|
||||
110/299 5.86G 0.03474 0.01674 0 0.05148 32 640 0.7358 0.63 0.7142 0.5967 0.04259 0.008156 0
|
||||
111/299 5.87G 0.03485 0.01683 0 0.05167 39 640 0.7557 0.6146 0.7131 0.5958 0.0426 0.008164 0
|
||||
112/299 5.86G 0.03482 0.01682 0 0.05163 45 640 0.7505 0.6161 0.7128 0.5955 0.04255 0.008177 0
|
||||
113/299 5.86G 0.03466 0.01678 0 0.05144 70 640 0.7376 0.6259 0.7134 0.5959 0.04251 0.008177 0
|
||||
114/299 5.87G 0.03467 0.01683 0 0.0515 71 640 0.7335 0.6289 0.7139 0.5967 0.04249 0.008177 0
|
||||
115/299 5.92G 0.03465 0.01677 0 0.05142 64 640 0.7361 0.6259 0.7128 0.5957 0.04247 0.008187 0
|
||||
116/299 5.92G 0.03461 0.01675 0 0.05135 37 640 0.738 0.6265 0.7123 0.5952 0.04245 0.008192 0
|
||||
117/299 5.92G 0.03467 0.01678 0 0.05146 43 640 0.7682 0.6053 0.7122 0.5953 0.04244 0.008199 0
|
||||
118/299 5.87G 0.03462 0.01674 0 0.05136 64 640 0.7717 0.6039 0.7118 0.5948 0.04241 0.008208 0
|
||||
119/299 5.86G 0.0346 0.01678 0 0.05138 56 640 0.7445 0.6204 0.7115 0.5943 0.04244 0.008208 0
|
||||
120/299 5.92G 0.03462 0.01675 0 0.05137 42 640 0.7447 0.6223 0.7124 0.595 0.04244 0.008204 0
|
||||
121/299 5.91G 0.03455 0.01677 0 0.05133 101 640 0.7216 0.6379 0.7121 0.5947 0.04246 0.008216 0
|
||||
122/299 5.86G 0.03458 0.01674 0 0.05132 56 640 0.7364 0.6281 0.7118 0.5948 0.04246 0.008228 0
|
||||
123/299 5.86G 0.03457 0.01675 0 0.05132 61 640 0.7332 0.6305 0.7123 0.5953 0.04245 0.008225 0
|
||||
124/299 5.87G 0.03457 0.01679 0 0.05137 47 640 0.7331 0.629 0.712 0.5949 0.04244 0.008231 0
|
||||
125/299 5.87G 0.03447 0.01667 0 0.05114 38 640 0.7001 0.6572 0.7133 0.5955 0.04239 0.008231 0
|
||||
126/299 5.87G 0.03452 0.0168 0 0.05133 40 640 0.7018 0.6591 0.7138 0.5961 0.04236 0.008239 0
|
||||
127/299 5.86G 0.03449 0.01669 0 0.05118 47 640 0.7198 0.6422 0.7138 0.5963 0.04232 0.008249 0
|
||||
128/299 5.85G 0.03441 0.01668 0 0.05109 50 640 0.7178 0.6437 0.714 0.5965 0.0423 0.008253 0
|
||||
129/299 5.87G 0.03448 0.01666 0 0.05114 52 640 0.7266 0.6386 0.7145 0.5965 0.04226 0.008256 0
|
||||
130/299 5.92G 0.03449 0.01664 0 0.05113 50 640 0.7258 0.6413 0.7143 0.5968 0.04224 0.008257 0
|
||||
131/299 5.87G 0.03443 0.01673 0 0.05115 39 640 0.7422 0.629 0.7143 0.5971 0.04223 0.008265 0
|
||||
132/299 5.87G 0.03443 0.01663 0 0.05106 45 640 0.7287 0.6381 0.7146 0.5968 0.04228 0.008257 0
|
||||
133/299 5.91G 0.03449 0.01664 0 0.05113 67 640 0.7551 0.6178 0.7153 0.5981 0.04229 0.00825 0
|
||||
134/299 5.91G 0.03438 0.01665 0 0.05104 38 640 0.7468 0.6215 0.7153 0.598 0.04224 0.008255 0
|
||||
135/299 5.9G 0.03435 0.01664 0 0.051 76 640 0.7478 0.6205 0.7148 0.5975 0.04224 0.008258 0
|
||||
136/299 5.86G 0.03434 0.01663 0 0.05097 60 640 0.7374 0.6271 0.7146 0.5964 0.04221 0.008256 0
|
||||
137/299 5.92G 0.03435 0.01665 0 0.05099 45 640 0.7573 0.6145 0.7145 0.5964 0.04225 0.008255 0
|
||||
138/299 5.91G 0.03436 0.01661 0 0.05097 45 640 0.7551 0.6173 0.7148 0.5975 0.04227 0.008244 0
|
||||
139/299 5.91G 0.03423 0.01658 0 0.05081 41 640 0.7502 0.6217 0.7151 0.5974 0.04227 0.008242 0
|
||||
140/299 5.92G 0.03436 0.01664 0 0.05101 51 640 0.7521 0.6192 0.7149 0.5969 0.04225 0.008244 0
|
||||
141/299 5.86G 0.03431 0.01654 0 0.05085 51 640 0.7545 0.618 0.7148 0.5962 0.04225 0.008241 0
|
||||
142/299 5.87G 0.03427 0.0166 0 0.05087 51 640 0.764 0.6109 0.7144 0.5955 0.04221 0.00825 0
|
||||
143/299 5.87G 0.03422 0.01658 0 0.0508 59 640 0.7605 0.6134 0.7137 0.5951 0.04221 0.008243 0
|
||||
144/299 5.91G 0.03425 0.01657 0 0.05082 47 640 0.7754 0.6056 0.7134 0.594 0.04222 0.008243 0
|
||||
145/299 5.92G 0.03425 0.01652 0 0.05077 59 640 0.7374 0.6292 0.7124 0.5931 0.0422 0.008253 0
|
||||
146/299 5.87G 0.0342 0.01653 0 0.05074 47 640 0.7351 0.6322 0.7119 0.5934 0.04216 0.008257 0
|
||||
147/299 5.92G 0.03414 0.01652 0 0.05067 80 640 0.7652 0.6124 0.7121 0.593 0.04213 0.008264 0
|
||||
148/299 5.92G 0.03413 0.01652 0 0.05066 81 640 0.7541 0.6173 0.7124 0.5941 0.04206 0.008263 0
|
||||
149/299 5.91G 0.03408 0.01644 0 0.05052 62 640 0.7514 0.6204 0.7125 0.594 0.04209 0.008265 0
|
||||
150/299 5.87G 0.03409 0.01653 0 0.05062 55 640 0.7437 0.6261 0.7126 0.5943 0.04208 0.00827 0
|
||||
151/299 5.87G 0.03417 0.0165 0 0.05067 74 640 0.7358 0.6348 0.7132 0.5936 0.04211 0.008276 0
|
||||
152/299 5.91G 0.0341 0.0165 0 0.05059 54 640 0.7532 0.6204 0.7121 0.593 0.04211 0.008286 0
|
||||
153/299 5.86G 0.03399 0.01644 0 0.05043 64 640 0.7348 0.6327 0.7109 0.5918 0.0421 0.008292 0
|
||||
154/299 5.86G 0.03405 0.01645 0 0.05051 50 640 0.7359 0.6326 0.7104 0.5914 0.04208 0.0083 0
|
||||
155/299 5.93G 0.03405 0.01653 0 0.05058 61 640 0.7584 0.6153 0.7092 0.5911 0.04209 0.008307 0
|
||||
156/299 5.87G 0.03401 0.0165 0 0.05051 52 640 0.7632 0.6136 0.7092 0.5912 0.04212 0.008309 0
|
||||
157/299 5.86G 0.034 0.0164 0 0.0504 38 640 0.7465 0.6253 0.7093 0.5909 0.04214 0.008313 0
|
||||
158/299 5.87G 0.03392 0.01641 0 0.05032 42 640 0.7372 0.6303 0.7085 0.5907 0.0421 0.008324 0
|
||||
159/299 5.86G 0.03395 0.01635 0 0.0503 58 640 0.7295 0.6359 0.7083 0.5911 0.04209 0.008324 0
|
||||
160/299 5.87G 0.03392 0.01639 0 0.0503 70 640 0.751 0.6198 0.7085 0.5904 0.04204 0.008329 0
|
||||
161/299 5.87G 0.03381 0.0164 0 0.05021 62 640 0.7347 0.632 0.7082 0.5899 0.04202 0.008341 0
|
||||
162/299 5.91G 0.03394 0.01632 0 0.05026 32 640 0.7287 0.6359 0.709 0.5912 0.04203 0.008342 0
|
||||
163/299 5.91G 0.03386 0.01632 0 0.05019 41 640 0.7229 0.6395 0.7097 0.5923 0.0421 0.008334 0
|
||||
164/299 5.86G 0.03388 0.01634 0 0.05022 74 640 0.7188 0.642 0.7096 0.5921 0.04215 0.008329 0
|
||||
165/299 5.87G 0.03383 0.01631 0 0.05014 43 640 0.7219 0.6395 0.7098 0.5924 0.04222 0.008331 0
|
||||
166/299 5.92G 0.0338 0.01632 0 0.05012 52 640 0.7081 0.6483 0.71 0.5917 0.04221 0.008338 0
|
||||
167/299 5.87G 0.03381 0.01632 0 0.05013 47 640 0.7129 0.6452 0.7091 0.5907 0.04216 0.008348 0
|
||||
168/299 5.87G 0.03367 0.01634 0 0.05001 80 640 0.7085 0.6475 0.7088 0.5913 0.04214 0.008348 0
|
||||
169/299 5.91G 0.0337 0.01629 0 0.04999 69 640 0.7136 0.6471 0.71 0.5918 0.04214 0.008343 0
|
||||
170/299 5.86G 0.0338 0.01634 0 0.05013 36 640 0.7166 0.6431 0.7099 0.5913 0.04214 0.008359 0
|
||||
171/299 5.86G 0.03374 0.01634 0 0.05008 71 640 0.7312 0.6325 0.7106 0.5913 0.04205 0.008372 0
|
||||
172/299 5.92G 0.0336 0.01625 0 0.04985 33 640 0.7327 0.6305 0.7101 0.5918 0.04202 0.008379 0
|
||||
173/299 5.91G 0.03359 0.01627 0 0.04987 38 640 0.7372 0.6273 0.7098 0.5913 0.04201 0.008386 0
|
||||
174/299 5.86G 0.03359 0.01627 0 0.04987 90 640 0.7314 0.628 0.7087 0.5908 0.04198 0.0084 0
|
||||
175/299 5.87G 0.03355 0.01624 0 0.04979 46 640 0.7247 0.6339 0.7077 0.5897 0.04196 0.008412 0
|
||||
176/299 5.91G 0.03354 0.01619 0 0.04973 33 640 0.7197 0.6378 0.7075 0.5896 0.04193 0.008424 0
|
||||
177/299 5.86G 0.03349 0.01623 0 0.04972 141 640 0.7251 0.6344 0.7077 0.5888 0.04195 0.008425 0
|
||||
178/299 5.87G 0.03354 0.01621 0 0.04976 54 640 0.7179 0.6375 0.7074 0.5876 0.04198 0.008428 0
|
||||
179/299 5.86G 0.03351 0.01612 0 0.04963 58 640 0.7305 0.63 0.7076 0.5875 0.04199 0.008433 0
|
||||
180/299 5.87G 0.03346 0.01619 0 0.04965 58 640 0.7299 0.6315 0.7078 0.5882 0.04199 0.008431 0
|
||||
181/299 5.87G 0.03343 0.01618 0 0.04961 61 640 0.7253 0.6343 0.7081 0.5881 0.04201 0.008428 0
|
||||
182/299 5.87G 0.03335 0.01621 0 0.04956 42 640 0.7357 0.6271 0.7075 0.5881 0.04203 0.008435 0
|
||||
183/299 5.86G 0.03337 0.01608 0 0.04946 60 640 0.7394 0.6266 0.7072 0.588 0.04205 0.008437 0
|
||||
184/299 5.93G 0.03339 0.01606 0 0.04945 40 640 0.7373 0.6266 0.7072 0.5875 0.04211 0.008436 0
|
||||
185/299 5.92G 0.03331 0.01605 0 0.04936 72 640 0.7328 0.6298 0.7069 0.587 0.0421 0.008438 0
|
||||
186/299 5.85G 0.0333 0.0161 0 0.0494 73 640 0.7601 0.6102 0.7073 0.5869 0.04216 0.008436 0
|
||||
187/299 5.87G 0.03328 0.01601 0 0.04929 61 640 0.7555 0.614 0.7074 0.5876 0.04215 0.008434 0
|
||||
188/299 5.87G 0.03325 0.01604 0 0.04929 44 640 0.7519 0.617 0.7079 0.5872 0.04216 0.008447 0
|
||||
189/299 5.85G 0.03315 0.01602 0 0.04917 93 640 0.753 0.6158 0.7074 0.5871 0.04212 0.00845 0
|
||||
190/299 5.86G 0.03309 0.01603 0 0.04912 112 640 0.7519 0.6183 0.7081 0.5871 0.0421 0.008458 0
|
||||
191/299 5.86G 0.03313 0.01598 0 0.04911 56 640 0.7485 0.6195 0.7083 0.5876 0.0421 0.008464 0
|
||||
192/299 5.92G 0.03308 0.01597 0 0.04905 65 640 0.7449 0.6185 0.7088 0.5884 0.04208 0.008463 0
|
||||
193/299 5.91G 0.03309 0.01599 0 0.04907 56 640 0.7409 0.619 0.7083 0.5881 0.04203 0.008469 0
|
||||
194/299 5.91G 0.03293 0.01591 0 0.04884 76 640 0.7209 0.6344 0.7074 0.5879 0.04203 0.008483 0
|
||||
195/299 5.91G 0.03304 0.01587 0 0.04891 29 640 0.7269 0.6289 0.7078 0.5878 0.04205 0.008488 0
|
||||
196/299 5.92G 0.03297 0.01588 0 0.04885 60 640 0.7187 0.6332 0.7063 0.5868 0.04206 0.008498 0
|
||||
197/299 5.86G 0.03296 0.01592 0 0.04887 53 640 0.7268 0.6268 0.7056 0.5858 0.04206 0.008515 0
|
||||
198/299 5.87G 0.03297 0.01592 0 0.04888 49 640 0.7383 0.619 0.7051 0.5854 0.042 0.008531 0
|
||||
199/299 5.93G 0.03295 0.01578 0 0.04873 50 640 0.7433 0.6162 0.705 0.5848 0.04201 0.008535 0
|
||||
200/299 5.92G 0.03282 0.01584 0 0.04866 38 640 0.7571 0.6065 0.7041 0.5845 0.04196 0.008548 0
|
||||
201/299 5.91G 0.03283 0.01577 0 0.04859 78 640 0.7309 0.6246 0.7033 0.5838 0.04194 0.008554 0
|
||||
202/299 5.86G 0.03282 0.01578 0 0.04861 55 640 0.738 0.6215 0.7027 0.5828 0.04192 0.008566 0
|
||||
203/299 5.86G 0.03278 0.01575 0 0.04853 75 640 0.7294 0.6259 0.7023 0.5831 0.04189 0.008573 0
|
||||
204/299 5.85G 0.03277 0.01576 0 0.04853 30 640 0.7334 0.6231 0.7025 0.583 0.04185 0.008572 0
|
||||
205/299 5.92G 0.03274 0.01569 0 0.04843 41 640 0.7618 0.6056 0.7021 0.5827 0.04185 0.008574 0
|
||||
206/299 5.85G 0.0327 0.01571 0 0.04841 33 640 0.7533 0.6114 0.7026 0.5831 0.04183 0.008573 0
|
||||
207/299 5.91G 0.03264 0.01573 0 0.04837 86 640 0.7368 0.6225 0.7027 0.5836 0.04175 0.008578 0
|
||||
208/299 5.91G 0.03258 0.01566 0 0.04823 75 640 0.7374 0.6215 0.7034 0.5844 0.04167 0.008581 0
|
||||
209/299 5.85G 0.03261 0.01567 0 0.04828 43 640 0.7464 0.6156 0.7026 0.5836 0.04164 0.008597 0
|
||||
210/299 5.85G 0.03255 0.01559 0 0.04814 34 640 0.7426 0.619 0.7023 0.5832 0.04161 0.008605 0
|
||||
211/299 5.91G 0.03247 0.01554 0 0.04801 46 640 0.7411 0.6207 0.7018 0.5833 0.04164 0.008599 0
|
||||
212/299 5.92G 0.0324 0.01555 0 0.04795 64 640 0.7501 0.6161 0.7014 0.5823 0.04164 0.008607 0
|
||||
213/299 5.92G 0.03238 0.01556 0 0.04794 52 640 0.7411 0.6212 0.7011 0.5813 0.04165 0.008612 0
|
||||
214/299 5.85G 0.03237 0.0155 0 0.04787 36 640 0.7357 0.626 0.7006 0.5808 0.04163 0.008619 0
|
||||
215/299 5.85G 0.03234 0.0155 0 0.04784 41 640 0.7415 0.619 0.7003 0.5803 0.04166 0.008622 0
|
||||
216/299 5.91G 0.03223 0.01551 0 0.04775 53 640 0.738 0.6202 0.6989 0.5802 0.0417 0.008629 0
|
||||
217/299 5.91G 0.03217 0.01541 0 0.04759 67 640 0.7323 0.6251 0.6985 0.5796 0.04166 0.008636 0
|
||||
218/299 5.9G 0.03218 0.01547 0 0.04764 71 640 0.7199 0.63 0.6978 0.5785 0.04165 0.008658 0
|
||||
219/299 5.86G 0.03214 0.01532 0 0.04745 60 640 0.7446 0.6095 0.6964 0.5778 0.04166 0.008665 0
|
||||
220/299 5.86G 0.0322 0.0154 0 0.0476 43 640 0.7447 0.6124 0.6954 0.5771 0.04166 0.008677 0
|
||||
221/299 5.86G 0.03209 0.01534 0 0.04743 56 640 0.7383 0.6164 0.6963 0.5772 0.04164 0.008685 0
|
||||
222/299 5.87G 0.03207 0.01527 0 0.04734 43 640 0.7477 0.6075 0.696 0.577 0.04166 0.008692 0
|
||||
223/299 5.87G 0.03203 0.0152 0 0.04723 60 640 0.7344 0.6173 0.6968 0.5769 0.04166 0.008703 0
|
||||
224/299 5.87G 0.03202 0.01532 0 0.04733 52 640 0.7465 0.6065 0.6958 0.5759 0.04168 0.008722 0
|
||||
225/299 5.92G 0.03193 0.01518 0 0.0471 36 640 0.7362 0.6119 0.6942 0.5747 0.0417 0.008732 0
|
||||
226/299 5.86G 0.03183 0.01521 0 0.04704 65 640 0.752 0.5997 0.6928 0.573 0.04173 0.008741 0
|
||||
227/299 5.87G 0.03194 0.01513 0 0.04706 53 640 0.7469 0.6019 0.6917 0.5724 0.04173 0.00875 0
|
||||
228/299 5.87G 0.03181 0.01517 0 0.04697 53 640 0.7439 0.6035 0.6917 0.5718 0.04181 0.008752 0
|
||||
229/299 5.87G 0.03174 0.01509 0 0.04683 35 640 0.7532 0.5975 0.6911 0.571 0.04191 0.008751 0
|
||||
230/299 5.87G 0.03171 0.01516 0 0.04688 52 640 0.7457 0.6053 0.6914 0.5713 0.0419 0.008756 0
|
||||
231/299 5.91G 0.03161 0.01502 0 0.04664 39 640 0.736 0.6085 0.6897 0.5707 0.04195 0.008766 0
|
||||
232/299 5.87G 0.03159 0.01497 0 0.04656 24 640 0.7198 0.6207 0.6888 0.5703 0.04199 0.008772 0
|
||||
233/299 5.91G 0.03157 0.01502 0 0.0466 86 640 0.716 0.6229 0.6886 0.5698 0.04204 0.008774 0
|
||||
234/299 5.85G 0.03154 0.01499 0 0.04653 68 640 0.7439 0.6041 0.6882 0.5696 0.04208 0.008777 0
|
||||
235/299 5.91G 0.0315 0.01497 0 0.04646 28 640 0.7351 0.6104 0.6886 0.5692 0.04213 0.008787 0
|
||||
236/299 5.86G 0.03138 0.01487 0 0.04625 38 640 0.7393 0.6077 0.6875 0.5685 0.04219 0.008796 0
|
||||
237/299 5.87G 0.03136 0.01492 0 0.04628 56 640 0.7507 0.6018 0.6878 0.5685 0.04222 0.008795 0
|
||||
238/299 5.87G 0.03136 0.01488 0 0.04624 41 640 0.7604 0.5948 0.6875 0.5681 0.04221 0.0088 0
|
||||
239/299 5.92G 0.03131 0.01483 0 0.04615 76 640 0.7265 0.6185 0.6874 0.5673 0.0422 0.008816 0
|
||||
240/299 5.87G 0.03123 0.01479 0 0.04601 25 640 0.7276 0.6173 0.6867 0.5667 0.04219 0.008833 0
|
||||
241/299 5.86G 0.0312 0.01477 0 0.04597 53 640 0.7048 0.6349 0.687 0.5664 0.04221 0.008844 0
|
||||
242/299 5.91G 0.03108 0.01466 0 0.04574 55 640 0.7054 0.6344 0.6864 0.5661 0.04221 0.008852 0
|
||||
243/299 5.91G 0.03105 0.0147 0 0.04576 58 640 0.7171 0.6244 0.6862 0.566 0.04226 0.00885 0
|
||||
244/299 5.92G 0.03107 0.01461 0 0.04568 27 640 0.7101 0.6272 0.6852 0.565 0.04223 0.008862 0
|
||||
245/299 5.86G 0.03104 0.01464 0 0.04568 52 640 0.7073 0.6252 0.6831 0.5634 0.04226 0.008877 0
|
||||
246/299 5.91G 0.03092 0.01464 0 0.04556 76 640 0.7017 0.6288 0.6826 0.5624 0.04225 0.008885 0
|
||||
247/299 5.92G 0.03089 0.01458 0 0.04546 72 640 0.6953 0.6329 0.6811 0.5611 0.04223 0.008909 0
|
||||
248/299 5.92G 0.03079 0.0145 0 0.04529 42 640 0.6812 0.6427 0.6811 0.5612 0.04222 0.008921 0
|
||||
249/299 5.91G 0.03078 0.01448 0 0.04525 75 640 0.6841 0.6395 0.6809 0.5615 0.04216 0.00893 0
|
||||
250/299 5.91G 0.03075 0.01441 0 0.04515 42 640 0.7165 0.6131 0.68 0.5614 0.04211 0.008935 0
|
||||
251/299 5.91G 0.03076 0.01445 0 0.04521 46 640 0.6728 0.6491 0.6809 0.5618 0.04212 0.008945 0
|
||||
252/299 5.92G 0.03071 0.01438 0 0.04509 50 640 0.6867 0.6389 0.6797 0.5609 0.04216 0.008952 0
|
||||
253/299 5.86G 0.03062 0.01434 0 0.04496 96 640 0.6919 0.6356 0.6786 0.56 0.04215 0.008965 0
|
||||
254/299 5.86G 0.0306 0.0143 0 0.04491 41 640 0.6932 0.6283 0.6768 0.5578 0.04212 0.008993 0
|
||||
255/299 5.92G 0.03049 0.01424 0 0.04473 96 640 0.6993 0.6207 0.6749 0.5551 0.04205 0.009022 0
|
||||
256/299 5.91G 0.03043 0.01423 0 0.04466 48 640 0.7027 0.619 0.6738 0.5541 0.04202 0.009041 0
|
||||
257/299 5.86G 0.0304 0.01425 0 0.04465 63 640 0.7033 0.6149 0.6722 0.5523 0.04204 0.009059 0
|
||||
258/299 5.87G 0.03043 0.01421 0 0.04464 42 640 0.7019 0.6166 0.6724 0.5517 0.04205 0.009069 0
|
||||
259/299 5.87G 0.03032 0.01415 0 0.04447 47 640 0.7026 0.6172 0.6734 0.5525 0.04205 0.009083 0
|
||||
260/299 5.91G 0.03026 0.01408 0 0.04435 111 640 0.6985 0.6222 0.6732 0.552 0.04203 0.009096 0
|
||||
261/299 5.91G 0.03022 0.01408 0 0.04431 39 640 0.6977 0.6224 0.6731 0.5519 0.04202 0.009121 0
|
||||
262/299 5.86G 0.03023 0.01403 0 0.04425 79 640 0.6995 0.6185 0.6715 0.5512 0.04205 0.009147 0
|
||||
263/299 5.91G 0.03011 0.01407 0 0.04418 46 640 0.7091 0.6109 0.6714 0.5501 0.04204 0.009162 0
|
||||
264/299 5.91G 0.03007 0.01403 0 0.04411 53 640 0.6996 0.6195 0.6688 0.5483 0.04208 0.009174 0
|
||||
265/299 5.87G 0.03004 0.01394 0 0.04398 28 640 0.682 0.6351 0.6678 0.5479 0.0421 0.009182 0
|
||||
266/299 5.85G 0.02998 0.0139 0 0.04388 48 640 0.6893 0.6268 0.6663 0.5458 0.0421 0.009201 0
|
||||
267/299 5.86G 0.02998 0.01389 0 0.04387 81 640 0.6558 0.6535 0.6657 0.5446 0.0421 0.009226 0
|
||||
268/299 5.87G 0.02988 0.01385 0 0.04373 59 640 0.6553 0.654 0.6645 0.5427 0.04214 0.009248 0
|
||||
269/299 5.87G 0.02985 0.0138 0 0.04365 48 640 0.6552 0.6525 0.6638 0.5418 0.0422 0.009262 0
|
||||
270/299 5.91G 0.02971 0.01382 0 0.04354 50 640 0.6821 0.6293 0.6654 0.5429 0.04223 0.009267 0
|
||||
271/299 5.86G 0.02979 0.01374 0 0.04353 41 640 0.6799 0.6329 0.6652 0.5429 0.04227 0.009284 0
|
||||
272/299 5.85G 0.02974 0.01376 0 0.0435 47 640 0.6686 0.6408 0.6648 0.5418 0.04233 0.009302 0
|
||||
273/299 5.91G 0.02961 0.01364 0 0.04326 44 640 0.659 0.6478 0.6641 0.5419 0.04237 0.009311 0
|
||||
274/299 5.86G 0.02959 0.01365 0 0.04324 45 640 0.6858 0.6249 0.6644 0.5422 0.04236 0.009321 0
|
||||
275/299 5.86G 0.02962 0.01371 0 0.04333 74 640 0.6925 0.6202 0.6647 0.5429 0.04234 0.00933 0
|
||||
276/299 5.86G 0.02958 0.0137 0 0.04328 60 640 0.689 0.6232 0.665 0.5433 0.04235 0.009345 0
|
||||
277/299 5.86G 0.02952 0.0136 0 0.04312 68 640 0.6987 0.6149 0.6642 0.5433 0.04231 0.009358 0
|
||||
278/299 5.92G 0.02953 0.0136 0 0.04313 58 640 0.7026 0.6109 0.6637 0.5427 0.04234 0.009375 0
|
||||
279/299 5.85G 0.02936 0.0135 0 0.04286 43 640 0.7233 0.597 0.6629 0.543 0.0424 0.009387 0
|
||||
280/299 5.87G 0.02941 0.01357 0 0.04298 86 640 0.7221 0.5987 0.663 0.5426 0.04248 0.009393 0
|
||||
281/299 5.87G 0.02931 0.01351 0 0.04282 51 640 0.6854 0.6261 0.6636 0.5436 0.04253 0.009399 0
|
||||
282/299 5.86G 0.02935 0.01349 0 0.04284 42 640 0.6888 0.6227 0.6632 0.5427 0.04257 0.009413 0
|
||||
283/299 5.86G 0.02924 0.01346 0 0.04269 46 640 0.6918 0.6181 0.6621 0.5417 0.04259 0.00943 0
|
||||
284/299 5.92G 0.02919 0.01341 0 0.0426 30 640 0.6908 0.6222 0.6626 0.5415 0.04258 0.009437 0
|
||||
285/299 5.86G 0.02913 0.01337 0 0.04251 52 640 0.6898 0.6221 0.6635 0.5421 0.04259 0.009444 0
|
||||
286/299 5.92G 0.02911 0.01345 0 0.04256 70 640 0.6768 0.6351 0.6629 0.5419 0.04259 0.009454 0
|
||||
287/299 5.87G 0.02909 0.01338 0 0.04247 61 640 0.6667 0.6439 0.6628 0.5415 0.04258 0.009465 0
|
||||
288/299 5.86G 0.02903 0.01327 0 0.0423 77 640 0.6738 0.6376 0.6629 0.5412 0.04259 0.009474 0
|
||||
289/299 5.86G 0.02896 0.01333 0 0.04228 37 640 0.669 0.6378 0.6619 0.5407 0.04262 0.009483 0
|
||||
290/299 5.87G 0.02902 0.01331 0 0.04233 79 640 0.6626 0.6413 0.6618 0.5411 0.04267 0.009484 0
|
||||
291/299 5.87G 0.02892 0.01325 0 0.04216 57 640 0.674 0.6312 0.6615 0.5413 0.0427 0.009496 0
|
||||
292/299 5.92G 0.02886 0.01326 0 0.04212 51 640 0.6699 0.6356 0.6604 0.5406 0.04274 0.009513 0
|
||||
293/299 5.87G 0.02884 0.01321 0 0.04205 74 640 0.674 0.6332 0.6611 0.54 0.0428 0.009521 0
|
||||
294/299 5.87G 0.02878 0.01322 0 0.042 46 640 0.6715 0.6375 0.661 0.5396 0.04283 0.009522 0
|
||||
295/299 5.91G 0.02877 0.01319 0 0.04196 46 640 0.6611 0.6464 0.6605 0.5391 0.04283 0.009531 0
|
||||
296/299 5.86G 0.02869 0.01313 0 0.04182 35 640 0.6792 0.6322 0.6616 0.5396 0.04283 0.009532 0
|
||||
297/299 5.91G 0.02872 0.01319 0 0.04191 98 640 0.701 0.6163 0.6619 0.5394 0.04282 0.009539 0
|
||||
298/299 5.92G 0.0287 0.01315 0 0.04185 47 640 0.6962 0.6193 0.6637 0.5406 0.04284 0.009546 0
|
||||
299/299 5.9G 0.02875 0.01319 0 0.04195 44 640 0.6892 0.6242 0.6642 0.5413 0.04285 0.009554 0
|
||||
Binary file not shown.
|
After Width: | Height: | Size: 195 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 204 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 251 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 254 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 249 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 253 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 303 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 349 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 355 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 357 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 340 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 334 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 330 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 345 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 359 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 357 KiB |
@@ -0,0 +1,353 @@
|
||||
import argparse
|
||||
import json
|
||||
import os
|
||||
from pathlib import Path
|
||||
from threading import Thread
|
||||
|
||||
import numpy as np
|
||||
import torch
|
||||
import yaml
|
||||
from tqdm import tqdm
|
||||
|
||||
from models.experimental import attempt_load
|
||||
from utils.datasets import create_dataloader
|
||||
from utils.general import coco80_to_coco91_class, check_dataset, check_file, check_img_size, check_requirements, \
|
||||
box_iou, non_max_suppression, scale_coords, xyxy2xywh, xywh2xyxy, set_logging, increment_path, colorstr
|
||||
from utils.metrics import ap_per_class, ConfusionMatrix
|
||||
from utils.plots import plot_images, output_to_target, plot_study_txt
|
||||
from utils.torch_utils import select_device, time_synchronized, TracedModel
|
||||
|
||||
|
||||
def test(data,
|
||||
weights=None,
|
||||
batch_size=32,
|
||||
imgsz=640,
|
||||
conf_thres=0.001,
|
||||
iou_thres=0.6, # for NMS
|
||||
save_json=False,
|
||||
single_cls=False,
|
||||
augment=False,
|
||||
verbose=False,
|
||||
model=None,
|
||||
dataloader=None,
|
||||
save_dir=Path(''), # for saving images
|
||||
save_txt=False, # for auto-labelling
|
||||
save_hybrid=False, # for hybrid auto-labelling
|
||||
save_conf=False, # save auto-label confidences
|
||||
plots=True,
|
||||
wandb_logger=None,
|
||||
compute_loss=None,
|
||||
half_precision=True,
|
||||
trace=False,
|
||||
is_coco=False,
|
||||
v5_metric=False):
|
||||
# Initialize/load model and set device
|
||||
training = model is not None
|
||||
if training: # called by train.py
|
||||
device = next(model.parameters()).device # get model device
|
||||
|
||||
else: # called directly
|
||||
set_logging()
|
||||
device = select_device(opt.device, batch_size=batch_size)
|
||||
|
||||
# Directories
|
||||
save_dir = Path(increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok)) # increment run
|
||||
(save_dir / 'labels' if save_txt else save_dir).mkdir(parents=True, exist_ok=True) # make dir
|
||||
|
||||
# Load model
|
||||
model = attempt_load(weights, map_location=device) # load FP32 model
|
||||
gs = max(int(model.stride.max()), 32) # grid size (max stride)
|
||||
imgsz = check_img_size(imgsz, s=gs) # check img_size
|
||||
|
||||
if trace:
|
||||
model = TracedModel(model, device, imgsz)
|
||||
|
||||
# Half
|
||||
half = device.type != 'cpu' and half_precision # half precision only supported on CUDA
|
||||
if half:
|
||||
model.half()
|
||||
|
||||
# Configure
|
||||
model.eval()
|
||||
if isinstance(data, str):
|
||||
is_coco = data.endswith('coco.yaml')
|
||||
with open(data) as f:
|
||||
data = yaml.load(f, Loader=yaml.SafeLoader)
|
||||
check_dataset(data) # check
|
||||
nc = 1 if single_cls else int(data['nc']) # number of classes
|
||||
iouv = torch.linspace(0.5, 0.95, 10).to(device) # iou vector for mAP@0.5:0.95
|
||||
niou = iouv.numel()
|
||||
|
||||
# Logging
|
||||
log_imgs = 0
|
||||
if wandb_logger and wandb_logger.wandb:
|
||||
log_imgs = min(wandb_logger.log_imgs, 100)
|
||||
# Dataloader
|
||||
if not training:
|
||||
if device.type != 'cpu':
|
||||
model(torch.zeros(1, 3, imgsz, imgsz).to(device).type_as(next(model.parameters()))) # run once
|
||||
task = opt.task if opt.task in ('train', 'val', 'test') else 'val' # path to train/val/test images
|
||||
dataloader = create_dataloader(data[task], imgsz, batch_size, gs, opt, pad=0.5, rect=True,
|
||||
prefix=colorstr(f'{task}: '))[0]
|
||||
|
||||
if v5_metric:
|
||||
print("Testing with YOLOv5 AP metric...")
|
||||
|
||||
seen = 0
|
||||
confusion_matrix = ConfusionMatrix(nc=nc)
|
||||
names = {k: v for k, v in enumerate(model.names if hasattr(model, 'names') else model.module.names)}
|
||||
coco91class = coco80_to_coco91_class()
|
||||
s = ('%20s' + '%12s' * 6) % ('Class', 'Images', 'Labels', 'P', 'R', 'mAP@.5', 'mAP@.5:.95')
|
||||
p, r, f1, mp, mr, map50, map, t0, t1 = 0., 0., 0., 0., 0., 0., 0., 0., 0.
|
||||
loss = torch.zeros(3, device=device)
|
||||
jdict, stats, ap, ap_class, wandb_images = [], [], [], [], []
|
||||
for batch_i, (img, targets, paths, shapes) in enumerate(tqdm(dataloader, desc=s)):
|
||||
img = img.to(device, non_blocking=True)
|
||||
img = img.half() if half else img.float() # uint8 to fp16/32
|
||||
img /= 255.0 # 0 - 255 to 0.0 - 1.0
|
||||
targets = targets.to(device)
|
||||
nb, _, height, width = img.shape # batch size, channels, height, width
|
||||
|
||||
with torch.no_grad():
|
||||
# Run model
|
||||
t = time_synchronized()
|
||||
out, train_out = model(img, augment=augment) # inference and training outputs
|
||||
t0 += time_synchronized() - t
|
||||
|
||||
# Compute loss
|
||||
if compute_loss:
|
||||
loss += compute_loss([x.float() for x in train_out], targets)[1][:3] # box, obj, cls
|
||||
|
||||
# Run NMS
|
||||
targets[:, 2:] *= torch.Tensor([width, height, width, height]).to(device) # to pixels
|
||||
lb = [targets[targets[:, 0] == i, 1:] for i in range(nb)] if save_hybrid else [] # for autolabelling
|
||||
t = time_synchronized()
|
||||
out = non_max_suppression(out, conf_thres=conf_thres, iou_thres=iou_thres, labels=lb, multi_label=True)
|
||||
t1 += time_synchronized() - t
|
||||
|
||||
# Statistics per image
|
||||
for si, pred in enumerate(out):
|
||||
labels = targets[targets[:, 0] == si, 1:]
|
||||
nl = len(labels)
|
||||
tcls = labels[:, 0].tolist() if nl else [] # target class
|
||||
path = Path(paths[si])
|
||||
seen += 1
|
||||
|
||||
if len(pred) == 0:
|
||||
if nl:
|
||||
stats.append((torch.zeros(0, niou, dtype=torch.bool), torch.Tensor(), torch.Tensor(), tcls))
|
||||
continue
|
||||
|
||||
# Predictions
|
||||
predn = pred.clone()
|
||||
scale_coords(img[si].shape[1:], predn[:, :4], shapes[si][0], shapes[si][1]) # native-space pred
|
||||
|
||||
# Append to text file
|
||||
if save_txt:
|
||||
gn = torch.tensor(shapes[si][0])[[1, 0, 1, 0]] # normalization gain whwh
|
||||
for *xyxy, conf, cls in predn.tolist():
|
||||
xywh = (xyxy2xywh(torch.tensor(xyxy).view(1, 4)) / gn).view(-1).tolist() # normalized xywh
|
||||
line = (cls, *xywh, conf) if save_conf else (cls, *xywh) # label format
|
||||
with open(save_dir / 'labels' / (path.stem + '.txt'), 'a') as f:
|
||||
f.write(('%g ' * len(line)).rstrip() % line + '\n')
|
||||
|
||||
# W&B logging - Media Panel Plots
|
||||
if len(wandb_images) < log_imgs and wandb_logger.current_epoch > 0: # Check for test operation
|
||||
if wandb_logger.current_epoch % wandb_logger.bbox_interval == 0:
|
||||
box_data = [{"position": {"minX": xyxy[0], "minY": xyxy[1], "maxX": xyxy[2], "maxY": xyxy[3]},
|
||||
"class_id": int(cls),
|
||||
"box_caption": "%s %.3f" % (names[cls], conf),
|
||||
"scores": {"class_score": conf},
|
||||
"domain": "pixel"} for *xyxy, conf, cls in pred.tolist()]
|
||||
boxes = {"predictions": {"box_data": box_data, "class_labels": names}} # inference-space
|
||||
wandb_images.append(wandb_logger.wandb.Image(img[si], boxes=boxes, caption=path.name))
|
||||
wandb_logger.log_training_progress(predn, path, names) if wandb_logger and wandb_logger.wandb_run else None
|
||||
|
||||
# Append to pycocotools JSON dictionary
|
||||
if save_json:
|
||||
# [{"image_id": 42, "category_id": 18, "bbox": [258.15, 41.29, 348.26, 243.78], "score": 0.236}, ...
|
||||
image_id = int(path.stem) if path.stem.isnumeric() else path.stem
|
||||
box = xyxy2xywh(predn[:, :4]) # xywh
|
||||
box[:, :2] -= box[:, 2:] / 2 # xy center to top-left corner
|
||||
for p, b in zip(pred.tolist(), box.tolist()):
|
||||
jdict.append({'image_id': image_id,
|
||||
'category_id': coco91class[int(p[5])] if is_coco else int(p[5]),
|
||||
'bbox': [round(x, 3) for x in b],
|
||||
'score': round(p[4], 5)})
|
||||
|
||||
# Assign all predictions as incorrect
|
||||
correct = torch.zeros(pred.shape[0], niou, dtype=torch.bool, device=device)
|
||||
if nl:
|
||||
detected = [] # target indices
|
||||
tcls_tensor = labels[:, 0]
|
||||
|
||||
# target boxes
|
||||
tbox = xywh2xyxy(labels[:, 1:5])
|
||||
scale_coords(img[si].shape[1:], tbox, shapes[si][0], shapes[si][1]) # native-space labels
|
||||
if plots:
|
||||
confusion_matrix.process_batch(predn, torch.cat((labels[:, 0:1], tbox), 1))
|
||||
|
||||
# Per target class
|
||||
for cls in torch.unique(tcls_tensor):
|
||||
ti = (cls == tcls_tensor).nonzero(as_tuple=False).view(-1) # prediction indices
|
||||
pi = (cls == pred[:, 5]).nonzero(as_tuple=False).view(-1) # target indices
|
||||
|
||||
# Search for detections
|
||||
if pi.shape[0]:
|
||||
# Prediction to target ious
|
||||
ious, i = box_iou(predn[pi, :4], tbox[ti]).max(1) # best ious, indices
|
||||
|
||||
# Append detections
|
||||
detected_set = set()
|
||||
for j in (ious > iouv[0]).nonzero(as_tuple=False):
|
||||
d = ti[i[j]] # detected target
|
||||
if d.item() not in detected_set:
|
||||
detected_set.add(d.item())
|
||||
detected.append(d)
|
||||
correct[pi[j]] = ious[j] > iouv # iou_thres is 1xn
|
||||
if len(detected) == nl: # all targets already located in image
|
||||
break
|
||||
|
||||
# Append statistics (correct, conf, pcls, tcls)
|
||||
stats.append((correct.cpu(), pred[:, 4].cpu(), pred[:, 5].cpu(), tcls))
|
||||
|
||||
# Plot images
|
||||
if plots and batch_i < 3:
|
||||
f = save_dir / f'test_batch{batch_i}_labels.jpg' # labels
|
||||
Thread(target=plot_images, args=(img, targets, paths, f, names), daemon=True).start()
|
||||
f = save_dir / f'test_batch{batch_i}_pred.jpg' # predictions
|
||||
Thread(target=plot_images, args=(img, output_to_target(out), paths, f, names), daemon=True).start()
|
||||
|
||||
# Compute statistics
|
||||
stats = [np.concatenate(x, 0) for x in zip(*stats)] # to numpy
|
||||
if len(stats) and stats[0].any():
|
||||
p, r, ap, f1, ap_class = ap_per_class(*stats, plot=plots, v5_metric=v5_metric, save_dir=save_dir, names=names)
|
||||
ap50, ap = ap[:, 0], ap.mean(1) # AP@0.5, AP@0.5:0.95
|
||||
mp, mr, map50, map = p.mean(), r.mean(), ap50.mean(), ap.mean()
|
||||
nt = np.bincount(stats[3].astype(np.int64), minlength=nc) # number of targets per class
|
||||
else:
|
||||
nt = torch.zeros(1)
|
||||
|
||||
# Print results
|
||||
pf = '%20s' + '%12i' * 2 + '%12.3g' * 4 # print format
|
||||
print(pf % ('all', seen, nt.sum(), mp, mr, map50, map))
|
||||
|
||||
# Print results per class
|
||||
if (verbose or (nc < 50 and not training)) and nc > 1 and len(stats):
|
||||
for i, c in enumerate(ap_class):
|
||||
print(pf % (names[c], seen, nt[c], p[i], r[i], ap50[i], ap[i]))
|
||||
|
||||
# Print speeds
|
||||
t = tuple(x / seen * 1E3 for x in (t0, t1, t0 + t1)) + (imgsz, imgsz, batch_size) # tuple
|
||||
if not training:
|
||||
print('Speed: %.1f/%.1f/%.1f ms inference/NMS/total per %gx%g image at batch-size %g' % t)
|
||||
|
||||
# Plots
|
||||
if plots:
|
||||
confusion_matrix.plot(save_dir=save_dir, names=list(names.values()))
|
||||
if wandb_logger and wandb_logger.wandb:
|
||||
val_batches = [wandb_logger.wandb.Image(str(f), caption=f.name) for f in sorted(save_dir.glob('test*.jpg'))]
|
||||
wandb_logger.log({"Validation": val_batches})
|
||||
if wandb_images:
|
||||
wandb_logger.log({"Bounding Box Debugger/Images": wandb_images})
|
||||
|
||||
# Save JSON
|
||||
if save_json and len(jdict):
|
||||
w = Path(weights[0] if isinstance(weights, list) else weights).stem if weights is not None else '' # weights
|
||||
anno_json = './coco/annotations/instances_val2017.json' # annotations json
|
||||
pred_json = str(save_dir / f"{w}_predictions.json") # predictions json
|
||||
print('\nEvaluating pycocotools mAP... saving %s...' % pred_json)
|
||||
with open(pred_json, 'w') as f:
|
||||
json.dump(jdict, f)
|
||||
|
||||
try: # https://github.com/cocodataset/cocoapi/blob/master/PythonAPI/pycocoEvalDemo.ipynb
|
||||
from pycocotools.coco import COCO
|
||||
from pycocotools.cocoeval import COCOeval
|
||||
|
||||
anno = COCO(anno_json) # init annotations api
|
||||
pred = anno.loadRes(pred_json) # init predictions api
|
||||
eval = COCOeval(anno, pred, 'bbox')
|
||||
if is_coco:
|
||||
eval.params.imgIds = [int(Path(x).stem) for x in dataloader.dataset.img_files] # image IDs to evaluate
|
||||
eval.evaluate()
|
||||
eval.accumulate()
|
||||
eval.summarize()
|
||||
map, map50 = eval.stats[:2] # update results (mAP@0.5:0.95, mAP@0.5)
|
||||
except Exception as e:
|
||||
print(f'pycocotools unable to run: {e}')
|
||||
|
||||
# Return results
|
||||
model.float() # for training
|
||||
if not training:
|
||||
s = f"\n{len(list(save_dir.glob('labels/*.txt')))} labels saved to {save_dir / 'labels'}" if save_txt else ''
|
||||
print(f"Results saved to {save_dir}{s}")
|
||||
maps = np.zeros(nc) + map
|
||||
for i, c in enumerate(ap_class):
|
||||
maps[c] = ap[i]
|
||||
return (mp, mr, map50, map, *(loss.cpu() / len(dataloader)).tolist()), maps, t
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser(prog='test.py')
|
||||
parser.add_argument('--weights', nargs='+', type=str, default='yolov7.pt', help='model.pt path(s)')
|
||||
parser.add_argument('--data', type=str, default='data/coco.yaml', help='*.data path')
|
||||
parser.add_argument('--batch-size', type=int, default=32, help='size of each image batch')
|
||||
parser.add_argument('--img-size', type=int, default=640, help='inference size (pixels)')
|
||||
parser.add_argument('--conf-thres', type=float, default=0.001, help='object confidence threshold')
|
||||
parser.add_argument('--iou-thres', type=float, default=0.65, help='IOU threshold for NMS')
|
||||
parser.add_argument('--task', default='val', help='train, val, test, speed or study')
|
||||
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
|
||||
parser.add_argument('--single-cls', action='store_true', help='treat as single-class dataset')
|
||||
parser.add_argument('--augment', action='store_true', help='augmented inference')
|
||||
parser.add_argument('--verbose', action='store_true', help='report mAP by class')
|
||||
parser.add_argument('--save-txt', action='store_true', help='save results to *.txt')
|
||||
parser.add_argument('--save-hybrid', action='store_true', help='save label+prediction hybrid results to *.txt')
|
||||
parser.add_argument('--save-conf', action='store_true', help='save confidences in --save-txt labels')
|
||||
parser.add_argument('--save-json', action='store_true', help='save a cocoapi-compatible JSON results file')
|
||||
parser.add_argument('--project', default='runs/test', help='save to project/name')
|
||||
parser.add_argument('--name', default='exp', help='save to project/name')
|
||||
parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
|
||||
parser.add_argument('--no-trace', action='store_true', help='don`t trace model')
|
||||
parser.add_argument('--v5-metric', action='store_true', help='assume maximum recall as 1.0 in AP calculation')
|
||||
opt = parser.parse_args()
|
||||
opt.save_json |= opt.data.endswith('coco.yaml')
|
||||
opt.data = check_file(opt.data) # check file
|
||||
print(opt)
|
||||
#check_requirements()
|
||||
|
||||
if opt.task in ('train', 'val', 'test'): # run normally
|
||||
test(opt.data,
|
||||
opt.weights,
|
||||
opt.batch_size,
|
||||
opt.img_size,
|
||||
opt.conf_thres,
|
||||
opt.iou_thres,
|
||||
opt.save_json,
|
||||
opt.single_cls,
|
||||
opt.augment,
|
||||
opt.verbose,
|
||||
save_txt=opt.save_txt | opt.save_hybrid,
|
||||
save_hybrid=opt.save_hybrid,
|
||||
save_conf=opt.save_conf,
|
||||
trace=not opt.no_trace,
|
||||
v5_metric=opt.v5_metric
|
||||
)
|
||||
|
||||
elif opt.task == 'speed': # speed benchmarks
|
||||
for w in opt.weights:
|
||||
test(opt.data, w, opt.batch_size, opt.img_size, 0.25, 0.45, save_json=False, plots=False, v5_metric=opt.v5_metric)
|
||||
|
||||
elif opt.task == 'study': # run over a range of settings and save/plot
|
||||
# python test.py --task study --data coco.yaml --iou 0.65 --weights yolov7.pt
|
||||
x = list(range(256, 1536 + 128, 128)) # x axis (image sizes)
|
||||
for w in opt.weights:
|
||||
f = f'study_{Path(opt.data).stem}_{Path(w).stem}.txt' # filename to save to
|
||||
y = [] # y axis
|
||||
for i in x: # img-size
|
||||
print(f'\nRunning {f} point {i}...')
|
||||
r, _, t = test(opt.data, w, opt.batch_size, i, opt.conf_thres, opt.iou_thres, opt.save_json,
|
||||
plots=False, v5_metric=opt.v5_metric)
|
||||
y.append(r + t) # results and times
|
||||
np.savetxt(f, y, fmt='%10.4g') # save
|
||||
os.system('zip -r study.zip study_*.txt')
|
||||
plot_study_txt(x=x) # plot
|
||||
@@ -0,0 +1,706 @@
|
||||
import argparse
|
||||
import logging
|
||||
import math
|
||||
import os
|
||||
import random
|
||||
import time
|
||||
from copy import deepcopy
|
||||
from pathlib import Path
|
||||
from threading import Thread
|
||||
|
||||
import numpy as np
|
||||
import torch.distributed as dist
|
||||
import torch.nn as nn
|
||||
import torch.nn.functional as F
|
||||
import torch.optim as optim
|
||||
import torch.optim.lr_scheduler as lr_scheduler
|
||||
import torch.utils.data
|
||||
import yaml
|
||||
from torch.cuda import amp
|
||||
from torch.nn.parallel import DistributedDataParallel as DDP
|
||||
from torch.utils.tensorboard import SummaryWriter
|
||||
from tqdm import tqdm
|
||||
|
||||
import test # import test.py to get mAP after each epoch
|
||||
from models.experimental import attempt_load
|
||||
from models.yolo import Model
|
||||
from utils.autoanchor import check_anchors
|
||||
from utils.datasets import create_dataloader
|
||||
from utils.general import labels_to_class_weights, increment_path, labels_to_image_weights, init_seeds, \
|
||||
fitness, strip_optimizer, get_latest_run, check_dataset, check_file, check_git_status, check_img_size, \
|
||||
check_requirements, print_mutation, set_logging, one_cycle, colorstr
|
||||
from utils.google_utils import attempt_download
|
||||
from utils.loss import ComputeLoss, ComputeLossOTA
|
||||
from utils.plots import plot_images, plot_labels, plot_results, plot_evolution
|
||||
from utils.torch_utils import ModelEMA, select_device, intersect_dicts, torch_distributed_zero_first, is_parallel
|
||||
from utils.wandb_logging.wandb_utils import WandbLogger, check_wandb_resume
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
def train(hyp, opt, device, tb_writer=None):
|
||||
logger.info(colorstr('hyperparameters: ') + ', '.join(f'{k}={v}' for k, v in hyp.items()))
|
||||
save_dir, epochs, batch_size, total_batch_size, weights, rank, freeze = \
|
||||
Path(opt.save_dir), opt.epochs, opt.batch_size, opt.total_batch_size, opt.weights, opt.global_rank, opt.freeze
|
||||
|
||||
# Directories
|
||||
wdir = save_dir / 'weights'
|
||||
wdir.mkdir(parents=True, exist_ok=True) # make dir
|
||||
last = wdir / 'last.pt'
|
||||
best = wdir / 'best.pt'
|
||||
results_file = save_dir / 'results.txt'
|
||||
|
||||
# Save run settings
|
||||
with open(save_dir / 'hyp.yaml', 'w') as f:
|
||||
yaml.dump(hyp, f, sort_keys=False)
|
||||
with open(save_dir / 'opt.yaml', 'w') as f:
|
||||
yaml.dump(vars(opt), f, sort_keys=False)
|
||||
|
||||
# Configure
|
||||
plots = not opt.evolve # create plots
|
||||
cuda = device.type != 'cpu'
|
||||
init_seeds(2 + rank)
|
||||
with open(opt.data) as f:
|
||||
data_dict = yaml.load(f, Loader=yaml.SafeLoader) # data dict
|
||||
is_coco = opt.data.endswith('coco.yaml')
|
||||
|
||||
# Logging- Doing this before checking the dataset. Might update data_dict
|
||||
loggers = {'wandb': None} # loggers dict
|
||||
if rank in [-1, 0]:
|
||||
opt.hyp = hyp # add hyperparameters
|
||||
run_id = torch.load(weights, map_location=device).get('wandb_id') if weights.endswith('.pt') and os.path.isfile(weights) else None
|
||||
wandb_logger = WandbLogger(opt, Path(opt.save_dir).stem, run_id, data_dict)
|
||||
loggers['wandb'] = wandb_logger.wandb
|
||||
data_dict = wandb_logger.data_dict
|
||||
if wandb_logger.wandb:
|
||||
weights, epochs, hyp = opt.weights, opt.epochs, opt.hyp # WandbLogger might update weights, epochs if resuming
|
||||
|
||||
nc = 1 if opt.single_cls else int(data_dict['nc']) # number of classes
|
||||
names = ['item'] if opt.single_cls and len(data_dict['names']) != 1 else data_dict['names'] # class names
|
||||
assert len(names) == nc, '%g names found for nc=%g dataset in %s' % (len(names), nc, opt.data) # check
|
||||
|
||||
# Model
|
||||
pretrained = weights.endswith('.pt')
|
||||
if pretrained:
|
||||
# with torch_distributed_zero_first(rank):
|
||||
# attempt_download(weights) # download if not found locally
|
||||
ckpt = torch.load(weights, map_location=device) # load checkpoint
|
||||
model = Model(opt.cfg or ckpt['model'].yaml, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) # create
|
||||
exclude = ['anchor'] if (opt.cfg or hyp.get('anchors')) and not opt.resume else [] # exclude keys
|
||||
state_dict = ckpt['model'].float().state_dict() # to FP32
|
||||
state_dict = intersect_dicts(state_dict, model.state_dict(), exclude=exclude) # intersect
|
||||
model.load_state_dict(state_dict, strict=False) # load
|
||||
logger.info('Transferred %g/%g items from %s' % (len(state_dict), len(model.state_dict()), weights)) # report
|
||||
else:
|
||||
model = Model(opt.cfg, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) # create
|
||||
with torch_distributed_zero_first(rank):
|
||||
check_dataset(data_dict) # check
|
||||
train_path = data_dict['train']
|
||||
test_path = data_dict['val']
|
||||
|
||||
# Freeze
|
||||
freeze = [f'model.{x}.' for x in (freeze if len(freeze) > 1 else range(freeze[0]))] # parameter names to freeze (full or partial)
|
||||
for k, v in model.named_parameters():
|
||||
v.requires_grad = True # train all layers
|
||||
if any(x in k for x in freeze):
|
||||
print('freezing %s' % k)
|
||||
v.requires_grad = False
|
||||
|
||||
# Optimizer
|
||||
nbs = 64 # nominal batch size
|
||||
accumulate = max(round(nbs / total_batch_size), 1) # accumulate loss before optimizing
|
||||
hyp['weight_decay'] *= total_batch_size * accumulate / nbs # scale weight_decay
|
||||
logger.info(f"Scaled weight_decay = {hyp['weight_decay']}")
|
||||
|
||||
pg0, pg1, pg2 = [], [], [] # optimizer parameter groups
|
||||
for k, v in model.named_modules():
|
||||
if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter):
|
||||
pg2.append(v.bias) # biases
|
||||
if isinstance(v, nn.BatchNorm2d):
|
||||
pg0.append(v.weight) # no decay
|
||||
elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter):
|
||||
pg1.append(v.weight) # apply decay
|
||||
if hasattr(v, 'im'):
|
||||
if hasattr(v.im, 'implicit'):
|
||||
pg0.append(v.im.implicit)
|
||||
else:
|
||||
for iv in v.im:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'imc'):
|
||||
if hasattr(v.imc, 'implicit'):
|
||||
pg0.append(v.imc.implicit)
|
||||
else:
|
||||
for iv in v.imc:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'imb'):
|
||||
if hasattr(v.imb, 'implicit'):
|
||||
pg0.append(v.imb.implicit)
|
||||
else:
|
||||
for iv in v.imb:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'imo'):
|
||||
if hasattr(v.imo, 'implicit'):
|
||||
pg0.append(v.imo.implicit)
|
||||
else:
|
||||
for iv in v.imo:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'ia'):
|
||||
if hasattr(v.ia, 'implicit'):
|
||||
pg0.append(v.ia.implicit)
|
||||
else:
|
||||
for iv in v.ia:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'attn'):
|
||||
if hasattr(v.attn, 'logit_scale'):
|
||||
pg0.append(v.attn.logit_scale)
|
||||
if hasattr(v.attn, 'q_bias'):
|
||||
pg0.append(v.attn.q_bias)
|
||||
if hasattr(v.attn, 'v_bias'):
|
||||
pg0.append(v.attn.v_bias)
|
||||
if hasattr(v.attn, 'relative_position_bias_table'):
|
||||
pg0.append(v.attn.relative_position_bias_table)
|
||||
if hasattr(v, 'rbr_dense'):
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_origin'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_origin)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_avg_conv'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_avg_conv)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_pfir_conv'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_pfir_conv)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_1x1_kxk_idconv1'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_1x1_kxk_idconv1)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_1x1_kxk_conv2'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_1x1_kxk_conv2)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_gconv_dw'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_gconv_dw)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_gconv_pw'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_gconv_pw)
|
||||
if hasattr(v.rbr_dense, 'vector'):
|
||||
pg0.append(v.rbr_dense.vector)
|
||||
|
||||
if opt.adam:
|
||||
optimizer = optim.Adam(pg0, lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum
|
||||
else:
|
||||
optimizer = optim.SGD(pg0, lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True)
|
||||
|
||||
optimizer.add_param_group({'params': pg1, 'weight_decay': hyp['weight_decay']}) # add pg1 with weight_decay
|
||||
optimizer.add_param_group({'params': pg2}) # add pg2 (biases)
|
||||
logger.info('Optimizer groups: %g .bias, %g conv.weight, %g other' % (len(pg2), len(pg1), len(pg0)))
|
||||
del pg0, pg1, pg2
|
||||
|
||||
# Scheduler https://arxiv.org/pdf/1812.01187.pdf
|
||||
# https://pytorch.org/docs/stable/_modules/torch/optim/lr_scheduler.html#OneCycleLR
|
||||
if opt.linear_lr:
|
||||
lf = lambda x: (1 - x / (epochs - 1)) * (1.0 - hyp['lrf']) + hyp['lrf'] # linear
|
||||
else:
|
||||
lf = one_cycle(1, hyp['lrf'], epochs) # cosine 1->hyp['lrf']
|
||||
scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lf)
|
||||
# plot_lr_scheduler(optimizer, scheduler, epochs)
|
||||
|
||||
# EMA
|
||||
ema = ModelEMA(model) if rank in [-1, 0] else None
|
||||
|
||||
# Resume
|
||||
start_epoch, best_fitness = 0, 0.0
|
||||
if pretrained:
|
||||
# Optimizer
|
||||
if ckpt['optimizer'] is not None:
|
||||
optimizer.load_state_dict(ckpt['optimizer'])
|
||||
best_fitness = ckpt['best_fitness']
|
||||
|
||||
# EMA
|
||||
if ema and ckpt.get('ema'):
|
||||
ema.ema.load_state_dict(ckpt['ema'].float().state_dict())
|
||||
ema.updates = ckpt['updates']
|
||||
|
||||
# Results
|
||||
if ckpt.get('training_results') is not None:
|
||||
results_file.write_text(ckpt['training_results']) # write results.txt
|
||||
|
||||
# Epochs
|
||||
start_epoch = ckpt['epoch'] + 1
|
||||
if opt.resume:
|
||||
assert start_epoch > 0, '%s training to %g epochs is finished, nothing to resume.' % (weights, epochs)
|
||||
if epochs < start_epoch:
|
||||
logger.info('%s has been trained for %g epochs. Fine-tuning for %g additional epochs.' %
|
||||
(weights, ckpt['epoch'], epochs))
|
||||
epochs += ckpt['epoch'] # finetune additional epochs
|
||||
|
||||
del ckpt, state_dict
|
||||
|
||||
# Image sizes
|
||||
gs = max(int(model.stride.max()), 32) # grid size (max stride)
|
||||
nl = model.model[-1].nl # number of detection layers (used for scaling hyp['obj'])
|
||||
imgsz, imgsz_test = [check_img_size(x, gs) for x in opt.img_size] # verify imgsz are gs-multiples
|
||||
|
||||
# DP mode
|
||||
if cuda and rank == -1 and torch.cuda.device_count() > 1:
|
||||
model = torch.nn.DataParallel(model)
|
||||
|
||||
# SyncBatchNorm
|
||||
if opt.sync_bn and cuda and rank != -1:
|
||||
model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model).to(device)
|
||||
logger.info('Using SyncBatchNorm()')
|
||||
|
||||
# Trainloader
|
||||
dataloader, dataset = create_dataloader(train_path, imgsz, batch_size, gs, opt,
|
||||
hyp=hyp, augment=True, cache=opt.cache_images, rect=opt.rect, rank=rank,
|
||||
world_size=opt.world_size, workers=opt.workers,
|
||||
image_weights=opt.image_weights, quad=opt.quad, prefix=colorstr('train: '))
|
||||
mlc = np.concatenate(dataset.labels, 0)[:, 0].max() # max label class
|
||||
nb = len(dataloader) # number of batches
|
||||
assert mlc < nc, 'Label class %g exceeds nc=%g in %s. Possible class labels are 0-%g' % (mlc, nc, opt.data, nc - 1)
|
||||
|
||||
# Process 0
|
||||
if rank in [-1, 0]:
|
||||
testloader = create_dataloader(test_path, imgsz_test, batch_size * 2, gs, opt, # testloader
|
||||
hyp=hyp, cache=opt.cache_images and not opt.notest, rect=True, rank=-1,
|
||||
world_size=opt.world_size, workers=opt.workers,
|
||||
pad=0.5, prefix=colorstr('val: '))[0]
|
||||
|
||||
if not opt.resume:
|
||||
labels = np.concatenate(dataset.labels, 0)
|
||||
c = torch.tensor(labels[:, 0]) # classes
|
||||
# cf = torch.bincount(c.long(), minlength=nc) + 1. # frequency
|
||||
# model._initialize_biases(cf.to(device))
|
||||
if plots:
|
||||
#plot_labels(labels, names, save_dir, loggers)
|
||||
if tb_writer:
|
||||
tb_writer.add_histogram('classes', c, 0)
|
||||
|
||||
# Anchors
|
||||
if not opt.noautoanchor:
|
||||
check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz)
|
||||
model.half().float() # pre-reduce anchor precision
|
||||
|
||||
# DDP mode
|
||||
if cuda and rank != -1:
|
||||
model = DDP(model, device_ids=[opt.local_rank], output_device=opt.local_rank,
|
||||
# nn.MultiheadAttention incompatibility with DDP https://github.com/pytorch/pytorch/issues/26698
|
||||
find_unused_parameters=any(isinstance(layer, nn.MultiheadAttention) for layer in model.modules()))
|
||||
|
||||
# Model parameters
|
||||
hyp['box'] *= 3. / nl # scale to layers
|
||||
hyp['cls'] *= nc / 80. * 3. / nl # scale to classes and layers
|
||||
hyp['obj'] *= (imgsz / 640) ** 2 * 3. / nl # scale to image size and layers
|
||||
hyp['label_smoothing'] = opt.label_smoothing
|
||||
model.nc = nc # attach number of classes to model
|
||||
model.hyp = hyp # attach hyperparameters to model
|
||||
model.gr = 1.0 # iou loss ratio (obj_loss = 1.0 or iou)
|
||||
model.class_weights = labels_to_class_weights(dataset.labels, nc).to(device) * nc # attach class weights
|
||||
model.names = names
|
||||
|
||||
# Start training
|
||||
t0 = time.time()
|
||||
nw = max(round(hyp['warmup_epochs'] * nb), 1000) # number of warmup iterations, max(3 epochs, 1k iterations)
|
||||
# nw = min(nw, (epochs - start_epoch) / 2 * nb) # limit warmup to < 1/2 of training
|
||||
maps = np.zeros(nc) # mAP per class
|
||||
results = (0, 0, 0, 0, 0, 0, 0) # P, R, mAP@.5, mAP@.5-.95, val_loss(box, obj, cls)
|
||||
scheduler.last_epoch = start_epoch - 1 # do not move
|
||||
scaler = amp.GradScaler(enabled=cuda)
|
||||
compute_loss_ota = ComputeLossOTA(model) # init loss class
|
||||
compute_loss = ComputeLoss(model) # init loss class
|
||||
logger.info(f'Image sizes {imgsz} train, {imgsz_test} test\n'
|
||||
f'Using {dataloader.num_workers} dataloader workers\n'
|
||||
f'Logging results to {save_dir}\n'
|
||||
f'Starting training for {epochs} epochs...')
|
||||
torch.save(model, wdir / 'init.pt')
|
||||
for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------
|
||||
model.train()
|
||||
|
||||
# Update image weights (optional)
|
||||
if opt.image_weights:
|
||||
# Generate indices
|
||||
if rank in [-1, 0]:
|
||||
cw = model.class_weights.cpu().numpy() * (1 - maps) ** 2 / nc # class weights
|
||||
iw = labels_to_image_weights(dataset.labels, nc=nc, class_weights=cw) # image weights
|
||||
dataset.indices = random.choices(range(dataset.n), weights=iw, k=dataset.n) # rand weighted idx
|
||||
# Broadcast if DDP
|
||||
if rank != -1:
|
||||
indices = (torch.tensor(dataset.indices) if rank == 0 else torch.zeros(dataset.n)).int()
|
||||
dist.broadcast(indices, 0)
|
||||
if rank != 0:
|
||||
dataset.indices = indices.cpu().numpy()
|
||||
|
||||
# Update mosaic border
|
||||
# b = int(random.uniform(0.25 * imgsz, 0.75 * imgsz + gs) // gs * gs)
|
||||
# dataset.mosaic_border = [b - imgsz, -b] # height, width borders
|
||||
|
||||
mloss = torch.zeros(4, device=device) # mean losses
|
||||
if rank != -1:
|
||||
dataloader.sampler.set_epoch(epoch)
|
||||
pbar = enumerate(dataloader)
|
||||
logger.info(('\n' + '%10s' * 8) % ('Epoch', 'gpu_mem', 'box', 'obj', 'cls', 'total', 'labels', 'img_size'))
|
||||
if rank in [-1, 0]:
|
||||
pbar = tqdm(pbar, total=nb) # progress bar
|
||||
optimizer.zero_grad()
|
||||
for i, (imgs, targets, paths, _) in pbar: # batch -------------------------------------------------------------
|
||||
ni = i + nb * epoch # number integrated batches (since train start)
|
||||
imgs = imgs.to(device, non_blocking=True).float() / 255.0 # uint8 to float32, 0-255 to 0.0-1.0
|
||||
|
||||
# Warmup
|
||||
if ni <= nw:
|
||||
xi = [0, nw] # x interp
|
||||
# model.gr = np.interp(ni, xi, [0.0, 1.0]) # iou loss ratio (obj_loss = 1.0 or iou)
|
||||
accumulate = max(1, np.interp(ni, xi, [1, nbs / total_batch_size]).round())
|
||||
for j, x in enumerate(optimizer.param_groups):
|
||||
# bias lr falls from 0.1 to lr0, all other lrs rise from 0.0 to lr0
|
||||
x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 2 else 0.0, x['initial_lr'] * lf(epoch)])
|
||||
if 'momentum' in x:
|
||||
x['momentum'] = np.interp(ni, xi, [hyp['warmup_momentum'], hyp['momentum']])
|
||||
|
||||
# Multi-scale
|
||||
if opt.multi_scale:
|
||||
sz = random.randrange(imgsz * 0.5, imgsz * 1.5 + gs) // gs * gs # size
|
||||
sf = sz / max(imgs.shape[2:]) # scale factor
|
||||
if sf != 1:
|
||||
ns = [math.ceil(x * sf / gs) * gs for x in imgs.shape[2:]] # new shape (stretched to gs-multiple)
|
||||
imgs = F.interpolate(imgs, size=ns, mode='bilinear', align_corners=False)
|
||||
|
||||
# Forward
|
||||
with amp.autocast(enabled=cuda):
|
||||
pred = model(imgs) # forward
|
||||
if 'loss_ota' not in hyp or hyp['loss_ota'] == 1:
|
||||
loss, loss_items = compute_loss_ota(pred, targets.to(device), imgs) # loss scaled by batch_size
|
||||
else:
|
||||
loss, loss_items = compute_loss(pred, targets.to(device)) # loss scaled by batch_size
|
||||
if rank != -1:
|
||||
loss *= opt.world_size # gradient averaged between devices in DDP mode
|
||||
if opt.quad:
|
||||
loss *= 4.
|
||||
|
||||
# Backward
|
||||
scaler.scale(loss).backward()
|
||||
|
||||
# Optimize
|
||||
if ni % accumulate == 0:
|
||||
scaler.step(optimizer) # optimizer.step
|
||||
scaler.update()
|
||||
optimizer.zero_grad()
|
||||
if ema:
|
||||
ema.update(model)
|
||||
|
||||
# Print
|
||||
if rank in [-1, 0]:
|
||||
mloss = (mloss * i + loss_items) / (i + 1) # update mean losses
|
||||
mem = '%.3gG' % (torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0) # (GB)
|
||||
s = ('%10s' * 2 + '%10.4g' * 6) % (
|
||||
'%g/%g' % (epoch, epochs - 1), mem, *mloss, targets.shape[0], imgs.shape[-1])
|
||||
pbar.set_description(s)
|
||||
|
||||
# Plot
|
||||
if plots and ni < 10:
|
||||
f = save_dir / f'train_batch{ni}.jpg' # filename
|
||||
Thread(target=plot_images, args=(imgs, targets, paths, f), daemon=True).start()
|
||||
# if tb_writer:
|
||||
# tb_writer.add_image(f, result, dataformats='HWC', global_step=epoch)
|
||||
# tb_writer.add_graph(torch.jit.trace(model, imgs, strict=False), []) # add model graph
|
||||
elif plots and ni == 10 and wandb_logger.wandb:
|
||||
wandb_logger.log({"Mosaics": [wandb_logger.wandb.Image(str(x), caption=x.name) for x in
|
||||
save_dir.glob('train*.jpg') if x.exists()]})
|
||||
|
||||
# end batch ------------------------------------------------------------------------------------------------
|
||||
# end epoch ----------------------------------------------------------------------------------------------------
|
||||
|
||||
# Scheduler
|
||||
lr = [x['lr'] for x in optimizer.param_groups] # for tensorboard
|
||||
scheduler.step()
|
||||
|
||||
# DDP process 0 or single-GPU
|
||||
if rank in [-1, 0]:
|
||||
# mAP
|
||||
ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'gr', 'names', 'stride', 'class_weights'])
|
||||
final_epoch = epoch + 1 == epochs
|
||||
if not opt.notest or final_epoch: # Calculate mAP
|
||||
wandb_logger.current_epoch = epoch + 1
|
||||
results, maps, times = test.test(data_dict,
|
||||
batch_size=batch_size * 2,
|
||||
imgsz=imgsz_test,
|
||||
model=ema.ema,
|
||||
single_cls=opt.single_cls,
|
||||
dataloader=testloader,
|
||||
save_dir=save_dir,
|
||||
verbose=nc < 50 and final_epoch,
|
||||
plots=plots and final_epoch,
|
||||
wandb_logger=wandb_logger,
|
||||
compute_loss=compute_loss,
|
||||
is_coco=is_coco,
|
||||
v5_metric=opt.v5_metric)
|
||||
|
||||
# Write
|
||||
with open(results_file, 'a') as f:
|
||||
f.write(s + '%10.4g' * 7 % results + '\n') # append metrics, val_loss
|
||||
if len(opt.name) and opt.bucket:
|
||||
os.system('gsutil cp %s gs://%s/results/results%s.txt' % (results_file, opt.bucket, opt.name))
|
||||
|
||||
# Log
|
||||
tags = ['train/box_loss', 'train/obj_loss', 'train/cls_loss', # train loss
|
||||
'metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95',
|
||||
'val/box_loss', 'val/obj_loss', 'val/cls_loss', # val loss
|
||||
'x/lr0', 'x/lr1', 'x/lr2'] # params
|
||||
for x, tag in zip(list(mloss[:-1]) + list(results) + lr, tags):
|
||||
if tb_writer:
|
||||
tb_writer.add_scalar(tag, x, epoch) # tensorboard
|
||||
if wandb_logger.wandb:
|
||||
wandb_logger.log({tag: x}) # W&B
|
||||
|
||||
# Update best mAP
|
||||
fi = fitness(np.array(results).reshape(1, -1)) # weighted combination of [P, R, mAP@.5, mAP@.5-.95]
|
||||
if fi > best_fitness:
|
||||
best_fitness = fi
|
||||
wandb_logger.end_epoch(best_result=best_fitness == fi)
|
||||
|
||||
# Save model
|
||||
if (not opt.nosave) or (final_epoch and not opt.evolve): # if save
|
||||
ckpt = {'epoch': epoch,
|
||||
'best_fitness': best_fitness,
|
||||
'training_results': results_file.read_text(),
|
||||
'model': deepcopy(model.module if is_parallel(model) else model).half(),
|
||||
'ema': deepcopy(ema.ema).half(),
|
||||
'updates': ema.updates,
|
||||
'optimizer': optimizer.state_dict(),
|
||||
'wandb_id': wandb_logger.wandb_run.id if wandb_logger.wandb else None}
|
||||
|
||||
# Save last, best and delete
|
||||
torch.save(ckpt, last)
|
||||
if best_fitness == fi:
|
||||
torch.save(ckpt, best)
|
||||
if (best_fitness == fi) and (epoch >= 200):
|
||||
torch.save(ckpt, wdir / 'best_{:03d}.pt'.format(epoch))
|
||||
if epoch == 0:
|
||||
torch.save(ckpt, wdir / 'epoch_{:03d}.pt'.format(epoch))
|
||||
elif ((epoch+1) % 25) == 0:
|
||||
torch.save(ckpt, wdir / 'epoch_{:03d}.pt'.format(epoch))
|
||||
elif epoch >= (epochs-5):
|
||||
torch.save(ckpt, wdir / 'epoch_{:03d}.pt'.format(epoch))
|
||||
if wandb_logger.wandb:
|
||||
if ((epoch + 1) % opt.save_period == 0 and not final_epoch) and opt.save_period != -1:
|
||||
wandb_logger.log_model(
|
||||
last.parent, opt, epoch, fi, best_model=best_fitness == fi)
|
||||
del ckpt
|
||||
torch.cuda.empty_cache()
|
||||
|
||||
# end epoch ----------------------------------------------------------------------------------------------------
|
||||
# end training
|
||||
if rank in [-1, 0]:
|
||||
# Plots
|
||||
if plots:
|
||||
plot_results(save_dir=save_dir) # save as results.png
|
||||
if wandb_logger.wandb:
|
||||
files = ['results.png', 'confusion_matrix.png', *[f'{x}_curve.png' for x in ('F1', 'PR', 'P', 'R')]]
|
||||
wandb_logger.log({"Results": [wandb_logger.wandb.Image(str(save_dir / f), caption=f) for f in files
|
||||
if (save_dir / f).exists()]})
|
||||
# Test best.pt
|
||||
logger.info('%g epochs completed in %.3f hours.\n' % (epoch - start_epoch + 1, (time.time() - t0) / 3600))
|
||||
if opt.data.endswith('coco.yaml') and nc == 80: # if COCO
|
||||
for m in (last, best) if best.exists() else (last): # speed, mAP tests
|
||||
results, _, _ = test.test(opt.data,
|
||||
batch_size=batch_size * 2,
|
||||
imgsz=imgsz_test,
|
||||
conf_thres=0.001,
|
||||
iou_thres=0.7,
|
||||
model=attempt_load(m, device).half(),
|
||||
single_cls=opt.single_cls,
|
||||
dataloader=testloader,
|
||||
save_dir=save_dir,
|
||||
save_json=True,
|
||||
plots=False,
|
||||
is_coco=is_coco,
|
||||
v5_metric=opt.v5_metric)
|
||||
|
||||
# Strip optimizers
|
||||
final = best if best.exists() else last # final model
|
||||
for f in last, best:
|
||||
if f.exists():
|
||||
strip_optimizer(f) # strip optimizers
|
||||
if opt.bucket:
|
||||
os.system(f'gsutil cp {final} gs://{opt.bucket}/weights') # upload
|
||||
if wandb_logger.wandb and not opt.evolve: # Log the stripped model
|
||||
wandb_logger.wandb.log_artifact(str(final), type='model',
|
||||
name='run_' + wandb_logger.wandb_run.id + '_model',
|
||||
aliases=['last', 'best', 'stripped'])
|
||||
wandb_logger.finish_run()
|
||||
else:
|
||||
dist.destroy_process_group()
|
||||
torch.cuda.empty_cache()
|
||||
return results
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--weights', type=str, default='yolo7.pt', help='initial weights path')
|
||||
parser.add_argument('--cfg', type=str, default='', help='model.yaml path')
|
||||
parser.add_argument('--data', type=str, default='data/coco.yaml', help='data.yaml path')
|
||||
parser.add_argument('--hyp', type=str, default='data/hyp.scratch.p5.yaml', help='hyperparameters path')
|
||||
parser.add_argument('--epochs', type=int, default=300)
|
||||
parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs')
|
||||
parser.add_argument('--img-size', nargs='+', type=int, default=[640, 640], help='[train, test] image sizes')
|
||||
parser.add_argument('--rect', action='store_true', help='rectangular training')
|
||||
parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training')
|
||||
parser.add_argument('--nosave', action='store_true', help='only save final checkpoint')
|
||||
parser.add_argument('--notest', action='store_true', help='only test final epoch')
|
||||
parser.add_argument('--noautoanchor', action='store_true', help='disable autoanchor check')
|
||||
parser.add_argument('--evolve', action='store_true', help='evolve hyperparameters')
|
||||
parser.add_argument('--bucket', type=str, default='', help='gsutil bucket')
|
||||
parser.add_argument('--cache-images', action='store_true', help='cache images for faster training')
|
||||
parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training')
|
||||
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
|
||||
parser.add_argument('--multi-scale', action='store_true', help='vary img-size +/- 50%%')
|
||||
parser.add_argument('--single-cls', action='store_true', help='train multi-class data as single-class')
|
||||
parser.add_argument('--adam', action='store_true', help='use torch.optim.Adam() optimizer')
|
||||
parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode')
|
||||
parser.add_argument('--local_rank', type=int, default=-1, help='DDP parameter, do not modify')
|
||||
parser.add_argument('--workers', type=int, default=8, help='maximum number of dataloader workers')
|
||||
parser.add_argument('--project', default='runs/train', help='save to project/name')
|
||||
parser.add_argument('--entity', default=None, help='W&B entity')
|
||||
parser.add_argument('--name', default='exp', help='save to project/name')
|
||||
parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
|
||||
parser.add_argument('--quad', action='store_true', help='quad dataloader')
|
||||
parser.add_argument('--linear-lr', action='store_true', help='linear LR')
|
||||
parser.add_argument('--label-smoothing', type=float, default=0.0, help='Label smoothing epsilon')
|
||||
parser.add_argument('--upload_dataset', action='store_true', help='Upload dataset as W&B artifact table')
|
||||
parser.add_argument('--bbox_interval', type=int, default=-1, help='Set bounding-box image logging interval for W&B')
|
||||
parser.add_argument('--save_period', type=int, default=-1, help='Log model after every "save_period" epoch')
|
||||
parser.add_argument('--artifact_alias', type=str, default="latest", help='version of dataset artifact to be used')
|
||||
parser.add_argument('--freeze', nargs='+', type=int, default=[0], help='Freeze layers: backbone of yolov7=50, first3=0 1 2')
|
||||
parser.add_argument('--v5-metric', action='store_true', help='assume maximum recall as 1.0 in AP calculation')
|
||||
opt = parser.parse_args()
|
||||
|
||||
# Set DDP variables
|
||||
opt.world_size = int(os.environ['WORLD_SIZE']) if 'WORLD_SIZE' in os.environ else 1
|
||||
opt.global_rank = int(os.environ['RANK']) if 'RANK' in os.environ else -1
|
||||
set_logging(opt.global_rank)
|
||||
#if opt.global_rank in [-1, 0]:
|
||||
# check_git_status()
|
||||
# check_requirements()
|
||||
|
||||
# Resume
|
||||
wandb_run = check_wandb_resume(opt)
|
||||
if opt.resume and not wandb_run: # resume an interrupted run
|
||||
ckpt = opt.resume if isinstance(opt.resume, str) else get_latest_run() # specified or most recent path
|
||||
assert os.path.isfile(ckpt), 'ERROR: --resume checkpoint does not exist'
|
||||
apriori = opt.global_rank, opt.local_rank
|
||||
with open(Path(ckpt).parent.parent / 'opt.yaml') as f:
|
||||
opt = argparse.Namespace(**yaml.load(f, Loader=yaml.SafeLoader)) # replace
|
||||
opt.cfg, opt.weights, opt.resume, opt.batch_size, opt.global_rank, opt.local_rank = '', ckpt, True, opt.total_batch_size, *apriori # reinstate
|
||||
logger.info('Resuming training from %s' % ckpt)
|
||||
else:
|
||||
# opt.hyp = opt.hyp or ('hyp.finetune.yaml' if opt.weights else 'hyp.scratch.yaml')
|
||||
opt.data, opt.cfg, opt.hyp = check_file(opt.data), check_file(opt.cfg), check_file(opt.hyp) # check files
|
||||
assert len(opt.cfg) or len(opt.weights), 'either --cfg or --weights must be specified'
|
||||
opt.img_size.extend([opt.img_size[-1]] * (2 - len(opt.img_size))) # extend to 2 sizes (train, test)
|
||||
opt.name = 'evolve' if opt.evolve else opt.name
|
||||
opt.save_dir = increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok | opt.evolve) # increment run
|
||||
|
||||
# DDP mode
|
||||
opt.total_batch_size = opt.batch_size
|
||||
device = select_device(opt.device, batch_size=opt.batch_size)
|
||||
if opt.local_rank != -1:
|
||||
assert torch.cuda.device_count() > opt.local_rank
|
||||
torch.cuda.set_device(opt.local_rank)
|
||||
device = torch.device('cuda', opt.local_rank)
|
||||
dist.init_process_group(backend='nccl', init_method='env://') # distributed backend
|
||||
assert opt.batch_size % opt.world_size == 0, '--batch-size must be multiple of CUDA device count'
|
||||
opt.batch_size = opt.total_batch_size // opt.world_size
|
||||
|
||||
# Hyperparameters
|
||||
with open(opt.hyp) as f:
|
||||
hyp = yaml.load(f, Loader=yaml.SafeLoader) # load hyps
|
||||
|
||||
# Train
|
||||
logger.info(opt)
|
||||
if not opt.evolve:
|
||||
tb_writer = None # init loggers
|
||||
if opt.global_rank in [-1, 0]:
|
||||
prefix = colorstr('tensorboard: ')
|
||||
logger.info(f"{prefix}Start with 'tensorboard --logdir {opt.project}', view at http://localhost:6006/")
|
||||
tb_writer = SummaryWriter(opt.save_dir) # Tensorboard
|
||||
train(hyp, opt, device, tb_writer)
|
||||
|
||||
# Evolve hyperparameters (optional)
|
||||
else:
|
||||
# Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit)
|
||||
meta = {'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3)
|
||||
'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf)
|
||||
'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1
|
||||
'weight_decay': (1, 0.0, 0.001), # optimizer weight decay
|
||||
'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok)
|
||||
'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum
|
||||
'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr
|
||||
'box': (1, 0.02, 0.2), # box loss gain
|
||||
'cls': (1, 0.2, 4.0), # cls loss gain
|
||||
'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight
|
||||
'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels)
|
||||
'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight
|
||||
'iou_t': (0, 0.1, 0.7), # IoU training threshold
|
||||
'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold
|
||||
'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore)
|
||||
'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5)
|
||||
'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction)
|
||||
'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction)
|
||||
'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction)
|
||||
'degrees': (1, 0.0, 45.0), # image rotation (+/- deg)
|
||||
'translate': (1, 0.0, 0.9), # image translation (+/- fraction)
|
||||
'scale': (1, 0.0, 0.9), # image scale (+/- gain)
|
||||
'shear': (1, 0.0, 10.0), # image shear (+/- deg)
|
||||
'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001
|
||||
'flipud': (1, 0.0, 1.0), # image flip up-down (probability)
|
||||
'fliplr': (0, 0.0, 1.0), # image flip left-right (probability)
|
||||
'mosaic': (1, 0.0, 1.0), # image mixup (probability)
|
||||
'mixup': (1, 0.0, 1.0), # image mixup (probability)
|
||||
'copy_paste': (1, 0.0, 1.0), # segment copy-paste (probability)
|
||||
'paste_in': (1, 0.0, 1.0)} # segment copy-paste (probability)
|
||||
|
||||
with open(opt.hyp, errors='ignore') as f:
|
||||
hyp = yaml.safe_load(f) # load hyps dict
|
||||
if 'anchors' not in hyp: # anchors commented in hyp.yaml
|
||||
hyp['anchors'] = 3
|
||||
|
||||
assert opt.local_rank == -1, 'DDP mode not implemented for --evolve'
|
||||
opt.notest, opt.nosave = True, True # only test/save final epoch
|
||||
# ei = [isinstance(x, (int, float)) for x in hyp.values()] # evolvable indices
|
||||
yaml_file = Path(opt.save_dir) / 'hyp_evolved.yaml' # save best result here
|
||||
if opt.bucket:
|
||||
os.system('gsutil cp gs://%s/evolve.txt .' % opt.bucket) # download evolve.txt if exists
|
||||
|
||||
for _ in range(300): # generations to evolve
|
||||
if Path('evolve.txt').exists(): # if evolve.txt exists: select best hyps and mutate
|
||||
# Select parent(s)
|
||||
parent = 'single' # parent selection method: 'single' or 'weighted'
|
||||
x = np.loadtxt('evolve.txt', ndmin=2)
|
||||
n = min(5, len(x)) # number of previous results to consider
|
||||
x = x[np.argsort(-fitness(x))][:n] # top n mutations
|
||||
w = fitness(x) - fitness(x).min() # weights
|
||||
if parent == 'single' or len(x) == 1:
|
||||
# x = x[random.randint(0, n - 1)] # random selection
|
||||
x = x[random.choices(range(n), weights=w)[0]] # weighted selection
|
||||
elif parent == 'weighted':
|
||||
x = (x * w.reshape(n, 1)).sum(0) / w.sum() # weighted combination
|
||||
|
||||
# Mutate
|
||||
mp, s = 0.8, 0.2 # mutation probability, sigma
|
||||
npr = np.random
|
||||
npr.seed(int(time.time()))
|
||||
g = np.array([x[0] for x in meta.values()]) # gains 0-1
|
||||
ng = len(meta)
|
||||
v = np.ones(ng)
|
||||
while all(v == 1): # mutate until a change occurs (prevent duplicates)
|
||||
v = (g * (npr.random(ng) < mp) * npr.randn(ng) * npr.random() * s + 1).clip(0.3, 3.0)
|
||||
for i, k in enumerate(hyp.keys()): # plt.hist(v.ravel(), 300)
|
||||
hyp[k] = float(x[i + 7] * v[i]) # mutate
|
||||
|
||||
# Constrain to limits
|
||||
for k, v in meta.items():
|
||||
hyp[k] = max(hyp[k], v[1]) # lower limit
|
||||
hyp[k] = min(hyp[k], v[2]) # upper limit
|
||||
hyp[k] = round(hyp[k], 5) # significant digits
|
||||
|
||||
# Train mutation
|
||||
results = train(hyp.copy(), opt, device)
|
||||
|
||||
# Write mutation results
|
||||
print_mutation(hyp.copy(), results, yaml_file, opt.bucket)
|
||||
|
||||
# Plot results
|
||||
plot_evolution(yaml_file)
|
||||
print(f'Hyperparameter evolution complete. Best results saved as: {yaml_file}\n'
|
||||
f'Command to train a new model with these hyperparameters: $ python train.py --hyp {yaml_file}')
|
||||
@@ -0,0 +1,699 @@
|
||||
import argparse
|
||||
import logging
|
||||
import math
|
||||
import os
|
||||
import random
|
||||
import time
|
||||
from copy import deepcopy
|
||||
from pathlib import Path
|
||||
from threading import Thread
|
||||
|
||||
import numpy as np
|
||||
import torch.distributed as dist
|
||||
import torch.nn as nn
|
||||
import torch.nn.functional as F
|
||||
import torch.optim as optim
|
||||
import torch.optim.lr_scheduler as lr_scheduler
|
||||
import torch.utils.data
|
||||
import yaml
|
||||
from torch.cuda import amp
|
||||
from torch.nn.parallel import DistributedDataParallel as DDP
|
||||
from torch.utils.tensorboard import SummaryWriter
|
||||
from tqdm import tqdm
|
||||
|
||||
import test # import test.py to get mAP after each epoch
|
||||
from models.experimental import attempt_load
|
||||
from models.yolo import Model
|
||||
from utils.autoanchor import check_anchors
|
||||
from utils.datasets import create_dataloader
|
||||
from utils.general import labels_to_class_weights, increment_path, labels_to_image_weights, init_seeds, \
|
||||
fitness, strip_optimizer, get_latest_run, check_dataset, check_file, check_git_status, check_img_size, \
|
||||
check_requirements, print_mutation, set_logging, one_cycle, colorstr
|
||||
from utils.google_utils import attempt_download
|
||||
from utils.loss import ComputeLoss, ComputeLossAuxOTA
|
||||
from utils.plots import plot_images, plot_labels, plot_results, plot_evolution
|
||||
from utils.torch_utils import ModelEMA, select_device, intersect_dicts, torch_distributed_zero_first, is_parallel
|
||||
from utils.wandb_logging.wandb_utils import WandbLogger, check_wandb_resume
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
def train(hyp, opt, device, tb_writer=None):
|
||||
logger.info(colorstr('hyperparameters: ') + ', '.join(f'{k}={v}' for k, v in hyp.items()))
|
||||
save_dir, epochs, batch_size, total_batch_size, weights, rank = \
|
||||
Path(opt.save_dir), opt.epochs, opt.batch_size, opt.total_batch_size, opt.weights, opt.global_rank
|
||||
|
||||
# Directories
|
||||
wdir = save_dir / 'weights'
|
||||
wdir.mkdir(parents=True, exist_ok=True) # make dir
|
||||
last = wdir / 'last.pt'
|
||||
best = wdir / 'best.pt'
|
||||
results_file = save_dir / 'results.txt'
|
||||
|
||||
# Save run settings
|
||||
with open(save_dir / 'hyp.yaml', 'w') as f:
|
||||
yaml.dump(hyp, f, sort_keys=False)
|
||||
with open(save_dir / 'opt.yaml', 'w') as f:
|
||||
yaml.dump(vars(opt), f, sort_keys=False)
|
||||
|
||||
# Configure
|
||||
plots = not opt.evolve # create plots
|
||||
cuda = device.type != 'cpu'
|
||||
init_seeds(2 + rank)
|
||||
with open(opt.data) as f:
|
||||
data_dict = yaml.load(f, Loader=yaml.SafeLoader) # data dict
|
||||
is_coco = opt.data.endswith('coco.yaml')
|
||||
|
||||
# Logging- Doing this before checking the dataset. Might update data_dict
|
||||
loggers = {'wandb': None} # loggers dict
|
||||
if rank in [-1, 0]:
|
||||
opt.hyp = hyp # add hyperparameters
|
||||
run_id = torch.load(weights).get('wandb_id') if weights.endswith('.pt') and os.path.isfile(weights) else None
|
||||
wandb_logger = WandbLogger(opt, Path(opt.save_dir).stem, run_id, data_dict)
|
||||
loggers['wandb'] = wandb_logger.wandb
|
||||
data_dict = wandb_logger.data_dict
|
||||
if wandb_logger.wandb:
|
||||
weights, epochs, hyp = opt.weights, opt.epochs, opt.hyp # WandbLogger might update weights, epochs if resuming
|
||||
|
||||
nc = 1 if opt.single_cls else int(data_dict['nc']) # number of classes
|
||||
names = ['item'] if opt.single_cls and len(data_dict['names']) != 1 else data_dict['names'] # class names
|
||||
assert len(names) == nc, '%g names found for nc=%g dataset in %s' % (len(names), nc, opt.data) # check
|
||||
|
||||
# Model
|
||||
pretrained = weights.endswith('.pt')
|
||||
if pretrained:
|
||||
with torch_distributed_zero_first(rank):
|
||||
attempt_download(weights) # download if not found locally
|
||||
ckpt = torch.load(weights, map_location=device) # load checkpoint
|
||||
model = Model(opt.cfg or ckpt['model'].yaml, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) # create
|
||||
exclude = ['anchor'] if (opt.cfg or hyp.get('anchors')) and not opt.resume else [] # exclude keys
|
||||
state_dict = ckpt['model'].float().state_dict() # to FP32
|
||||
state_dict = intersect_dicts(state_dict, model.state_dict(), exclude=exclude) # intersect
|
||||
model.load_state_dict(state_dict, strict=False) # load
|
||||
logger.info('Transferred %g/%g items from %s' % (len(state_dict), len(model.state_dict()), weights)) # report
|
||||
else:
|
||||
model = Model(opt.cfg, ch=3, nc=nc, anchors=hyp.get('anchors')).to(device) # create
|
||||
with torch_distributed_zero_first(rank):
|
||||
check_dataset(data_dict) # check
|
||||
train_path = data_dict['train']
|
||||
test_path = data_dict['val']
|
||||
|
||||
# Freeze
|
||||
freeze = [] # parameter names to freeze (full or partial)
|
||||
for k, v in model.named_parameters():
|
||||
v.requires_grad = True # train all layers
|
||||
if any(x in k for x in freeze):
|
||||
print('freezing %s' % k)
|
||||
v.requires_grad = False
|
||||
|
||||
# Optimizer
|
||||
nbs = 64 # nominal batch size
|
||||
accumulate = max(round(nbs / total_batch_size), 1) # accumulate loss before optimizing
|
||||
hyp['weight_decay'] *= total_batch_size * accumulate / nbs # scale weight_decay
|
||||
logger.info(f"Scaled weight_decay = {hyp['weight_decay']}")
|
||||
|
||||
pg0, pg1, pg2 = [], [], [] # optimizer parameter groups
|
||||
for k, v in model.named_modules():
|
||||
if hasattr(v, 'bias') and isinstance(v.bias, nn.Parameter):
|
||||
pg2.append(v.bias) # biases
|
||||
if isinstance(v, nn.BatchNorm2d):
|
||||
pg0.append(v.weight) # no decay
|
||||
elif hasattr(v, 'weight') and isinstance(v.weight, nn.Parameter):
|
||||
pg1.append(v.weight) # apply decay
|
||||
if hasattr(v, 'im'):
|
||||
if hasattr(v.im, 'implicit'):
|
||||
pg0.append(v.im.implicit)
|
||||
else:
|
||||
for iv in v.im:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'imc'):
|
||||
if hasattr(v.imc, 'implicit'):
|
||||
pg0.append(v.imc.implicit)
|
||||
else:
|
||||
for iv in v.imc:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'imb'):
|
||||
if hasattr(v.imb, 'implicit'):
|
||||
pg0.append(v.imb.implicit)
|
||||
else:
|
||||
for iv in v.imb:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'imo'):
|
||||
if hasattr(v.imo, 'implicit'):
|
||||
pg0.append(v.imo.implicit)
|
||||
else:
|
||||
for iv in v.imo:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'ia'):
|
||||
if hasattr(v.ia, 'implicit'):
|
||||
pg0.append(v.ia.implicit)
|
||||
else:
|
||||
for iv in v.ia:
|
||||
pg0.append(iv.implicit)
|
||||
if hasattr(v, 'attn'):
|
||||
if hasattr(v.attn, 'logit_scale'):
|
||||
pg0.append(v.attn.logit_scale)
|
||||
if hasattr(v.attn, 'q_bias'):
|
||||
pg0.append(v.attn.q_bias)
|
||||
if hasattr(v.attn, 'v_bias'):
|
||||
pg0.append(v.attn.v_bias)
|
||||
if hasattr(v.attn, 'relative_position_bias_table'):
|
||||
pg0.append(v.attn.relative_position_bias_table)
|
||||
if hasattr(v, 'rbr_dense'):
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_origin'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_origin)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_avg_conv'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_avg_conv)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_pfir_conv'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_pfir_conv)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_1x1_kxk_idconv1'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_1x1_kxk_idconv1)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_1x1_kxk_conv2'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_1x1_kxk_conv2)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_gconv_dw'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_gconv_dw)
|
||||
if hasattr(v.rbr_dense, 'weight_rbr_gconv_pw'):
|
||||
pg0.append(v.rbr_dense.weight_rbr_gconv_pw)
|
||||
if hasattr(v.rbr_dense, 'vector'):
|
||||
pg0.append(v.rbr_dense.vector)
|
||||
|
||||
if opt.adam:
|
||||
optimizer = optim.Adam(pg0, lr=hyp['lr0'], betas=(hyp['momentum'], 0.999)) # adjust beta1 to momentum
|
||||
else:
|
||||
optimizer = optim.SGD(pg0, lr=hyp['lr0'], momentum=hyp['momentum'], nesterov=True)
|
||||
|
||||
optimizer.add_param_group({'params': pg1, 'weight_decay': hyp['weight_decay']}) # add pg1 with weight_decay
|
||||
optimizer.add_param_group({'params': pg2}) # add pg2 (biases)
|
||||
logger.info('Optimizer groups: %g .bias, %g conv.weight, %g other' % (len(pg2), len(pg1), len(pg0)))
|
||||
del pg0, pg1, pg2
|
||||
|
||||
# Scheduler https://arxiv.org/pdf/1812.01187.pdf
|
||||
# https://pytorch.org/docs/stable/_modules/torch/optim/lr_scheduler.html#OneCycleLR
|
||||
if opt.linear_lr:
|
||||
lf = lambda x: (1 - x / (epochs - 1)) * (1.0 - hyp['lrf']) + hyp['lrf'] # linear
|
||||
else:
|
||||
lf = one_cycle(1, hyp['lrf'], epochs) # cosine 1->hyp['lrf']
|
||||
scheduler = lr_scheduler.LambdaLR(optimizer, lr_lambda=lf)
|
||||
# plot_lr_scheduler(optimizer, scheduler, epochs)
|
||||
|
||||
# EMA
|
||||
ema = ModelEMA(model) if rank in [-1, 0] else None
|
||||
|
||||
# Resume
|
||||
start_epoch, best_fitness = 0, 0.0
|
||||
if pretrained:
|
||||
# Optimizer
|
||||
if ckpt['optimizer'] is not None:
|
||||
optimizer.load_state_dict(ckpt['optimizer'])
|
||||
best_fitness = ckpt['best_fitness']
|
||||
|
||||
# EMA
|
||||
if ema and ckpt.get('ema'):
|
||||
ema.ema.load_state_dict(ckpt['ema'].float().state_dict())
|
||||
ema.updates = ckpt['updates']
|
||||
|
||||
# Results
|
||||
if ckpt.get('training_results') is not None:
|
||||
results_file.write_text(ckpt['training_results']) # write results.txt
|
||||
|
||||
# Epochs
|
||||
start_epoch = ckpt['epoch'] + 1
|
||||
if opt.resume:
|
||||
assert start_epoch > 0, '%s training to %g epochs is finished, nothing to resume.' % (weights, epochs)
|
||||
if epochs < start_epoch:
|
||||
logger.info('%s has been trained for %g epochs. Fine-tuning for %g additional epochs.' %
|
||||
(weights, ckpt['epoch'], epochs))
|
||||
epochs += ckpt['epoch'] # finetune additional epochs
|
||||
|
||||
del ckpt, state_dict
|
||||
|
||||
# Image sizes
|
||||
gs = max(int(model.stride.max()), 32) # grid size (max stride)
|
||||
nl = model.model[-1].nl # number of detection layers (used for scaling hyp['obj'])
|
||||
imgsz, imgsz_test = [check_img_size(x, gs) for x in opt.img_size] # verify imgsz are gs-multiples
|
||||
|
||||
# DP mode
|
||||
if cuda and rank == -1 and torch.cuda.device_count() > 1:
|
||||
model = torch.nn.DataParallel(model)
|
||||
|
||||
# SyncBatchNorm
|
||||
if opt.sync_bn and cuda and rank != -1:
|
||||
model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model).to(device)
|
||||
logger.info('Using SyncBatchNorm()')
|
||||
|
||||
# Trainloader
|
||||
dataloader, dataset = create_dataloader(train_path, imgsz, batch_size, gs, opt,
|
||||
hyp=hyp, augment=True, cache=opt.cache_images, rect=opt.rect, rank=rank,
|
||||
world_size=opt.world_size, workers=opt.workers,
|
||||
image_weights=opt.image_weights, quad=opt.quad, prefix=colorstr('train: '))
|
||||
mlc = np.concatenate(dataset.labels, 0)[:, 0].max() # max label class
|
||||
nb = len(dataloader) # number of batches
|
||||
assert mlc < nc, 'Label class %g exceeds nc=%g in %s. Possible class labels are 0-%g' % (mlc, nc, opt.data, nc - 1)
|
||||
|
||||
# Process 0
|
||||
if rank in [-1, 0]:
|
||||
testloader = create_dataloader(test_path, imgsz_test, batch_size * 2, gs, opt, # testloader
|
||||
hyp=hyp, cache=opt.cache_images and not opt.notest, rect=True, rank=-1,
|
||||
world_size=opt.world_size, workers=opt.workers,
|
||||
pad=0.5, prefix=colorstr('val: '))[0]
|
||||
|
||||
if not opt.resume:
|
||||
labels = np.concatenate(dataset.labels, 0)
|
||||
c = torch.tensor(labels[:, 0]) # classes
|
||||
# cf = torch.bincount(c.long(), minlength=nc) + 1. # frequency
|
||||
# model._initialize_biases(cf.to(device))
|
||||
if plots:
|
||||
#plot_labels(labels, names, save_dir, loggers)
|
||||
if tb_writer:
|
||||
tb_writer.add_histogram('classes', c, 0)
|
||||
|
||||
# Anchors
|
||||
if not opt.noautoanchor:
|
||||
check_anchors(dataset, model=model, thr=hyp['anchor_t'], imgsz=imgsz)
|
||||
model.half().float() # pre-reduce anchor precision
|
||||
|
||||
# DDP mode
|
||||
if cuda and rank != -1:
|
||||
model = DDP(model, device_ids=[opt.local_rank], output_device=opt.local_rank,
|
||||
# nn.MultiheadAttention incompatibility with DDP https://github.com/pytorch/pytorch/issues/26698
|
||||
find_unused_parameters=any(isinstance(layer, nn.MultiheadAttention) for layer in model.modules()))
|
||||
|
||||
# Model parameters
|
||||
hyp['box'] *= 3. / nl # scale to layers
|
||||
hyp['cls'] *= nc / 80. * 3. / nl # scale to classes and layers
|
||||
hyp['obj'] *= (imgsz / 640) ** 2 * 3. / nl # scale to image size and layers
|
||||
hyp['label_smoothing'] = opt.label_smoothing
|
||||
model.nc = nc # attach number of classes to model
|
||||
model.hyp = hyp # attach hyperparameters to model
|
||||
model.gr = 1.0 # iou loss ratio (obj_loss = 1.0 or iou)
|
||||
model.class_weights = labels_to_class_weights(dataset.labels, nc).to(device) * nc # attach class weights
|
||||
model.names = names
|
||||
|
||||
# Start training
|
||||
t0 = time.time()
|
||||
nw = max(round(hyp['warmup_epochs'] * nb), 1000) # number of warmup iterations, max(3 epochs, 1k iterations)
|
||||
# nw = min(nw, (epochs - start_epoch) / 2 * nb) # limit warmup to < 1/2 of training
|
||||
maps = np.zeros(nc) # mAP per class
|
||||
results = (0, 0, 0, 0, 0, 0, 0) # P, R, mAP@.5, mAP@.5-.95, val_loss(box, obj, cls)
|
||||
scheduler.last_epoch = start_epoch - 1 # do not move
|
||||
scaler = amp.GradScaler(enabled=cuda)
|
||||
compute_loss_ota = ComputeLossAuxOTA(model) # init loss class
|
||||
compute_loss = ComputeLoss(model) # init loss class
|
||||
logger.info(f'Image sizes {imgsz} train, {imgsz_test} test\n'
|
||||
f'Using {dataloader.num_workers} dataloader workers\n'
|
||||
f'Logging results to {save_dir}\n'
|
||||
f'Starting training for {epochs} epochs...')
|
||||
torch.save(model, wdir / 'init.pt')
|
||||
for epoch in range(start_epoch, epochs): # epoch ------------------------------------------------------------------
|
||||
model.train()
|
||||
|
||||
# Update image weights (optional)
|
||||
if opt.image_weights:
|
||||
# Generate indices
|
||||
if rank in [-1, 0]:
|
||||
cw = model.class_weights.cpu().numpy() * (1 - maps) ** 2 / nc # class weights
|
||||
iw = labels_to_image_weights(dataset.labels, nc=nc, class_weights=cw) # image weights
|
||||
dataset.indices = random.choices(range(dataset.n), weights=iw, k=dataset.n) # rand weighted idx
|
||||
# Broadcast if DDP
|
||||
if rank != -1:
|
||||
indices = (torch.tensor(dataset.indices) if rank == 0 else torch.zeros(dataset.n)).int()
|
||||
dist.broadcast(indices, 0)
|
||||
if rank != 0:
|
||||
dataset.indices = indices.cpu().numpy()
|
||||
|
||||
# Update mosaic border
|
||||
# b = int(random.uniform(0.25 * imgsz, 0.75 * imgsz + gs) // gs * gs)
|
||||
# dataset.mosaic_border = [b - imgsz, -b] # height, width borders
|
||||
|
||||
mloss = torch.zeros(4, device=device) # mean losses
|
||||
if rank != -1:
|
||||
dataloader.sampler.set_epoch(epoch)
|
||||
pbar = enumerate(dataloader)
|
||||
logger.info(('\n' + '%10s' * 8) % ('Epoch', 'gpu_mem', 'box', 'obj', 'cls', 'total', 'labels', 'img_size'))
|
||||
if rank in [-1, 0]:
|
||||
pbar = tqdm(pbar, total=nb) # progress bar
|
||||
optimizer.zero_grad()
|
||||
for i, (imgs, targets, paths, _) in pbar: # batch -------------------------------------------------------------
|
||||
ni = i + nb * epoch # number integrated batches (since train start)
|
||||
imgs = imgs.to(device, non_blocking=True).float() / 255.0 # uint8 to float32, 0-255 to 0.0-1.0
|
||||
|
||||
# Warmup
|
||||
if ni <= nw:
|
||||
xi = [0, nw] # x interp
|
||||
# model.gr = np.interp(ni, xi, [0.0, 1.0]) # iou loss ratio (obj_loss = 1.0 or iou)
|
||||
accumulate = max(1, np.interp(ni, xi, [1, nbs / total_batch_size]).round())
|
||||
for j, x in enumerate(optimizer.param_groups):
|
||||
# bias lr falls from 0.1 to lr0, all other lrs rise from 0.0 to lr0
|
||||
x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 2 else 0.0, x['initial_lr'] * lf(epoch)])
|
||||
if 'momentum' in x:
|
||||
x['momentum'] = np.interp(ni, xi, [hyp['warmup_momentum'], hyp['momentum']])
|
||||
|
||||
# Multi-scale
|
||||
if opt.multi_scale:
|
||||
sz = random.randrange(imgsz * 0.5, imgsz * 1.5 + gs) // gs * gs # size
|
||||
sf = sz / max(imgs.shape[2:]) # scale factor
|
||||
if sf != 1:
|
||||
ns = [math.ceil(x * sf / gs) * gs for x in imgs.shape[2:]] # new shape (stretched to gs-multiple)
|
||||
imgs = F.interpolate(imgs, size=ns, mode='bilinear', align_corners=False)
|
||||
|
||||
# Forward
|
||||
with amp.autocast(enabled=cuda):
|
||||
pred = model(imgs) # forward
|
||||
loss, loss_items = compute_loss_ota(pred, targets.to(device), imgs) # loss scaled by batch_size
|
||||
if rank != -1:
|
||||
loss *= opt.world_size # gradient averaged between devices in DDP mode
|
||||
if opt.quad:
|
||||
loss *= 4.
|
||||
|
||||
# Backward
|
||||
scaler.scale(loss).backward()
|
||||
|
||||
# Optimize
|
||||
if ni % accumulate == 0:
|
||||
scaler.step(optimizer) # optimizer.step
|
||||
scaler.update()
|
||||
optimizer.zero_grad()
|
||||
if ema:
|
||||
ema.update(model)
|
||||
|
||||
# Print
|
||||
if rank in [-1, 0]:
|
||||
mloss = (mloss * i + loss_items) / (i + 1) # update mean losses
|
||||
mem = '%.3gG' % (torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0) # (GB)
|
||||
s = ('%10s' * 2 + '%10.4g' * 6) % (
|
||||
'%g/%g' % (epoch, epochs - 1), mem, *mloss, targets.shape[0], imgs.shape[-1])
|
||||
pbar.set_description(s)
|
||||
|
||||
# Plot
|
||||
if plots and ni < 10:
|
||||
f = save_dir / f'train_batch{ni}.jpg' # filename
|
||||
Thread(target=plot_images, args=(imgs, targets, paths, f), daemon=True).start()
|
||||
# if tb_writer:
|
||||
# tb_writer.add_image(f, result, dataformats='HWC', global_step=epoch)
|
||||
# tb_writer.add_graph(torch.jit.trace(model, imgs, strict=False), []) # add model graph
|
||||
elif plots and ni == 10 and wandb_logger.wandb:
|
||||
wandb_logger.log({"Mosaics": [wandb_logger.wandb.Image(str(x), caption=x.name) for x in
|
||||
save_dir.glob('train*.jpg') if x.exists()]})
|
||||
|
||||
# end batch ------------------------------------------------------------------------------------------------
|
||||
# end epoch ----------------------------------------------------------------------------------------------------
|
||||
|
||||
# Scheduler
|
||||
lr = [x['lr'] for x in optimizer.param_groups] # for tensorboard
|
||||
scheduler.step()
|
||||
|
||||
# DDP process 0 or single-GPU
|
||||
if rank in [-1, 0]:
|
||||
# mAP
|
||||
ema.update_attr(model, include=['yaml', 'nc', 'hyp', 'gr', 'names', 'stride', 'class_weights'])
|
||||
final_epoch = epoch + 1 == epochs
|
||||
if not opt.notest or final_epoch: # Calculate mAP
|
||||
wandb_logger.current_epoch = epoch + 1
|
||||
results, maps, times = test.test(data_dict,
|
||||
batch_size=batch_size * 2,
|
||||
imgsz=imgsz_test,
|
||||
model=ema.ema,
|
||||
single_cls=opt.single_cls,
|
||||
dataloader=testloader,
|
||||
save_dir=save_dir,
|
||||
verbose=nc < 50 and final_epoch,
|
||||
plots=plots and final_epoch,
|
||||
wandb_logger=wandb_logger,
|
||||
compute_loss=compute_loss,
|
||||
is_coco=is_coco,
|
||||
v5_metric=opt.v5_metric)
|
||||
|
||||
# Write
|
||||
with open(results_file, 'a') as f:
|
||||
f.write(s + '%10.4g' * 7 % results + '\n') # append metrics, val_loss
|
||||
if len(opt.name) and opt.bucket:
|
||||
os.system('gsutil cp %s gs://%s/results/results%s.txt' % (results_file, opt.bucket, opt.name))
|
||||
|
||||
# Log
|
||||
tags = ['train/box_loss', 'train/obj_loss', 'train/cls_loss', # train loss
|
||||
'metrics/precision', 'metrics/recall', 'metrics/mAP_0.5', 'metrics/mAP_0.5:0.95',
|
||||
'val/box_loss', 'val/obj_loss', 'val/cls_loss', # val loss
|
||||
'x/lr0', 'x/lr1', 'x/lr2'] # params
|
||||
for x, tag in zip(list(mloss[:-1]) + list(results) + lr, tags):
|
||||
if tb_writer:
|
||||
tb_writer.add_scalar(tag, x, epoch) # tensorboard
|
||||
if wandb_logger.wandb:
|
||||
wandb_logger.log({tag: x}) # W&B
|
||||
|
||||
# Update best mAP
|
||||
fi = fitness(np.array(results).reshape(1, -1)) # weighted combination of [P, R, mAP@.5, mAP@.5-.95]
|
||||
if fi > best_fitness:
|
||||
best_fitness = fi
|
||||
wandb_logger.end_epoch(best_result=best_fitness == fi)
|
||||
|
||||
# Save model
|
||||
if (not opt.nosave) or (final_epoch and not opt.evolve): # if save
|
||||
ckpt = {'epoch': epoch,
|
||||
'best_fitness': best_fitness,
|
||||
'training_results': results_file.read_text(),
|
||||
'model': deepcopy(model.module if is_parallel(model) else model).half(),
|
||||
'ema': deepcopy(ema.ema).half(),
|
||||
'updates': ema.updates,
|
||||
'optimizer': optimizer.state_dict(),
|
||||
'wandb_id': wandb_logger.wandb_run.id if wandb_logger.wandb else None}
|
||||
|
||||
# Save last, best and delete
|
||||
torch.save(ckpt, last)
|
||||
if best_fitness == fi:
|
||||
torch.save(ckpt, best)
|
||||
if (best_fitness == fi) and (epoch >= 200):
|
||||
torch.save(ckpt, wdir / 'best_{:03d}.pt'.format(epoch))
|
||||
if epoch == 0:
|
||||
torch.save(ckpt, wdir / 'epoch_{:03d}.pt'.format(epoch))
|
||||
elif ((epoch+1) % 25) == 0:
|
||||
torch.save(ckpt, wdir / 'epoch_{:03d}.pt'.format(epoch))
|
||||
elif epoch >= (epochs-5):
|
||||
torch.save(ckpt, wdir / 'epoch_{:03d}.pt'.format(epoch))
|
||||
if wandb_logger.wandb:
|
||||
if ((epoch + 1) % opt.save_period == 0 and not final_epoch) and opt.save_period != -1:
|
||||
wandb_logger.log_model(
|
||||
last.parent, opt, epoch, fi, best_model=best_fitness == fi)
|
||||
del ckpt
|
||||
|
||||
# end epoch ----------------------------------------------------------------------------------------------------
|
||||
# end training
|
||||
if rank in [-1, 0]:
|
||||
# Plots
|
||||
if plots:
|
||||
plot_results(save_dir=save_dir) # save as results.png
|
||||
if wandb_logger.wandb:
|
||||
files = ['results.png', 'confusion_matrix.png', *[f'{x}_curve.png' for x in ('F1', 'PR', 'P', 'R')]]
|
||||
wandb_logger.log({"Results": [wandb_logger.wandb.Image(str(save_dir / f), caption=f) for f in files
|
||||
if (save_dir / f).exists()]})
|
||||
# Test best.pt
|
||||
logger.info('%g epochs completed in %.3f hours.\n' % (epoch - start_epoch + 1, (time.time() - t0) / 3600))
|
||||
if opt.data.endswith('coco.yaml') and nc == 80: # if COCO
|
||||
for m in (last, best) if best.exists() else (last): # speed, mAP tests
|
||||
results, _, _ = test.test(opt.data,
|
||||
batch_size=batch_size * 2,
|
||||
imgsz=imgsz_test,
|
||||
conf_thres=0.001,
|
||||
iou_thres=0.7,
|
||||
model=attempt_load(m, device).half(),
|
||||
single_cls=opt.single_cls,
|
||||
dataloader=testloader,
|
||||
save_dir=save_dir,
|
||||
save_json=True,
|
||||
plots=False,
|
||||
is_coco=is_coco,
|
||||
v5_metric=opt.v5_metric)
|
||||
|
||||
# Strip optimizers
|
||||
final = best if best.exists() else last # final model
|
||||
for f in last, best:
|
||||
if f.exists():
|
||||
strip_optimizer(f) # strip optimizers
|
||||
if opt.bucket:
|
||||
os.system(f'gsutil cp {final} gs://{opt.bucket}/weights') # upload
|
||||
if wandb_logger.wandb and not opt.evolve: # Log the stripped model
|
||||
wandb_logger.wandb.log_artifact(str(final), type='model',
|
||||
name='run_' + wandb_logger.wandb_run.id + '_model',
|
||||
aliases=['last', 'best', 'stripped'])
|
||||
wandb_logger.finish_run()
|
||||
else:
|
||||
dist.destroy_process_group()
|
||||
torch.cuda.empty_cache()
|
||||
return results
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--weights', type=str, default='yolo7.pt', help='initial weights path')
|
||||
parser.add_argument('--cfg', type=str, default='', help='model.yaml path')
|
||||
parser.add_argument('--data', type=str, default='data/coco.yaml', help='data.yaml path')
|
||||
parser.add_argument('--hyp', type=str, default='data/hyp.scratch.p5.yaml', help='hyperparameters path')
|
||||
parser.add_argument('--epochs', type=int, default=300)
|
||||
parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs')
|
||||
parser.add_argument('--img-size', nargs='+', type=int, default=[640, 640], help='[train, test] image sizes')
|
||||
parser.add_argument('--rect', action='store_true', help='rectangular training')
|
||||
parser.add_argument('--resume', nargs='?', const=True, default=False, help='resume most recent training')
|
||||
parser.add_argument('--nosave', action='store_true', help='only save final checkpoint')
|
||||
parser.add_argument('--notest', action='store_true', help='only test final epoch')
|
||||
parser.add_argument('--noautoanchor', action='store_true', help='disable autoanchor check')
|
||||
parser.add_argument('--evolve', action='store_true', help='evolve hyperparameters')
|
||||
parser.add_argument('--bucket', type=str, default='', help='gsutil bucket')
|
||||
parser.add_argument('--cache-images', action='store_true', help='cache images for faster training')
|
||||
parser.add_argument('--image-weights', action='store_true', help='use weighted image selection for training')
|
||||
parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
|
||||
parser.add_argument('--multi-scale', action='store_true', help='vary img-size +/- 50%%')
|
||||
parser.add_argument('--single-cls', action='store_true', help='train multi-class data as single-class')
|
||||
parser.add_argument('--adam', action='store_true', help='use torch.optim.Adam() optimizer')
|
||||
parser.add_argument('--sync-bn', action='store_true', help='use SyncBatchNorm, only available in DDP mode')
|
||||
parser.add_argument('--local_rank', type=int, default=-1, help='DDP parameter, do not modify')
|
||||
parser.add_argument('--workers', type=int, default=8, help='maximum number of dataloader workers')
|
||||
parser.add_argument('--project', default='runs/train', help='save to project/name')
|
||||
parser.add_argument('--entity', default=None, help='W&B entity')
|
||||
parser.add_argument('--name', default='exp', help='save to project/name')
|
||||
parser.add_argument('--exist-ok', action='store_true', help='existing project/name ok, do not increment')
|
||||
parser.add_argument('--quad', action='store_true', help='quad dataloader')
|
||||
parser.add_argument('--linear-lr', action='store_true', help='linear LR')
|
||||
parser.add_argument('--label-smoothing', type=float, default=0.0, help='Label smoothing epsilon')
|
||||
parser.add_argument('--upload_dataset', action='store_true', help='Upload dataset as W&B artifact table')
|
||||
parser.add_argument('--bbox_interval', type=int, default=-1, help='Set bounding-box image logging interval for W&B')
|
||||
parser.add_argument('--save_period', type=int, default=-1, help='Log model after every "save_period" epoch')
|
||||
parser.add_argument('--artifact_alias', type=str, default="latest", help='version of dataset artifact to be used')
|
||||
parser.add_argument('--v5-metric', action='store_true', help='assume maximum recall as 1.0 in AP calculation')
|
||||
opt = parser.parse_args()
|
||||
|
||||
# Set DDP variables
|
||||
opt.world_size = int(os.environ['WORLD_SIZE']) if 'WORLD_SIZE' in os.environ else 1
|
||||
opt.global_rank = int(os.environ['RANK']) if 'RANK' in os.environ else -1
|
||||
set_logging(opt.global_rank)
|
||||
#if opt.global_rank in [-1, 0]:
|
||||
# check_git_status()
|
||||
# check_requirements()
|
||||
|
||||
# Resume
|
||||
wandb_run = check_wandb_resume(opt)
|
||||
if opt.resume and not wandb_run: # resume an interrupted run
|
||||
ckpt = opt.resume if isinstance(opt.resume, str) else get_latest_run() # specified or most recent path
|
||||
assert os.path.isfile(ckpt), 'ERROR: --resume checkpoint does not exist'
|
||||
apriori = opt.global_rank, opt.local_rank
|
||||
with open(Path(ckpt).parent.parent / 'opt.yaml') as f:
|
||||
opt = argparse.Namespace(**yaml.load(f, Loader=yaml.SafeLoader)) # replace
|
||||
opt.cfg, opt.weights, opt.resume, opt.batch_size, opt.global_rank, opt.local_rank = '', ckpt, True, opt.total_batch_size, *apriori # reinstate
|
||||
logger.info('Resuming training from %s' % ckpt)
|
||||
else:
|
||||
# opt.hyp = opt.hyp or ('hyp.finetune.yaml' if opt.weights else 'hyp.scratch.yaml')
|
||||
opt.data, opt.cfg, opt.hyp = check_file(opt.data), check_file(opt.cfg), check_file(opt.hyp) # check files
|
||||
assert len(opt.cfg) or len(opt.weights), 'either --cfg or --weights must be specified'
|
||||
opt.img_size.extend([opt.img_size[-1]] * (2 - len(opt.img_size))) # extend to 2 sizes (train, test)
|
||||
opt.name = 'evolve' if opt.evolve else opt.name
|
||||
opt.save_dir = increment_path(Path(opt.project) / opt.name, exist_ok=opt.exist_ok | opt.evolve) # increment run
|
||||
|
||||
# DDP mode
|
||||
opt.total_batch_size = opt.batch_size
|
||||
device = select_device(opt.device, batch_size=opt.batch_size)
|
||||
if opt.local_rank != -1:
|
||||
assert torch.cuda.device_count() > opt.local_rank
|
||||
torch.cuda.set_device(opt.local_rank)
|
||||
device = torch.device('cuda', opt.local_rank)
|
||||
dist.init_process_group(backend='nccl', init_method='env://') # distributed backend
|
||||
assert opt.batch_size % opt.world_size == 0, '--batch-size must be multiple of CUDA device count'
|
||||
opt.batch_size = opt.total_batch_size // opt.world_size
|
||||
|
||||
# Hyperparameters
|
||||
with open(opt.hyp) as f:
|
||||
hyp = yaml.load(f, Loader=yaml.SafeLoader) # load hyps
|
||||
|
||||
# Train
|
||||
logger.info(opt)
|
||||
if not opt.evolve:
|
||||
tb_writer = None # init loggers
|
||||
if opt.global_rank in [-1, 0]:
|
||||
prefix = colorstr('tensorboard: ')
|
||||
logger.info(f"{prefix}Start with 'tensorboard --logdir {opt.project}', view at http://localhost:6006/")
|
||||
tb_writer = SummaryWriter(opt.save_dir) # Tensorboard
|
||||
train(hyp, opt, device, tb_writer)
|
||||
|
||||
# Evolve hyperparameters (optional)
|
||||
else:
|
||||
# Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit)
|
||||
meta = {'lr0': (1, 1e-5, 1e-1), # initial learning rate (SGD=1E-2, Adam=1E-3)
|
||||
'lrf': (1, 0.01, 1.0), # final OneCycleLR learning rate (lr0 * lrf)
|
||||
'momentum': (0.3, 0.6, 0.98), # SGD momentum/Adam beta1
|
||||
'weight_decay': (1, 0.0, 0.001), # optimizer weight decay
|
||||
'warmup_epochs': (1, 0.0, 5.0), # warmup epochs (fractions ok)
|
||||
'warmup_momentum': (1, 0.0, 0.95), # warmup initial momentum
|
||||
'warmup_bias_lr': (1, 0.0, 0.2), # warmup initial bias lr
|
||||
'box': (1, 0.02, 0.2), # box loss gain
|
||||
'cls': (1, 0.2, 4.0), # cls loss gain
|
||||
'cls_pw': (1, 0.5, 2.0), # cls BCELoss positive_weight
|
||||
'obj': (1, 0.2, 4.0), # obj loss gain (scale with pixels)
|
||||
'obj_pw': (1, 0.5, 2.0), # obj BCELoss positive_weight
|
||||
'iou_t': (0, 0.1, 0.7), # IoU training threshold
|
||||
'anchor_t': (1, 2.0, 8.0), # anchor-multiple threshold
|
||||
'anchors': (2, 2.0, 10.0), # anchors per output grid (0 to ignore)
|
||||
'fl_gamma': (0, 0.0, 2.0), # focal loss gamma (efficientDet default gamma=1.5)
|
||||
'hsv_h': (1, 0.0, 0.1), # image HSV-Hue augmentation (fraction)
|
||||
'hsv_s': (1, 0.0, 0.9), # image HSV-Saturation augmentation (fraction)
|
||||
'hsv_v': (1, 0.0, 0.9), # image HSV-Value augmentation (fraction)
|
||||
'degrees': (1, 0.0, 45.0), # image rotation (+/- deg)
|
||||
'translate': (1, 0.0, 0.9), # image translation (+/- fraction)
|
||||
'scale': (1, 0.0, 0.9), # image scale (+/- gain)
|
||||
'shear': (1, 0.0, 10.0), # image shear (+/- deg)
|
||||
'perspective': (0, 0.0, 0.001), # image perspective (+/- fraction), range 0-0.001
|
||||
'flipud': (1, 0.0, 1.0), # image flip up-down (probability)
|
||||
'fliplr': (0, 0.0, 1.0), # image flip left-right (probability)
|
||||
'mosaic': (1, 0.0, 1.0), # image mixup (probability)
|
||||
'mixup': (1, 0.0, 1.0)} # image mixup (probability)
|
||||
|
||||
with open(opt.hyp, errors='ignore') as f:
|
||||
hyp = yaml.safe_load(f) # load hyps dict
|
||||
if 'anchors' not in hyp: # anchors commented in hyp.yaml
|
||||
hyp['anchors'] = 3
|
||||
|
||||
assert opt.local_rank == -1, 'DDP mode not implemented for --evolve'
|
||||
opt.notest, opt.nosave = True, True # only test/save final epoch
|
||||
# ei = [isinstance(x, (int, float)) for x in hyp.values()] # evolvable indices
|
||||
yaml_file = Path(opt.save_dir) / 'hyp_evolved.yaml' # save best result here
|
||||
if opt.bucket:
|
||||
os.system('gsutil cp gs://%s/evolve.txt .' % opt.bucket) # download evolve.txt if exists
|
||||
|
||||
for _ in range(300): # generations to evolve
|
||||
if Path('evolve.txt').exists(): # if evolve.txt exists: select best hyps and mutate
|
||||
# Select parent(s)
|
||||
parent = 'single' # parent selection method: 'single' or 'weighted'
|
||||
x = np.loadtxt('evolve.txt', ndmin=2)
|
||||
n = min(5, len(x)) # number of previous results to consider
|
||||
x = x[np.argsort(-fitness(x))][:n] # top n mutations
|
||||
w = fitness(x) - fitness(x).min() # weights
|
||||
if parent == 'single' or len(x) == 1:
|
||||
# x = x[random.randint(0, n - 1)] # random selection
|
||||
x = x[random.choices(range(n), weights=w)[0]] # weighted selection
|
||||
elif parent == 'weighted':
|
||||
x = (x * w.reshape(n, 1)).sum(0) / w.sum() # weighted combination
|
||||
|
||||
# Mutate
|
||||
mp, s = 0.8, 0.2 # mutation probability, sigma
|
||||
npr = np.random
|
||||
npr.seed(int(time.time()))
|
||||
g = np.array([x[0] for x in meta.values()]) # gains 0-1
|
||||
ng = len(meta)
|
||||
v = np.ones(ng)
|
||||
while all(v == 1): # mutate until a change occurs (prevent duplicates)
|
||||
v = (g * (npr.random(ng) < mp) * npr.randn(ng) * npr.random() * s + 1).clip(0.3, 3.0)
|
||||
for i, k in enumerate(hyp.keys()): # plt.hist(v.ravel(), 300)
|
||||
hyp[k] = float(x[i + 7] * v[i]) # mutate
|
||||
|
||||
# Constrain to limits
|
||||
for k, v in meta.items():
|
||||
hyp[k] = max(hyp[k], v[1]) # lower limit
|
||||
hyp[k] = min(hyp[k], v[2]) # upper limit
|
||||
hyp[k] = round(hyp[k], 5) # significant digits
|
||||
|
||||
# Train mutation
|
||||
results = train(hyp.copy(), opt, device)
|
||||
|
||||
# Write mutation results
|
||||
print_mutation(hyp.copy(), results, yaml_file, opt.bucket)
|
||||
|
||||
# Plot results
|
||||
plot_evolution(yaml_file)
|
||||
print(f'Hyperparameter evolution complete. Best results saved as: {yaml_file}\n'
|
||||
f'Command to train a new model with these hyperparameters: $ python train.py --hyp {yaml_file}')
|
||||
@@ -0,0 +1 @@
|
||||
# init
|
||||
@@ -0,0 +1,72 @@
|
||||
# Activation functions
|
||||
|
||||
import torch
|
||||
import torch.nn as nn
|
||||
import torch.nn.functional as F
|
||||
|
||||
|
||||
# SiLU https://arxiv.org/pdf/1606.08415.pdf ----------------------------------------------------------------------------
|
||||
class SiLU(nn.Module): # export-friendly version of nn.SiLU()
|
||||
@staticmethod
|
||||
def forward(x):
|
||||
return x * torch.sigmoid(x)
|
||||
|
||||
|
||||
class Hardswish(nn.Module): # export-friendly version of nn.Hardswish()
|
||||
@staticmethod
|
||||
def forward(x):
|
||||
# return x * F.hardsigmoid(x) # for torchscript and CoreML
|
||||
return x * F.hardtanh(x + 3, 0., 6.) / 6. # for torchscript, CoreML and ONNX
|
||||
|
||||
|
||||
class MemoryEfficientSwish(nn.Module):
|
||||
class F(torch.autograd.Function):
|
||||
@staticmethod
|
||||
def forward(ctx, x):
|
||||
ctx.save_for_backward(x)
|
||||
return x * torch.sigmoid(x)
|
||||
|
||||
@staticmethod
|
||||
def backward(ctx, grad_output):
|
||||
x = ctx.saved_tensors[0]
|
||||
sx = torch.sigmoid(x)
|
||||
return grad_output * (sx * (1 + x * (1 - sx)))
|
||||
|
||||
def forward(self, x):
|
||||
return self.F.apply(x)
|
||||
|
||||
|
||||
# Mish https://github.com/digantamisra98/Mish --------------------------------------------------------------------------
|
||||
class Mish(nn.Module):
|
||||
@staticmethod
|
||||
def forward(x):
|
||||
return x * F.softplus(x).tanh()
|
||||
|
||||
|
||||
class MemoryEfficientMish(nn.Module):
|
||||
class F(torch.autograd.Function):
|
||||
@staticmethod
|
||||
def forward(ctx, x):
|
||||
ctx.save_for_backward(x)
|
||||
return x.mul(torch.tanh(F.softplus(x))) # x * tanh(ln(1 + exp(x)))
|
||||
|
||||
@staticmethod
|
||||
def backward(ctx, grad_output):
|
||||
x = ctx.saved_tensors[0]
|
||||
sx = torch.sigmoid(x)
|
||||
fx = F.softplus(x).tanh()
|
||||
return grad_output * (fx + x * sx * (1 - fx * fx))
|
||||
|
||||
def forward(self, x):
|
||||
return self.F.apply(x)
|
||||
|
||||
|
||||
# FReLU https://arxiv.org/abs/2007.11824 -------------------------------------------------------------------------------
|
||||
class FReLU(nn.Module):
|
||||
def __init__(self, c1, k=3): # ch_in, kernel
|
||||
super().__init__()
|
||||
self.conv = nn.Conv2d(c1, c1, k, 1, 1, groups=c1, bias=False)
|
||||
self.bn = nn.BatchNorm2d(c1)
|
||||
|
||||
def forward(self, x):
|
||||
return torch.max(x, self.bn(self.conv(x)))
|
||||
@@ -0,0 +1,155 @@
|
||||
import numpy as np
|
||||
import onnx
|
||||
from onnx import shape_inference
|
||||
try:
|
||||
import onnx_graphsurgeon as gs
|
||||
except Exception as e:
|
||||
print('Import onnx_graphsurgeon failure: %s' % e)
|
||||
|
||||
import logging
|
||||
|
||||
LOGGER = logging.getLogger(__name__)
|
||||
|
||||
class RegisterNMS(object):
|
||||
def __init__(
|
||||
self,
|
||||
onnx_model_path: str,
|
||||
precision: str = "fp32",
|
||||
):
|
||||
|
||||
self.graph = gs.import_onnx(onnx.load(onnx_model_path))
|
||||
assert self.graph
|
||||
LOGGER.info("ONNX graph created successfully")
|
||||
# Fold constants via ONNX-GS that PyTorch2ONNX may have missed
|
||||
self.graph.fold_constants()
|
||||
self.precision = precision
|
||||
self.batch_size = 1
|
||||
def infer(self):
|
||||
"""
|
||||
Sanitize the graph by cleaning any unconnected nodes, do a topological resort,
|
||||
and fold constant inputs values. When possible, run shape inference on the
|
||||
ONNX graph to determine tensor shapes.
|
||||
"""
|
||||
for _ in range(3):
|
||||
count_before = len(self.graph.nodes)
|
||||
|
||||
self.graph.cleanup().toposort()
|
||||
try:
|
||||
for node in self.graph.nodes:
|
||||
for o in node.outputs:
|
||||
o.shape = None
|
||||
model = gs.export_onnx(self.graph)
|
||||
model = shape_inference.infer_shapes(model)
|
||||
self.graph = gs.import_onnx(model)
|
||||
except Exception as e:
|
||||
LOGGER.info(f"Shape inference could not be performed at this time:\n{e}")
|
||||
try:
|
||||
self.graph.fold_constants(fold_shapes=True)
|
||||
except TypeError as e:
|
||||
LOGGER.error(
|
||||
"This version of ONNX GraphSurgeon does not support folding shapes, "
|
||||
f"please upgrade your onnx_graphsurgeon module. Error:\n{e}"
|
||||
)
|
||||
raise
|
||||
|
||||
count_after = len(self.graph.nodes)
|
||||
if count_before == count_after:
|
||||
# No new folding occurred in this iteration, so we can stop for now.
|
||||
break
|
||||
|
||||
def save(self, output_path):
|
||||
"""
|
||||
Save the ONNX model to the given location.
|
||||
Args:
|
||||
output_path: Path pointing to the location where to write
|
||||
out the updated ONNX model.
|
||||
"""
|
||||
self.graph.cleanup().toposort()
|
||||
model = gs.export_onnx(self.graph)
|
||||
onnx.save(model, output_path)
|
||||
LOGGER.info(f"Saved ONNX model to {output_path}")
|
||||
|
||||
def register_nms(
|
||||
self,
|
||||
*,
|
||||
score_thresh: float = 0.25,
|
||||
nms_thresh: float = 0.45,
|
||||
detections_per_img: int = 100,
|
||||
):
|
||||
"""
|
||||
Register the ``EfficientNMS_TRT`` plugin node.
|
||||
NMS expects these shapes for its input tensors:
|
||||
- box_net: [batch_size, number_boxes, 4]
|
||||
- class_net: [batch_size, number_boxes, number_labels]
|
||||
Args:
|
||||
score_thresh (float): The scalar threshold for score (low scoring boxes are removed).
|
||||
nms_thresh (float): The scalar threshold for IOU (new boxes that have high IOU
|
||||
overlap with previously selected boxes are removed).
|
||||
detections_per_img (int): Number of best detections to keep after NMS.
|
||||
"""
|
||||
|
||||
self.infer()
|
||||
# Find the concat node at the end of the network
|
||||
op_inputs = self.graph.outputs
|
||||
op = "EfficientNMS_TRT"
|
||||
attrs = {
|
||||
"plugin_version": "1",
|
||||
"background_class": -1, # no background class
|
||||
"max_output_boxes": detections_per_img,
|
||||
"score_threshold": score_thresh,
|
||||
"iou_threshold": nms_thresh,
|
||||
"score_activation": False,
|
||||
"box_coding": 0,
|
||||
}
|
||||
|
||||
if self.precision == "fp32":
|
||||
dtype_output = np.float32
|
||||
elif self.precision == "fp16":
|
||||
dtype_output = np.float16
|
||||
else:
|
||||
raise NotImplementedError(f"Currently not supports precision: {self.precision}")
|
||||
|
||||
# NMS Outputs
|
||||
output_num_detections = gs.Variable(
|
||||
name="num_dets",
|
||||
dtype=np.int32,
|
||||
shape=[self.batch_size, 1],
|
||||
) # A scalar indicating the number of valid detections per batch image.
|
||||
output_boxes = gs.Variable(
|
||||
name="det_boxes",
|
||||
dtype=dtype_output,
|
||||
shape=[self.batch_size, detections_per_img, 4],
|
||||
)
|
||||
output_scores = gs.Variable(
|
||||
name="det_scores",
|
||||
dtype=dtype_output,
|
||||
shape=[self.batch_size, detections_per_img],
|
||||
)
|
||||
output_labels = gs.Variable(
|
||||
name="det_classes",
|
||||
dtype=np.int32,
|
||||
shape=[self.batch_size, detections_per_img],
|
||||
)
|
||||
|
||||
op_outputs = [output_num_detections, output_boxes, output_scores, output_labels]
|
||||
|
||||
# Create the NMS Plugin node with the selected inputs. The outputs of the node will also
|
||||
# become the final outputs of the graph.
|
||||
self.graph.layer(op=op, name="batched_nms", inputs=op_inputs, outputs=op_outputs, attrs=attrs)
|
||||
LOGGER.info(f"Created NMS plugin '{op}' with attributes: {attrs}")
|
||||
|
||||
self.graph.outputs = op_outputs
|
||||
|
||||
self.infer()
|
||||
|
||||
def save(self, output_path):
|
||||
"""
|
||||
Save the ONNX model to the given location.
|
||||
Args:
|
||||
output_path: Path pointing to the location where to write
|
||||
out the updated ONNX model.
|
||||
"""
|
||||
self.graph.cleanup().toposort()
|
||||
model = gs.export_onnx(self.graph)
|
||||
onnx.save(model, output_path)
|
||||
LOGGER.info(f"Saved ONNX model to {output_path}")
|
||||
@@ -0,0 +1,160 @@
|
||||
# Auto-anchor utils
|
||||
|
||||
import numpy as np
|
||||
import torch
|
||||
import yaml
|
||||
from scipy.cluster.vq import kmeans
|
||||
from tqdm import tqdm
|
||||
|
||||
from utils.general import colorstr
|
||||
|
||||
|
||||
def check_anchor_order(m):
|
||||
# Check anchor order against stride order for YOLO Detect() module m, and correct if necessary
|
||||
a = m.anchor_grid.prod(-1).view(-1) # anchor area
|
||||
da = a[-1] - a[0] # delta a
|
||||
ds = m.stride[-1] - m.stride[0] # delta s
|
||||
if da.sign() != ds.sign(): # same order
|
||||
print('Reversing anchor order')
|
||||
m.anchors[:] = m.anchors.flip(0)
|
||||
m.anchor_grid[:] = m.anchor_grid.flip(0)
|
||||
|
||||
|
||||
def check_anchors(dataset, model, thr=4.0, imgsz=640):
|
||||
# Check anchor fit to data, recompute if necessary
|
||||
prefix = colorstr('autoanchor: ')
|
||||
print(f'\n{prefix}Analyzing anchors... ', end='')
|
||||
m = model.module.model[-1] if hasattr(model, 'module') else model.model[-1] # Detect()
|
||||
shapes = imgsz * dataset.shapes / dataset.shapes.max(1, keepdims=True)
|
||||
scale = np.random.uniform(0.9, 1.1, size=(shapes.shape[0], 1)) # augment scale
|
||||
wh = torch.tensor(np.concatenate([l[:, 3:5] * s for s, l in zip(shapes * scale, dataset.labels)])).float() # wh
|
||||
|
||||
def metric(k): # compute metric
|
||||
r = wh[:, None] / k[None]
|
||||
x = torch.min(r, 1. / r).min(2)[0] # ratio metric
|
||||
best = x.max(1)[0] # best_x
|
||||
aat = (x > 1. / thr).float().sum(1).mean() # anchors above threshold
|
||||
bpr = (best > 1. / thr).float().mean() # best possible recall
|
||||
return bpr, aat
|
||||
|
||||
anchors = m.anchor_grid.clone().cpu().view(-1, 2) # current anchors
|
||||
bpr, aat = metric(anchors)
|
||||
print(f'anchors/target = {aat:.2f}, Best Possible Recall (BPR) = {bpr:.4f}', end='')
|
||||
if bpr < 0.98: # threshold to recompute
|
||||
print('. Attempting to improve anchors, please wait...')
|
||||
na = m.anchor_grid.numel() // 2 # number of anchors
|
||||
try:
|
||||
anchors = kmean_anchors(dataset, n=na, img_size=imgsz, thr=thr, gen=1000, verbose=False)
|
||||
except Exception as e:
|
||||
print(f'{prefix}ERROR: {e}')
|
||||
new_bpr = metric(anchors)[0]
|
||||
if new_bpr > bpr: # replace anchors
|
||||
anchors = torch.tensor(anchors, device=m.anchors.device).type_as(m.anchors)
|
||||
m.anchor_grid[:] = anchors.clone().view_as(m.anchor_grid) # for inference
|
||||
check_anchor_order(m)
|
||||
m.anchors[:] = anchors.clone().view_as(m.anchors) / m.stride.to(m.anchors.device).view(-1, 1, 1) # loss
|
||||
print(f'{prefix}New anchors saved to model. Update model *.yaml to use these anchors in the future.')
|
||||
else:
|
||||
print(f'{prefix}Original anchors better than new anchors. Proceeding with original anchors.')
|
||||
print('') # newline
|
||||
|
||||
|
||||
def kmean_anchors(path='./data/coco.yaml', n=9, img_size=640, thr=4.0, gen=1000, verbose=True):
|
||||
""" Creates kmeans-evolved anchors from training dataset
|
||||
|
||||
Arguments:
|
||||
path: path to dataset *.yaml, or a loaded dataset
|
||||
n: number of anchors
|
||||
img_size: image size used for training
|
||||
thr: anchor-label wh ratio threshold hyperparameter hyp['anchor_t'] used for training, default=4.0
|
||||
gen: generations to evolve anchors using genetic algorithm
|
||||
verbose: print all results
|
||||
|
||||
Return:
|
||||
k: kmeans evolved anchors
|
||||
|
||||
Usage:
|
||||
from utils.autoanchor import *; _ = kmean_anchors()
|
||||
"""
|
||||
thr = 1. / thr
|
||||
prefix = colorstr('autoanchor: ')
|
||||
|
||||
def metric(k, wh): # compute metrics
|
||||
r = wh[:, None] / k[None]
|
||||
x = torch.min(r, 1. / r).min(2)[0] # ratio metric
|
||||
# x = wh_iou(wh, torch.tensor(k)) # iou metric
|
||||
return x, x.max(1)[0] # x, best_x
|
||||
|
||||
def anchor_fitness(k): # mutation fitness
|
||||
_, best = metric(torch.tensor(k, dtype=torch.float32), wh)
|
||||
return (best * (best > thr).float()).mean() # fitness
|
||||
|
||||
def print_results(k):
|
||||
k = k[np.argsort(k.prod(1))] # sort small to large
|
||||
x, best = metric(k, wh0)
|
||||
bpr, aat = (best > thr).float().mean(), (x > thr).float().mean() * n # best possible recall, anch > thr
|
||||
print(f'{prefix}thr={thr:.2f}: {bpr:.4f} best possible recall, {aat:.2f} anchors past thr')
|
||||
print(f'{prefix}n={n}, img_size={img_size}, metric_all={x.mean():.3f}/{best.mean():.3f}-mean/best, '
|
||||
f'past_thr={x[x > thr].mean():.3f}-mean: ', end='')
|
||||
for i, x in enumerate(k):
|
||||
print('%i,%i' % (round(x[0]), round(x[1])), end=', ' if i < len(k) - 1 else '\n') # use in *.cfg
|
||||
return k
|
||||
|
||||
if isinstance(path, str): # *.yaml file
|
||||
with open(path) as f:
|
||||
data_dict = yaml.load(f, Loader=yaml.SafeLoader) # model dict
|
||||
from utils.datasets import LoadImagesAndLabels
|
||||
dataset = LoadImagesAndLabels(data_dict['train'], augment=True, rect=True)
|
||||
else:
|
||||
dataset = path # dataset
|
||||
|
||||
# Get label wh
|
||||
shapes = img_size * dataset.shapes / dataset.shapes.max(1, keepdims=True)
|
||||
wh0 = np.concatenate([l[:, 3:5] * s for s, l in zip(shapes, dataset.labels)]) # wh
|
||||
|
||||
# Filter
|
||||
i = (wh0 < 3.0).any(1).sum()
|
||||
if i:
|
||||
print(f'{prefix}WARNING: Extremely small objects found. {i} of {len(wh0)} labels are < 3 pixels in size.')
|
||||
wh = wh0[(wh0 >= 2.0).any(1)] # filter > 2 pixels
|
||||
# wh = wh * (np.random.rand(wh.shape[0], 1) * 0.9 + 0.1) # multiply by random scale 0-1
|
||||
|
||||
# Kmeans calculation
|
||||
print(f'{prefix}Running kmeans for {n} anchors on {len(wh)} points...')
|
||||
s = wh.std(0) # sigmas for whitening
|
||||
k, dist = kmeans(wh / s, n, iter=30) # points, mean distance
|
||||
assert len(k) == n, print(f'{prefix}ERROR: scipy.cluster.vq.kmeans requested {n} points but returned only {len(k)}')
|
||||
k *= s
|
||||
wh = torch.tensor(wh, dtype=torch.float32) # filtered
|
||||
wh0 = torch.tensor(wh0, dtype=torch.float32) # unfiltered
|
||||
k = print_results(k)
|
||||
|
||||
# Plot
|
||||
# k, d = [None] * 20, [None] * 20
|
||||
# for i in tqdm(range(1, 21)):
|
||||
# k[i-1], d[i-1] = kmeans(wh / s, i) # points, mean distance
|
||||
# fig, ax = plt.subplots(1, 2, figsize=(14, 7), tight_layout=True)
|
||||
# ax = ax.ravel()
|
||||
# ax[0].plot(np.arange(1, 21), np.array(d) ** 2, marker='.')
|
||||
# fig, ax = plt.subplots(1, 2, figsize=(14, 7)) # plot wh
|
||||
# ax[0].hist(wh[wh[:, 0]<100, 0],400)
|
||||
# ax[1].hist(wh[wh[:, 1]<100, 1],400)
|
||||
# fig.savefig('wh.png', dpi=200)
|
||||
|
||||
# Evolve
|
||||
npr = np.random
|
||||
f, sh, mp, s = anchor_fitness(k), k.shape, 0.9, 0.1 # fitness, generations, mutation prob, sigma
|
||||
pbar = tqdm(range(gen), desc=f'{prefix}Evolving anchors with Genetic Algorithm:') # progress bar
|
||||
for _ in pbar:
|
||||
v = np.ones(sh)
|
||||
while (v == 1).all(): # mutate until a change occurs (prevent duplicates)
|
||||
v = ((npr.random(sh) < mp) * npr.random() * npr.randn(*sh) * s + 1).clip(0.3, 3.0)
|
||||
kg = (k.copy() * v).clip(min=2.0)
|
||||
fg = anchor_fitness(kg)
|
||||
if fg > f:
|
||||
f, k = fg, kg.copy()
|
||||
pbar.desc = f'{prefix}Evolving anchors with Genetic Algorithm: fitness = {f:.4f}'
|
||||
if verbose:
|
||||
print_results(k)
|
||||
|
||||
return print_results(k)
|
||||
@@ -0,0 +1 @@
|
||||
#init
|
||||
@@ -0,0 +1,26 @@
|
||||
# AWS EC2 instance startup 'MIME' script https://aws.amazon.com/premiumsupport/knowledge-center/execute-user-data-ec2/
|
||||
# This script will run on every instance restart, not only on first start
|
||||
# --- DO NOT COPY ABOVE COMMENTS WHEN PASTING INTO USERDATA ---
|
||||
|
||||
Content-Type: multipart/mixed; boundary="//"
|
||||
MIME-Version: 1.0
|
||||
|
||||
--//
|
||||
Content-Type: text/cloud-config; charset="us-ascii"
|
||||
MIME-Version: 1.0
|
||||
Content-Transfer-Encoding: 7bit
|
||||
Content-Disposition: attachment; filename="cloud-config.txt"
|
||||
|
||||
#cloud-config
|
||||
cloud_final_modules:
|
||||
- [scripts-user, always]
|
||||
|
||||
--//
|
||||
Content-Type: text/x-shellscript; charset="us-ascii"
|
||||
MIME-Version: 1.0
|
||||
Content-Transfer-Encoding: 7bit
|
||||
Content-Disposition: attachment; filename="userdata.txt"
|
||||
|
||||
#!/bin/bash
|
||||
# --- paste contents of userdata.sh here ---
|
||||
--//
|
||||
@@ -0,0 +1,37 @@
|
||||
# Resume all interrupted trainings in yolor/ dir including DDP trainings
|
||||
# Usage: $ python utils/aws/resume.py
|
||||
|
||||
import os
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
import torch
|
||||
import yaml
|
||||
|
||||
sys.path.append('./') # to run '$ python *.py' files in subdirectories
|
||||
|
||||
port = 0 # --master_port
|
||||
path = Path('').resolve()
|
||||
for last in path.rglob('*/**/last.pt'):
|
||||
ckpt = torch.load(last)
|
||||
if ckpt['optimizer'] is None:
|
||||
continue
|
||||
|
||||
# Load opt.yaml
|
||||
with open(last.parent.parent / 'opt.yaml') as f:
|
||||
opt = yaml.load(f, Loader=yaml.SafeLoader)
|
||||
|
||||
# Get device count
|
||||
d = opt['device'].split(',') # devices
|
||||
nd = len(d) # number of devices
|
||||
ddp = nd > 1 or (nd == 0 and torch.cuda.device_count() > 1) # distributed data parallel
|
||||
|
||||
if ddp: # multi-GPU
|
||||
port += 1
|
||||
cmd = f'python -m torch.distributed.launch --nproc_per_node {nd} --master_port {port} train.py --resume {last}'
|
||||
else: # single-GPU
|
||||
cmd = f'python train.py --resume {last}'
|
||||
|
||||
cmd += ' > /dev/null 2>&1 &' # redirect output to dev/null and run in daemon thread
|
||||
print(cmd)
|
||||
os.system(cmd)
|
||||
@@ -0,0 +1,27 @@
|
||||
#!/bin/bash
|
||||
# AWS EC2 instance startup script https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/user-data.html
|
||||
# This script will run only once on first instance start (for a re-start script see mime.sh)
|
||||
# /home/ubuntu (ubuntu) or /home/ec2-user (amazon-linux) is working dir
|
||||
# Use >300 GB SSD
|
||||
|
||||
cd home/ubuntu
|
||||
if [ ! -d yolor ]; then
|
||||
echo "Running first-time script." # install dependencies, download COCO, pull Docker
|
||||
git clone -b main https://github.com/WongKinYiu/yolov7 && sudo chmod -R 777 yolov7
|
||||
cd yolov7
|
||||
bash data/scripts/get_coco.sh && echo "Data done." &
|
||||
sudo docker pull nvcr.io/nvidia/pytorch:21.08-py3 && echo "Docker done." &
|
||||
python -m pip install --upgrade pip && pip install -r requirements.txt && python detect.py && echo "Requirements done." &
|
||||
wait && echo "All tasks done." # finish background tasks
|
||||
else
|
||||
echo "Running re-start script." # resume interrupted runs
|
||||
i=0
|
||||
list=$(sudo docker ps -qa) # container list i.e. $'one\ntwo\nthree\nfour'
|
||||
while IFS= read -r id; do
|
||||
((i++))
|
||||
echo "restarting container $i: $id"
|
||||
sudo docker start $id
|
||||
# sudo docker exec -it $id python train.py --resume # single-GPU
|
||||
sudo docker exec -d $id python utils/aws/resume.py # multi-scenario
|
||||
done <<<"$list"
|
||||
fi
|
||||
File diff suppressed because it is too large
Load Diff
@@ -0,0 +1,892 @@
|
||||
# YOLOR general utils
|
||||
|
||||
import glob
|
||||
import logging
|
||||
import math
|
||||
import os
|
||||
import platform
|
||||
import random
|
||||
import re
|
||||
import subprocess
|
||||
import time
|
||||
from pathlib import Path
|
||||
|
||||
import cv2
|
||||
import numpy as np
|
||||
import pandas as pd
|
||||
import torch
|
||||
import torchvision
|
||||
import yaml
|
||||
|
||||
from utils.google_utils import gsutil_getsize
|
||||
from utils.metrics import fitness
|
||||
from utils.torch_utils import init_torch_seeds
|
||||
|
||||
# Settings
|
||||
torch.set_printoptions(linewidth=320, precision=5, profile='long')
|
||||
np.set_printoptions(linewidth=320, formatter={'float_kind': '{:11.5g}'.format}) # format short g, %precision=5
|
||||
pd.options.display.max_columns = 10
|
||||
cv2.setNumThreads(0) # prevent OpenCV from multithreading (incompatible with PyTorch DataLoader)
|
||||
os.environ['NUMEXPR_MAX_THREADS'] = str(min(os.cpu_count(), 8)) # NumExpr max threads
|
||||
|
||||
|
||||
def set_logging(rank=-1):
|
||||
logging.basicConfig(
|
||||
format="%(message)s",
|
||||
level=logging.INFO if rank in [-1, 0] else logging.WARN)
|
||||
|
||||
|
||||
def init_seeds(seed=0):
|
||||
# Initialize random number generator (RNG) seeds
|
||||
random.seed(seed)
|
||||
np.random.seed(seed)
|
||||
init_torch_seeds(seed)
|
||||
|
||||
|
||||
def get_latest_run(search_dir='.'):
|
||||
# Return path to most recent 'last.pt' in /runs (i.e. to --resume from)
|
||||
last_list = glob.glob(f'{search_dir}/**/last*.pt', recursive=True)
|
||||
return max(last_list, key=os.path.getctime) if last_list else ''
|
||||
|
||||
|
||||
def isdocker():
|
||||
# Is environment a Docker container
|
||||
return Path('/workspace').exists() # or Path('/.dockerenv').exists()
|
||||
|
||||
|
||||
def emojis(str=''):
|
||||
# Return platform-dependent emoji-safe version of string
|
||||
return str.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else str
|
||||
|
||||
|
||||
def check_online():
|
||||
# Check internet connectivity
|
||||
import socket
|
||||
try:
|
||||
socket.create_connection(("1.1.1.1", 443), 5) # check host accesability
|
||||
return True
|
||||
except OSError:
|
||||
return False
|
||||
|
||||
|
||||
def check_git_status():
|
||||
# Recommend 'git pull' if code is out of date
|
||||
print(colorstr('github: '), end='')
|
||||
try:
|
||||
assert Path('.git').exists(), 'skipping check (not a git repository)'
|
||||
assert not isdocker(), 'skipping check (Docker image)'
|
||||
assert check_online(), 'skipping check (offline)'
|
||||
|
||||
cmd = 'git fetch && git config --get remote.origin.url'
|
||||
url = subprocess.check_output(cmd, shell=True).decode().strip().rstrip('.git') # github repo url
|
||||
branch = subprocess.check_output('git rev-parse --abbrev-ref HEAD', shell=True).decode().strip() # checked out
|
||||
n = int(subprocess.check_output(f'git rev-list {branch}..origin/master --count', shell=True)) # commits behind
|
||||
if n > 0:
|
||||
s = f"⚠️ WARNING: code is out of date by {n} commit{'s' * (n > 1)}. " \
|
||||
f"Use 'git pull' to update or 'git clone {url}' to download latest."
|
||||
else:
|
||||
s = f'up to date with {url} ✅'
|
||||
print(emojis(s)) # emoji-safe
|
||||
except Exception as e:
|
||||
print(e)
|
||||
|
||||
|
||||
def check_requirements(requirements='requirements.txt', exclude=()):
|
||||
# Check installed dependencies meet requirements (pass *.txt file or list of packages)
|
||||
import pkg_resources as pkg
|
||||
prefix = colorstr('red', 'bold', 'requirements:')
|
||||
if isinstance(requirements, (str, Path)): # requirements.txt file
|
||||
file = Path(requirements)
|
||||
if not file.exists():
|
||||
print(f"{prefix} {file.resolve()} not found, check failed.")
|
||||
return
|
||||
requirements = [f'{x.name}{x.specifier}' for x in pkg.parse_requirements(file.open()) if x.name not in exclude]
|
||||
else: # list or tuple of packages
|
||||
requirements = [x for x in requirements if x not in exclude]
|
||||
|
||||
n = 0 # number of packages updates
|
||||
for r in requirements:
|
||||
try:
|
||||
pkg.require(r)
|
||||
except Exception as e: # DistributionNotFound or VersionConflict if requirements not met
|
||||
n += 1
|
||||
print(f"{prefix} {e.req} not found and is required by YOLOR, attempting auto-update...")
|
||||
print(subprocess.check_output(f"pip install '{e.req}'", shell=True).decode())
|
||||
|
||||
if n: # if packages updated
|
||||
source = file.resolve() if 'file' in locals() else requirements
|
||||
s = f"{prefix} {n} package{'s' * (n > 1)} updated per {source}\n" \
|
||||
f"{prefix} ⚠️ {colorstr('bold', 'Restart runtime or rerun command for updates to take effect')}\n"
|
||||
print(emojis(s)) # emoji-safe
|
||||
|
||||
|
||||
def check_img_size(img_size, s=32):
|
||||
# Verify img_size is a multiple of stride s
|
||||
new_size = make_divisible(img_size, int(s)) # ceil gs-multiple
|
||||
if new_size != img_size:
|
||||
print('WARNING: --img-size %g must be multiple of max stride %g, updating to %g' % (img_size, s, new_size))
|
||||
return new_size
|
||||
|
||||
|
||||
def check_imshow():
|
||||
# Check if environment supports image displays
|
||||
try:
|
||||
assert not isdocker(), 'cv2.imshow() is disabled in Docker environments'
|
||||
cv2.imshow('test', np.zeros((1, 1, 3)))
|
||||
cv2.waitKey(1)
|
||||
cv2.destroyAllWindows()
|
||||
cv2.waitKey(1)
|
||||
return True
|
||||
except Exception as e:
|
||||
print(f'WARNING: Environment does not support cv2.imshow() or PIL Image.show() image displays\n{e}')
|
||||
return False
|
||||
|
||||
|
||||
def check_file(file):
|
||||
# Search for file if not found
|
||||
if Path(file).is_file() or file == '':
|
||||
return file
|
||||
else:
|
||||
files = glob.glob('./**/' + file, recursive=True) # find file
|
||||
assert len(files), f'File Not Found: {file}' # assert file was found
|
||||
assert len(files) == 1, f"Multiple files match '{file}', specify exact path: {files}" # assert unique
|
||||
return files[0] # return file
|
||||
|
||||
|
||||
def check_dataset(dict):
|
||||
# Download dataset if not found locally
|
||||
val, s = dict.get('val'), dict.get('download')
|
||||
if val and len(val):
|
||||
val = [Path(x).resolve() for x in (val if isinstance(val, list) else [val])] # val path
|
||||
if not all(x.exists() for x in val):
|
||||
print('\nWARNING: Dataset not found, nonexistent paths: %s' % [str(x) for x in val if not x.exists()])
|
||||
if s and len(s): # download script
|
||||
print('Downloading %s ...' % s)
|
||||
if s.startswith('http') and s.endswith('.zip'): # URL
|
||||
f = Path(s).name # filename
|
||||
torch.hub.download_url_to_file(s, f)
|
||||
r = os.system('unzip -q %s -d ../ && rm %s' % (f, f)) # unzip
|
||||
else: # bash script
|
||||
r = os.system(s)
|
||||
print('Dataset autodownload %s\n' % ('success' if r == 0 else 'failure')) # analyze return value
|
||||
else:
|
||||
raise Exception('Dataset not found.')
|
||||
|
||||
|
||||
def make_divisible(x, divisor):
|
||||
# Returns x evenly divisible by divisor
|
||||
return math.ceil(x / divisor) * divisor
|
||||
|
||||
|
||||
def clean_str(s):
|
||||
# Cleans a string by replacing special characters with underscore _
|
||||
return re.sub(pattern="[|@#!¡·$€%&()=?¿^*;:,¨´><+]", repl="_", string=s)
|
||||
|
||||
|
||||
def one_cycle(y1=0.0, y2=1.0, steps=100):
|
||||
# lambda function for sinusoidal ramp from y1 to y2
|
||||
return lambda x: ((1 - math.cos(x * math.pi / steps)) / 2) * (y2 - y1) + y1
|
||||
|
||||
|
||||
def colorstr(*input):
|
||||
# Colors a string https://en.wikipedia.org/wiki/ANSI_escape_code, i.e. colorstr('blue', 'hello world')
|
||||
*args, string = input if len(input) > 1 else ('blue', 'bold', input[0]) # color arguments, string
|
||||
colors = {'black': '\033[30m', # basic colors
|
||||
'red': '\033[31m',
|
||||
'green': '\033[32m',
|
||||
'yellow': '\033[33m',
|
||||
'blue': '\033[34m',
|
||||
'magenta': '\033[35m',
|
||||
'cyan': '\033[36m',
|
||||
'white': '\033[37m',
|
||||
'bright_black': '\033[90m', # bright colors
|
||||
'bright_red': '\033[91m',
|
||||
'bright_green': '\033[92m',
|
||||
'bright_yellow': '\033[93m',
|
||||
'bright_blue': '\033[94m',
|
||||
'bright_magenta': '\033[95m',
|
||||
'bright_cyan': '\033[96m',
|
||||
'bright_white': '\033[97m',
|
||||
'end': '\033[0m', # misc
|
||||
'bold': '\033[1m',
|
||||
'underline': '\033[4m'}
|
||||
return ''.join(colors[x] for x in args) + f'{string}' + colors['end']
|
||||
|
||||
|
||||
def labels_to_class_weights(labels, nc=80):
|
||||
# Get class weights (inverse frequency) from training labels
|
||||
if labels[0] is None: # no labels loaded
|
||||
return torch.Tensor()
|
||||
|
||||
labels = np.concatenate(labels, 0) # labels.shape = (866643, 5) for COCO
|
||||
classes = labels[:, 0].astype(np.int) # labels = [class xywh]
|
||||
weights = np.bincount(classes, minlength=nc) # occurrences per class
|
||||
|
||||
# Prepend gridpoint count (for uCE training)
|
||||
# gpi = ((320 / 32 * np.array([1, 2, 4])) ** 2 * 3).sum() # gridpoints per image
|
||||
# weights = np.hstack([gpi * len(labels) - weights.sum() * 9, weights * 9]) ** 0.5 # prepend gridpoints to start
|
||||
|
||||
weights[weights == 0] = 1 # replace empty bins with 1
|
||||
weights = 1 / weights # number of targets per class
|
||||
weights /= weights.sum() # normalize
|
||||
return torch.from_numpy(weights)
|
||||
|
||||
|
||||
def labels_to_image_weights(labels, nc=80, class_weights=np.ones(80)):
|
||||
# Produces image weights based on class_weights and image contents
|
||||
class_counts = np.array([np.bincount(x[:, 0].astype(np.int), minlength=nc) for x in labels])
|
||||
image_weights = (class_weights.reshape(1, nc) * class_counts).sum(1)
|
||||
# index = random.choices(range(n), weights=image_weights, k=1) # weight image sample
|
||||
return image_weights
|
||||
|
||||
|
||||
def coco80_to_coco91_class(): # converts 80-index (val2014) to 91-index (paper)
|
||||
# https://tech.amikelive.com/node-718/what-object-categories-labels-are-in-coco-dataset/
|
||||
# a = np.loadtxt('data/coco.names', dtype='str', delimiter='\n')
|
||||
# b = np.loadtxt('data/coco_paper.names', dtype='str', delimiter='\n')
|
||||
# x1 = [list(a[i] == b).index(True) + 1 for i in range(80)] # darknet to coco
|
||||
# x2 = [list(b[i] == a).index(True) if any(b[i] == a) else None for i in range(91)] # coco to darknet
|
||||
x = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34,
|
||||
35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63,
|
||||
64, 65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 84, 85, 86, 87, 88, 89, 90]
|
||||
return x
|
||||
|
||||
|
||||
def xyxy2xywh(x):
|
||||
# Convert nx4 boxes from [x1, y1, x2, y2] to [x, y, w, h] where xy1=top-left, xy2=bottom-right
|
||||
y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
|
||||
y[:, 0] = (x[:, 0] + x[:, 2]) / 2 # x center
|
||||
y[:, 1] = (x[:, 1] + x[:, 3]) / 2 # y center
|
||||
y[:, 2] = x[:, 2] - x[:, 0] # width
|
||||
y[:, 3] = x[:, 3] - x[:, 1] # height
|
||||
return y
|
||||
|
||||
|
||||
def xywh2xyxy(x):
|
||||
# Convert nx4 boxes from [x, y, w, h] to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right
|
||||
y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
|
||||
y[:, 0] = x[:, 0] - x[:, 2] / 2 # top left x
|
||||
y[:, 1] = x[:, 1] - x[:, 3] / 2 # top left y
|
||||
y[:, 2] = x[:, 0] + x[:, 2] / 2 # bottom right x
|
||||
y[:, 3] = x[:, 1] + x[:, 3] / 2 # bottom right y
|
||||
return y
|
||||
|
||||
|
||||
def xywhn2xyxy(x, w=640, h=640, padw=0, padh=0):
|
||||
# Convert nx4 boxes from [x, y, w, h] normalized to [x1, y1, x2, y2] where xy1=top-left, xy2=bottom-right
|
||||
y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
|
||||
y[:, 0] = w * (x[:, 0] - x[:, 2] / 2) + padw # top left x
|
||||
y[:, 1] = h * (x[:, 1] - x[:, 3] / 2) + padh # top left y
|
||||
y[:, 2] = w * (x[:, 0] + x[:, 2] / 2) + padw # bottom right x
|
||||
y[:, 3] = h * (x[:, 1] + x[:, 3] / 2) + padh # bottom right y
|
||||
return y
|
||||
|
||||
|
||||
def xyn2xy(x, w=640, h=640, padw=0, padh=0):
|
||||
# Convert normalized segments into pixel segments, shape (n,2)
|
||||
y = x.clone() if isinstance(x, torch.Tensor) else np.copy(x)
|
||||
y[:, 0] = w * x[:, 0] + padw # top left x
|
||||
y[:, 1] = h * x[:, 1] + padh # top left y
|
||||
return y
|
||||
|
||||
|
||||
def segment2box(segment, width=640, height=640):
|
||||
# Convert 1 segment label to 1 box label, applying inside-image constraint, i.e. (xy1, xy2, ...) to (xyxy)
|
||||
x, y = segment.T # segment xy
|
||||
inside = (x >= 0) & (y >= 0) & (x <= width) & (y <= height)
|
||||
x, y, = x[inside], y[inside]
|
||||
return np.array([x.min(), y.min(), x.max(), y.max()]) if any(x) else np.zeros((1, 4)) # xyxy
|
||||
|
||||
|
||||
def segments2boxes(segments):
|
||||
# Convert segment labels to box labels, i.e. (cls, xy1, xy2, ...) to (cls, xywh)
|
||||
boxes = []
|
||||
for s in segments:
|
||||
x, y = s.T # segment xy
|
||||
boxes.append([x.min(), y.min(), x.max(), y.max()]) # cls, xyxy
|
||||
return xyxy2xywh(np.array(boxes)) # cls, xywh
|
||||
|
||||
|
||||
def resample_segments(segments, n=1000):
|
||||
# Up-sample an (n,2) segment
|
||||
for i, s in enumerate(segments):
|
||||
s = np.concatenate((s, s[0:1, :]), axis=0)
|
||||
x = np.linspace(0, len(s) - 1, n)
|
||||
xp = np.arange(len(s))
|
||||
segments[i] = np.concatenate([np.interp(x, xp, s[:, i]) for i in range(2)]).reshape(2, -1).T # segment xy
|
||||
return segments
|
||||
|
||||
|
||||
def scale_coords(img1_shape, coords, img0_shape, ratio_pad=None):
|
||||
# Rescale coords (xyxy) from img1_shape to img0_shape
|
||||
if ratio_pad is None: # calculate from img0_shape
|
||||
gain = min(img1_shape[0] / img0_shape[0], img1_shape[1] / img0_shape[1]) # gain = old / new
|
||||
pad = (img1_shape[1] - img0_shape[1] * gain) / 2, (img1_shape[0] - img0_shape[0] * gain) / 2 # wh padding
|
||||
else:
|
||||
gain = ratio_pad[0][0]
|
||||
pad = ratio_pad[1]
|
||||
|
||||
coords[:, [0, 2]] -= pad[0] # x padding
|
||||
coords[:, [1, 3]] -= pad[1] # y padding
|
||||
coords[:, :4] /= gain
|
||||
clip_coords(coords, img0_shape)
|
||||
return coords
|
||||
|
||||
|
||||
def clip_coords(boxes, img_shape):
|
||||
# Clip bounding xyxy bounding boxes to image shape (height, width)
|
||||
boxes[:, 0].clamp_(0, img_shape[1]) # x1
|
||||
boxes[:, 1].clamp_(0, img_shape[0]) # y1
|
||||
boxes[:, 2].clamp_(0, img_shape[1]) # x2
|
||||
boxes[:, 3].clamp_(0, img_shape[0]) # y2
|
||||
|
||||
|
||||
def bbox_iou(box1, box2, x1y1x2y2=True, GIoU=False, DIoU=False, CIoU=False, eps=1e-7):
|
||||
# Returns the IoU of box1 to box2. box1 is 4, box2 is nx4
|
||||
box2 = box2.T
|
||||
|
||||
# Get the coordinates of bounding boxes
|
||||
if x1y1x2y2: # x1, y1, x2, y2 = box1
|
||||
b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3]
|
||||
b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3]
|
||||
else: # transform from xywh to xyxy
|
||||
b1_x1, b1_x2 = box1[0] - box1[2] / 2, box1[0] + box1[2] / 2
|
||||
b1_y1, b1_y2 = box1[1] - box1[3] / 2, box1[1] + box1[3] / 2
|
||||
b2_x1, b2_x2 = box2[0] - box2[2] / 2, box2[0] + box2[2] / 2
|
||||
b2_y1, b2_y2 = box2[1] - box2[3] / 2, box2[1] + box2[3] / 2
|
||||
|
||||
# Intersection area
|
||||
inter = (torch.min(b1_x2, b2_x2) - torch.max(b1_x1, b2_x1)).clamp(0) * \
|
||||
(torch.min(b1_y2, b2_y2) - torch.max(b1_y1, b2_y1)).clamp(0)
|
||||
|
||||
# Union Area
|
||||
w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps
|
||||
w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps
|
||||
union = w1 * h1 + w2 * h2 - inter + eps
|
||||
|
||||
iou = inter / union
|
||||
|
||||
if GIoU or DIoU or CIoU:
|
||||
cw = torch.max(b1_x2, b2_x2) - torch.min(b1_x1, b2_x1) # convex (smallest enclosing box) width
|
||||
ch = torch.max(b1_y2, b2_y2) - torch.min(b1_y1, b2_y1) # convex height
|
||||
if CIoU or DIoU: # Distance or Complete IoU https://arxiv.org/abs/1911.08287v1
|
||||
c2 = cw ** 2 + ch ** 2 + eps # convex diagonal squared
|
||||
rho2 = ((b2_x1 + b2_x2 - b1_x1 - b1_x2) ** 2 +
|
||||
(b2_y1 + b2_y2 - b1_y1 - b1_y2) ** 2) / 4 # center distance squared
|
||||
if DIoU:
|
||||
return iou - rho2 / c2 # DIoU
|
||||
elif CIoU: # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47
|
||||
v = (4 / math.pi ** 2) * torch.pow(torch.atan(w2 / (h2 + eps)) - torch.atan(w1 / (h1 + eps)), 2)
|
||||
with torch.no_grad():
|
||||
alpha = v / (v - iou + (1 + eps))
|
||||
return iou - (rho2 / c2 + v * alpha) # CIoU
|
||||
else: # GIoU https://arxiv.org/pdf/1902.09630.pdf
|
||||
c_area = cw * ch + eps # convex area
|
||||
return iou - (c_area - union) / c_area # GIoU
|
||||
else:
|
||||
return iou # IoU
|
||||
|
||||
|
||||
|
||||
|
||||
def bbox_alpha_iou(box1, box2, x1y1x2y2=False, GIoU=False, DIoU=False, CIoU=False, alpha=2, eps=1e-9):
|
||||
# Returns tsqrt_he IoU of box1 to box2. box1 is 4, box2 is nx4
|
||||
box2 = box2.T
|
||||
|
||||
# Get the coordinates of bounding boxes
|
||||
if x1y1x2y2: # x1, y1, x2, y2 = box1
|
||||
b1_x1, b1_y1, b1_x2, b1_y2 = box1[0], box1[1], box1[2], box1[3]
|
||||
b2_x1, b2_y1, b2_x2, b2_y2 = box2[0], box2[1], box2[2], box2[3]
|
||||
else: # transform from xywh to xyxy
|
||||
b1_x1, b1_x2 = box1[0] - box1[2] / 2, box1[0] + box1[2] / 2
|
||||
b1_y1, b1_y2 = box1[1] - box1[3] / 2, box1[1] + box1[3] / 2
|
||||
b2_x1, b2_x2 = box2[0] - box2[2] / 2, box2[0] + box2[2] / 2
|
||||
b2_y1, b2_y2 = box2[1] - box2[3] / 2, box2[1] + box2[3] / 2
|
||||
|
||||
# Intersection area
|
||||
inter = (torch.min(b1_x2, b2_x2) - torch.max(b1_x1, b2_x1)).clamp(0) * \
|
||||
(torch.min(b1_y2, b2_y2) - torch.max(b1_y1, b2_y1)).clamp(0)
|
||||
|
||||
# Union Area
|
||||
w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps
|
||||
w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps
|
||||
union = w1 * h1 + w2 * h2 - inter + eps
|
||||
|
||||
# change iou into pow(iou+eps)
|
||||
# iou = inter / union
|
||||
iou = torch.pow(inter/union + eps, alpha)
|
||||
# beta = 2 * alpha
|
||||
if GIoU or DIoU or CIoU:
|
||||
cw = torch.max(b1_x2, b2_x2) - torch.min(b1_x1, b2_x1) # convex (smallest enclosing box) width
|
||||
ch = torch.max(b1_y2, b2_y2) - torch.min(b1_y1, b2_y1) # convex height
|
||||
if CIoU or DIoU: # Distance or Complete IoU https://arxiv.org/abs/1911.08287v1
|
||||
c2 = (cw ** 2 + ch ** 2) ** alpha + eps # convex diagonal
|
||||
rho_x = torch.abs(b2_x1 + b2_x2 - b1_x1 - b1_x2)
|
||||
rho_y = torch.abs(b2_y1 + b2_y2 - b1_y1 - b1_y2)
|
||||
rho2 = ((rho_x ** 2 + rho_y ** 2) / 4) ** alpha # center distance
|
||||
if DIoU:
|
||||
return iou - rho2 / c2 # DIoU
|
||||
elif CIoU: # https://github.com/Zzh-tju/DIoU-SSD-pytorch/blob/master/utils/box/box_utils.py#L47
|
||||
v = (4 / math.pi ** 2) * torch.pow(torch.atan(w2 / h2) - torch.atan(w1 / h1), 2)
|
||||
with torch.no_grad():
|
||||
alpha_ciou = v / ((1 + eps) - inter / union + v)
|
||||
# return iou - (rho2 / c2 + v * alpha_ciou) # CIoU
|
||||
return iou - (rho2 / c2 + torch.pow(v * alpha_ciou + eps, alpha)) # CIoU
|
||||
else: # GIoU https://arxiv.org/pdf/1902.09630.pdf
|
||||
# c_area = cw * ch + eps # convex area
|
||||
# return iou - (c_area - union) / c_area # GIoU
|
||||
c_area = torch.max(cw * ch + eps, union) # convex area
|
||||
return iou - torch.pow((c_area - union) / c_area + eps, alpha) # GIoU
|
||||
else:
|
||||
return iou # torch.log(iou+eps) or iou
|
||||
|
||||
|
||||
def box_iou(box1, box2):
|
||||
# https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py
|
||||
"""
|
||||
Return intersection-over-union (Jaccard index) of boxes.
|
||||
Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
|
||||
Arguments:
|
||||
box1 (Tensor[N, 4])
|
||||
box2 (Tensor[M, 4])
|
||||
Returns:
|
||||
iou (Tensor[N, M]): the NxM matrix containing the pairwise
|
||||
IoU values for every element in boxes1 and boxes2
|
||||
"""
|
||||
|
||||
def box_area(box):
|
||||
# box = 4xn
|
||||
return (box[2] - box[0]) * (box[3] - box[1])
|
||||
|
||||
area1 = box_area(box1.T)
|
||||
area2 = box_area(box2.T)
|
||||
|
||||
# inter(N,M) = (rb(N,M,2) - lt(N,M,2)).clamp(0).prod(2)
|
||||
inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2)
|
||||
return inter / (area1[:, None] + area2 - inter) # iou = inter / (area1 + area2 - inter)
|
||||
|
||||
|
||||
def wh_iou(wh1, wh2):
|
||||
# Returns the nxm IoU matrix. wh1 is nx2, wh2 is mx2
|
||||
wh1 = wh1[:, None] # [N,1,2]
|
||||
wh2 = wh2[None] # [1,M,2]
|
||||
inter = torch.min(wh1, wh2).prod(2) # [N,M]
|
||||
return inter / (wh1.prod(2) + wh2.prod(2) - inter) # iou = inter / (area1 + area2 - inter)
|
||||
|
||||
|
||||
def box_giou(box1, box2):
|
||||
"""
|
||||
Return generalized intersection-over-union (Jaccard index) between two sets of boxes.
|
||||
Both sets of boxes are expected to be in ``(x1, y1, x2, y2)`` format with
|
||||
``0 <= x1 < x2`` and ``0 <= y1 < y2``.
|
||||
Args:
|
||||
boxes1 (Tensor[N, 4]): first set of boxes
|
||||
boxes2 (Tensor[M, 4]): second set of boxes
|
||||
Returns:
|
||||
Tensor[N, M]: the NxM matrix containing the pairwise generalized IoU values
|
||||
for every element in boxes1 and boxes2
|
||||
"""
|
||||
|
||||
def box_area(box):
|
||||
# box = 4xn
|
||||
return (box[2] - box[0]) * (box[3] - box[1])
|
||||
|
||||
area1 = box_area(box1.T)
|
||||
area2 = box_area(box2.T)
|
||||
|
||||
inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2)
|
||||
union = (area1[:, None] + area2 - inter)
|
||||
|
||||
iou = inter / union
|
||||
|
||||
lti = torch.min(box1[:, None, :2], box2[:, :2])
|
||||
rbi = torch.max(box1[:, None, 2:], box2[:, 2:])
|
||||
|
||||
whi = (rbi - lti).clamp(min=0) # [N,M,2]
|
||||
areai = whi[:, :, 0] * whi[:, :, 1]
|
||||
|
||||
return iou - (areai - union) / areai
|
||||
|
||||
|
||||
def box_ciou(box1, box2, eps: float = 1e-7):
|
||||
"""
|
||||
Return complete intersection-over-union (Jaccard index) between two sets of boxes.
|
||||
Both sets of boxes are expected to be in ``(x1, y1, x2, y2)`` format with
|
||||
``0 <= x1 < x2`` and ``0 <= y1 < y2``.
|
||||
Args:
|
||||
boxes1 (Tensor[N, 4]): first set of boxes
|
||||
boxes2 (Tensor[M, 4]): second set of boxes
|
||||
eps (float, optional): small number to prevent division by zero. Default: 1e-7
|
||||
Returns:
|
||||
Tensor[N, M]: the NxM matrix containing the pairwise complete IoU values
|
||||
for every element in boxes1 and boxes2
|
||||
"""
|
||||
|
||||
def box_area(box):
|
||||
# box = 4xn
|
||||
return (box[2] - box[0]) * (box[3] - box[1])
|
||||
|
||||
area1 = box_area(box1.T)
|
||||
area2 = box_area(box2.T)
|
||||
|
||||
inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2)
|
||||
union = (area1[:, None] + area2 - inter)
|
||||
|
||||
iou = inter / union
|
||||
|
||||
lti = torch.min(box1[:, None, :2], box2[:, :2])
|
||||
rbi = torch.max(box1[:, None, 2:], box2[:, 2:])
|
||||
|
||||
whi = (rbi - lti).clamp(min=0) # [N,M,2]
|
||||
diagonal_distance_squared = (whi[:, :, 0] ** 2) + (whi[:, :, 1] ** 2) + eps
|
||||
|
||||
# centers of boxes
|
||||
x_p = (box1[:, None, 0] + box1[:, None, 2]) / 2
|
||||
y_p = (box1[:, None, 1] + box1[:, None, 3]) / 2
|
||||
x_g = (box2[:, 0] + box2[:, 2]) / 2
|
||||
y_g = (box2[:, 1] + box2[:, 3]) / 2
|
||||
# The distance between boxes' centers squared.
|
||||
centers_distance_squared = (x_p - x_g) ** 2 + (y_p - y_g) ** 2
|
||||
|
||||
w_pred = box1[:, None, 2] - box1[:, None, 0]
|
||||
h_pred = box1[:, None, 3] - box1[:, None, 1]
|
||||
|
||||
w_gt = box2[:, 2] - box2[:, 0]
|
||||
h_gt = box2[:, 3] - box2[:, 1]
|
||||
|
||||
v = (4 / (torch.pi ** 2)) * torch.pow((torch.atan(w_gt / h_gt) - torch.atan(w_pred / h_pred)), 2)
|
||||
with torch.no_grad():
|
||||
alpha = v / (1 - iou + v + eps)
|
||||
return iou - (centers_distance_squared / diagonal_distance_squared) - alpha * v
|
||||
|
||||
|
||||
def box_diou(box1, box2, eps: float = 1e-7):
|
||||
"""
|
||||
Return distance intersection-over-union (Jaccard index) between two sets of boxes.
|
||||
Both sets of boxes are expected to be in ``(x1, y1, x2, y2)`` format with
|
||||
``0 <= x1 < x2`` and ``0 <= y1 < y2``.
|
||||
Args:
|
||||
boxes1 (Tensor[N, 4]): first set of boxes
|
||||
boxes2 (Tensor[M, 4]): second set of boxes
|
||||
eps (float, optional): small number to prevent division by zero. Default: 1e-7
|
||||
Returns:
|
||||
Tensor[N, M]: the NxM matrix containing the pairwise distance IoU values
|
||||
for every element in boxes1 and boxes2
|
||||
"""
|
||||
|
||||
def box_area(box):
|
||||
# box = 4xn
|
||||
return (box[2] - box[0]) * (box[3] - box[1])
|
||||
|
||||
area1 = box_area(box1.T)
|
||||
area2 = box_area(box2.T)
|
||||
|
||||
inter = (torch.min(box1[:, None, 2:], box2[:, 2:]) - torch.max(box1[:, None, :2], box2[:, :2])).clamp(0).prod(2)
|
||||
union = (area1[:, None] + area2 - inter)
|
||||
|
||||
iou = inter / union
|
||||
|
||||
lti = torch.min(box1[:, None, :2], box2[:, :2])
|
||||
rbi = torch.max(box1[:, None, 2:], box2[:, 2:])
|
||||
|
||||
whi = (rbi - lti).clamp(min=0) # [N,M,2]
|
||||
diagonal_distance_squared = (whi[:, :, 0] ** 2) + (whi[:, :, 1] ** 2) + eps
|
||||
|
||||
# centers of boxes
|
||||
x_p = (box1[:, None, 0] + box1[:, None, 2]) / 2
|
||||
y_p = (box1[:, None, 1] + box1[:, None, 3]) / 2
|
||||
x_g = (box2[:, 0] + box2[:, 2]) / 2
|
||||
y_g = (box2[:, 1] + box2[:, 3]) / 2
|
||||
# The distance between boxes' centers squared.
|
||||
centers_distance_squared = (x_p - x_g) ** 2 + (y_p - y_g) ** 2
|
||||
|
||||
# The distance IoU is the IoU penalized by a normalized
|
||||
# distance between boxes' centers squared.
|
||||
return iou - (centers_distance_squared / diagonal_distance_squared)
|
||||
|
||||
|
||||
def non_max_suppression(prediction, conf_thres=0.25, iou_thres=0.45, classes=None, agnostic=False, multi_label=False,
|
||||
labels=()):
|
||||
"""Runs Non-Maximum Suppression (NMS) on inference results
|
||||
|
||||
Returns:
|
||||
list of detections, on (n,6) tensor per image [xyxy, conf, cls]
|
||||
"""
|
||||
|
||||
nc = prediction.shape[2] - 5 # number of classes
|
||||
xc = prediction[..., 4] > conf_thres # candidates
|
||||
|
||||
# Settings
|
||||
min_wh, max_wh = 2, 4096 # (pixels) minimum and maximum box width and height
|
||||
max_det = 300 # maximum number of detections per image
|
||||
max_nms = 30000 # maximum number of boxes into torchvision.ops.nms()
|
||||
time_limit = 10.0 # seconds to quit after
|
||||
redundant = True # require redundant detections
|
||||
multi_label &= nc > 1 # multiple labels per box (adds 0.5ms/img)
|
||||
merge = False # use merge-NMS
|
||||
|
||||
t = time.time()
|
||||
output = [torch.zeros((0, 6), device=prediction.device)] * prediction.shape[0]
|
||||
for xi, x in enumerate(prediction): # image index, image inference
|
||||
# Apply constraints
|
||||
# x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0 # width-height
|
||||
x = x[xc[xi]] # confidence
|
||||
|
||||
# Cat apriori labels if autolabelling
|
||||
if labels and len(labels[xi]):
|
||||
l = labels[xi]
|
||||
v = torch.zeros((len(l), nc + 5), device=x.device)
|
||||
v[:, :4] = l[:, 1:5] # box
|
||||
v[:, 4] = 1.0 # conf
|
||||
v[range(len(l)), l[:, 0].long() + 5] = 1.0 # cls
|
||||
x = torch.cat((x, v), 0)
|
||||
|
||||
# If none remain process next image
|
||||
if not x.shape[0]:
|
||||
continue
|
||||
|
||||
# Compute conf
|
||||
if nc == 1:
|
||||
x[:, 5:] = x[:, 4:5] # for models with one class, cls_loss is 0 and cls_conf is always 0.5,
|
||||
# so there is no need to multiplicate.
|
||||
else:
|
||||
x[:, 5:] *= x[:, 4:5] # conf = obj_conf * cls_conf
|
||||
|
||||
# Box (center x, center y, width, height) to (x1, y1, x2, y2)
|
||||
box = xywh2xyxy(x[:, :4])
|
||||
|
||||
# Detections matrix nx6 (xyxy, conf, cls)
|
||||
if multi_label:
|
||||
i, j = (x[:, 5:] > conf_thres).nonzero(as_tuple=False).T
|
||||
x = torch.cat((box[i], x[i, j + 5, None], j[:, None].float()), 1)
|
||||
else: # best class only
|
||||
conf, j = x[:, 5:].max(1, keepdim=True)
|
||||
x = torch.cat((box, conf, j.float()), 1)[conf.view(-1) > conf_thres]
|
||||
|
||||
# Filter by class
|
||||
if classes is not None:
|
||||
x = x[(x[:, 5:6] == torch.tensor(classes, device=x.device)).any(1)]
|
||||
|
||||
# Apply finite constraint
|
||||
# if not torch.isfinite(x).all():
|
||||
# x = x[torch.isfinite(x).all(1)]
|
||||
|
||||
# Check shape
|
||||
n = x.shape[0] # number of boxes
|
||||
if not n: # no boxes
|
||||
continue
|
||||
elif n > max_nms: # excess boxes
|
||||
x = x[x[:, 4].argsort(descending=True)[:max_nms]] # sort by confidence
|
||||
|
||||
# Batched NMS
|
||||
c = x[:, 5:6] * (0 if agnostic else max_wh) # classes
|
||||
boxes, scores = x[:, :4] + c, x[:, 4] # boxes (offset by class), scores
|
||||
i = torchvision.ops.nms(boxes, scores, iou_thres) # NMS
|
||||
if i.shape[0] > max_det: # limit detections
|
||||
i = i[:max_det]
|
||||
if merge and (1 < n < 3E3): # Merge NMS (boxes merged using weighted mean)
|
||||
# update boxes as boxes(i,4) = weights(i,n) * boxes(n,4)
|
||||
iou = box_iou(boxes[i], boxes) > iou_thres # iou matrix
|
||||
weights = iou * scores[None] # box weights
|
||||
x[i, :4] = torch.mm(weights, x[:, :4]).float() / weights.sum(1, keepdim=True) # merged boxes
|
||||
if redundant:
|
||||
i = i[iou.sum(1) > 1] # require redundancy
|
||||
|
||||
output[xi] = x[i]
|
||||
if (time.time() - t) > time_limit:
|
||||
print(f'WARNING: NMS time limit {time_limit}s exceeded')
|
||||
break # time limit exceeded
|
||||
|
||||
return output
|
||||
|
||||
|
||||
def non_max_suppression_kpt(prediction, conf_thres=0.25, iou_thres=0.45, classes=None, agnostic=False, multi_label=False,
|
||||
labels=(), kpt_label=False, nc=None, nkpt=None):
|
||||
"""Runs Non-Maximum Suppression (NMS) on inference results
|
||||
|
||||
Returns:
|
||||
list of detections, on (n,6) tensor per image [xyxy, conf, cls]
|
||||
"""
|
||||
if nc is None:
|
||||
nc = prediction.shape[2] - 5 if not kpt_label else prediction.shape[2] - 56 # number of classes
|
||||
xc = prediction[..., 4] > conf_thres # candidates
|
||||
|
||||
# Settings
|
||||
min_wh, max_wh = 2, 4096 # (pixels) minimum and maximum box width and height
|
||||
max_det = 300 # maximum number of detections per image
|
||||
max_nms = 30000 # maximum number of boxes into torchvision.ops.nms()
|
||||
time_limit = 10.0 # seconds to quit after
|
||||
redundant = True # require redundant detections
|
||||
multi_label &= nc > 1 # multiple labels per box (adds 0.5ms/img)
|
||||
merge = False # use merge-NMS
|
||||
|
||||
t = time.time()
|
||||
output = [torch.zeros((0,6), device=prediction.device)] * prediction.shape[0]
|
||||
for xi, x in enumerate(prediction): # image index, image inference
|
||||
# Apply constraints
|
||||
# x[((x[..., 2:4] < min_wh) | (x[..., 2:4] > max_wh)).any(1), 4] = 0 # width-height
|
||||
x = x[xc[xi]] # confidence
|
||||
|
||||
# Cat apriori labels if autolabelling
|
||||
if labels and len(labels[xi]):
|
||||
l = labels[xi]
|
||||
v = torch.zeros((len(l), nc + 5), device=x.device)
|
||||
v[:, :4] = l[:, 1:5] # box
|
||||
v[:, 4] = 1.0 # conf
|
||||
v[range(len(l)), l[:, 0].long() + 5] = 1.0 # cls
|
||||
x = torch.cat((x, v), 0)
|
||||
|
||||
# If none remain process next image
|
||||
if not x.shape[0]:
|
||||
continue
|
||||
|
||||
# Compute conf
|
||||
x[:, 5:5+nc] *= x[:, 4:5] # conf = obj_conf * cls_conf
|
||||
|
||||
# Box (center x, center y, width, height) to (x1, y1, x2, y2)
|
||||
box = xywh2xyxy(x[:, :4])
|
||||
|
||||
# Detections matrix nx6 (xyxy, conf, cls)
|
||||
if multi_label:
|
||||
i, j = (x[:, 5:] > conf_thres).nonzero(as_tuple=False).T
|
||||
x = torch.cat((box[i], x[i, j + 5, None], j[:, None].float()), 1)
|
||||
else: # best class only
|
||||
if not kpt_label:
|
||||
conf, j = x[:, 5:].max(1, keepdim=True)
|
||||
x = torch.cat((box, conf, j.float()), 1)[conf.view(-1) > conf_thres]
|
||||
else:
|
||||
kpts = x[:, 6:]
|
||||
conf, j = x[:, 5:6].max(1, keepdim=True)
|
||||
x = torch.cat((box, conf, j.float(), kpts), 1)[conf.view(-1) > conf_thres]
|
||||
|
||||
|
||||
# Filter by class
|
||||
if classes is not None:
|
||||
x = x[(x[:, 5:6] == torch.tensor(classes, device=x.device)).any(1)]
|
||||
|
||||
# Apply finite constraint
|
||||
# if not torch.isfinite(x).all():
|
||||
# x = x[torch.isfinite(x).all(1)]
|
||||
|
||||
# Check shape
|
||||
n = x.shape[0] # number of boxes
|
||||
if not n: # no boxes
|
||||
continue
|
||||
elif n > max_nms: # excess boxes
|
||||
x = x[x[:, 4].argsort(descending=True)[:max_nms]] # sort by confidence
|
||||
|
||||
# Batched NMS
|
||||
c = x[:, 5:6] * (0 if agnostic else max_wh) # classes
|
||||
boxes, scores = x[:, :4] + c, x[:, 4] # boxes (offset by class), scores
|
||||
i = torchvision.ops.nms(boxes, scores, iou_thres) # NMS
|
||||
if i.shape[0] > max_det: # limit detections
|
||||
i = i[:max_det]
|
||||
if merge and (1 < n < 3E3): # Merge NMS (boxes merged using weighted mean)
|
||||
# update boxes as boxes(i,4) = weights(i,n) * boxes(n,4)
|
||||
iou = box_iou(boxes[i], boxes) > iou_thres # iou matrix
|
||||
weights = iou * scores[None] # box weights
|
||||
x[i, :4] = torch.mm(weights, x[:, :4]).float() / weights.sum(1, keepdim=True) # merged boxes
|
||||
if redundant:
|
||||
i = i[iou.sum(1) > 1] # require redundancy
|
||||
|
||||
output[xi] = x[i]
|
||||
if (time.time() - t) > time_limit:
|
||||
print(f'WARNING: NMS time limit {time_limit}s exceeded')
|
||||
break # time limit exceeded
|
||||
|
||||
return output
|
||||
|
||||
|
||||
def strip_optimizer(f='best.pt', s=''): # from utils.general import *; strip_optimizer()
|
||||
# Strip optimizer from 'f' to finalize training, optionally save as 's'
|
||||
x = torch.load(f, map_location=torch.device('cpu'))
|
||||
if x.get('ema'):
|
||||
x['model'] = x['ema'] # replace model with ema
|
||||
for k in 'optimizer', 'training_results', 'wandb_id', 'ema', 'updates': # keys
|
||||
x[k] = None
|
||||
x['epoch'] = -1
|
||||
x['model'].half() # to FP16
|
||||
for p in x['model'].parameters():
|
||||
p.requires_grad = False
|
||||
torch.save(x, s or f)
|
||||
mb = os.path.getsize(s or f) / 1E6 # filesize
|
||||
print(f"Optimizer stripped from {f},{(' saved as %s,' % s) if s else ''} {mb:.1f}MB")
|
||||
|
||||
|
||||
def print_mutation(hyp, results, yaml_file='hyp_evolved.yaml', bucket=''):
|
||||
# Print mutation results to evolve.txt (for use with train.py --evolve)
|
||||
a = '%10s' * len(hyp) % tuple(hyp.keys()) # hyperparam keys
|
||||
b = '%10.3g' * len(hyp) % tuple(hyp.values()) # hyperparam values
|
||||
c = '%10.4g' * len(results) % results # results (P, R, mAP@0.5, mAP@0.5:0.95, val_losses x 3)
|
||||
print('\n%s\n%s\nEvolved fitness: %s\n' % (a, b, c))
|
||||
|
||||
if bucket:
|
||||
url = 'gs://%s/evolve.txt' % bucket
|
||||
if gsutil_getsize(url) > (os.path.getsize('evolve.txt') if os.path.exists('evolve.txt') else 0):
|
||||
os.system('gsutil cp %s .' % url) # download evolve.txt if larger than local
|
||||
|
||||
with open('evolve.txt', 'a') as f: # append result
|
||||
f.write(c + b + '\n')
|
||||
x = np.unique(np.loadtxt('evolve.txt', ndmin=2), axis=0) # load unique rows
|
||||
x = x[np.argsort(-fitness(x))] # sort
|
||||
np.savetxt('evolve.txt', x, '%10.3g') # save sort by fitness
|
||||
|
||||
# Save yaml
|
||||
for i, k in enumerate(hyp.keys()):
|
||||
hyp[k] = float(x[0, i + 7])
|
||||
with open(yaml_file, 'w') as f:
|
||||
results = tuple(x[0, :7])
|
||||
c = '%10.4g' * len(results) % results # results (P, R, mAP@0.5, mAP@0.5:0.95, val_losses x 3)
|
||||
f.write('# Hyperparameter Evolution Results\n# Generations: %g\n# Metrics: ' % len(x) + c + '\n\n')
|
||||
yaml.dump(hyp, f, sort_keys=False)
|
||||
|
||||
if bucket:
|
||||
os.system('gsutil cp evolve.txt %s gs://%s' % (yaml_file, bucket)) # upload
|
||||
|
||||
|
||||
def apply_classifier(x, model, img, im0):
|
||||
# applies a second stage classifier to yolo outputs
|
||||
im0 = [im0] if isinstance(im0, np.ndarray) else im0
|
||||
for i, d in enumerate(x): # per image
|
||||
if d is not None and len(d):
|
||||
d = d.clone()
|
||||
|
||||
# Reshape and pad cutouts
|
||||
b = xyxy2xywh(d[:, :4]) # boxes
|
||||
b[:, 2:] = b[:, 2:].max(1)[0].unsqueeze(1) # rectangle to square
|
||||
b[:, 2:] = b[:, 2:] * 1.3 + 30 # pad
|
||||
d[:, :4] = xywh2xyxy(b).long()
|
||||
|
||||
# Rescale boxes from img_size to im0 size
|
||||
scale_coords(img.shape[2:], d[:, :4], im0[i].shape)
|
||||
|
||||
# Classes
|
||||
pred_cls1 = d[:, 5].long()
|
||||
ims = []
|
||||
for j, a in enumerate(d): # per item
|
||||
cutout = im0[i][int(a[1]):int(a[3]), int(a[0]):int(a[2])]
|
||||
im = cv2.resize(cutout, (224, 224)) # BGR
|
||||
# cv2.imwrite('test%i.jpg' % j, cutout)
|
||||
|
||||
im = im[:, :, ::-1].transpose(2, 0, 1) # BGR to RGB, to 3x416x416
|
||||
im = np.ascontiguousarray(im, dtype=np.float32) # uint8 to float32
|
||||
im /= 255.0 # 0 - 255 to 0.0 - 1.0
|
||||
ims.append(im)
|
||||
|
||||
pred_cls2 = model(torch.Tensor(ims).to(d.device)).argmax(1) # classifier prediction
|
||||
x[i] = x[i][pred_cls1 == pred_cls2] # retain matching class detections
|
||||
|
||||
return x
|
||||
|
||||
|
||||
def increment_path(path, exist_ok=True, sep=''):
|
||||
# Increment path, i.e. runs/exp --> runs/exp{sep}0, runs/exp{sep}1 etc.
|
||||
path = Path(path) # os-agnostic
|
||||
if (path.exists() and exist_ok) or (not path.exists()):
|
||||
return str(path)
|
||||
else:
|
||||
dirs = glob.glob(f"{path}{sep}*") # similar paths
|
||||
matches = [re.search(rf"%s{sep}(\d+)" % path.stem, d) for d in dirs]
|
||||
i = [int(m.groups()[0]) for m in matches if m] # indices
|
||||
n = max(i) + 1 if i else 2 # increment number
|
||||
return f"{path}{sep}{n}" # update path
|
||||
@@ -0,0 +1,25 @@
|
||||
FROM gcr.io/google-appengine/python
|
||||
|
||||
# Create a virtualenv for dependencies. This isolates these packages from
|
||||
# system-level packages.
|
||||
# Use -p python3 or -p python3.7 to select python version. Default is version 2.
|
||||
RUN virtualenv /env -p python3
|
||||
|
||||
# Setting these environment variables are the same as running
|
||||
# source /env/bin/activate.
|
||||
ENV VIRTUAL_ENV /env
|
||||
ENV PATH /env/bin:$PATH
|
||||
|
||||
RUN apt-get update && apt-get install -y python-opencv
|
||||
|
||||
# Copy the application's requirements.txt and run pip to install all
|
||||
# dependencies into the virtualenv.
|
||||
ADD requirements.txt /app/requirements.txt
|
||||
RUN pip install -r /app/requirements.txt
|
||||
|
||||
# Add the application source code.
|
||||
ADD . /app
|
||||
|
||||
# Run a WSGI server to serve the application. gunicorn must be declared as
|
||||
# a dependency in requirements.txt.
|
||||
CMD gunicorn -b :$PORT main:app
|
||||
@@ -0,0 +1,4 @@
|
||||
# add these requirements in your app on top of the existing ones
|
||||
pip==18.1
|
||||
Flask==1.0.2
|
||||
gunicorn==19.9.0
|
||||
@@ -0,0 +1,14 @@
|
||||
runtime: custom
|
||||
env: flex
|
||||
|
||||
service: yolorapp
|
||||
|
||||
liveness_check:
|
||||
initial_delay_sec: 600
|
||||
|
||||
manual_scaling:
|
||||
instances: 1
|
||||
resources:
|
||||
cpu: 1
|
||||
memory_gb: 4
|
||||
disk_size_gb: 20
|
||||
@@ -0,0 +1,123 @@
|
||||
# Google utils: https://cloud.google.com/storage/docs/reference/libraries
|
||||
|
||||
import os
|
||||
import platform
|
||||
import subprocess
|
||||
import time
|
||||
from pathlib import Path
|
||||
|
||||
import requests
|
||||
import torch
|
||||
|
||||
|
||||
def gsutil_getsize(url=''):
|
||||
# gs://bucket/file size https://cloud.google.com/storage/docs/gsutil/commands/du
|
||||
s = subprocess.check_output(f'gsutil du {url}', shell=True).decode('utf-8')
|
||||
return eval(s.split(' ')[0]) if len(s) else 0 # bytes
|
||||
|
||||
|
||||
def attempt_download(file, repo='WongKinYiu/yolov7'):
|
||||
# Attempt file download if does not exist
|
||||
file = Path(str(file).strip().replace("'", '').lower())
|
||||
|
||||
if not file.exists():
|
||||
try:
|
||||
response = requests.get(f'https://api.github.com/repos/{repo}/releases/latest').json() # github api
|
||||
assets = [x['name'] for x in response['assets']] # release assets
|
||||
tag = response['tag_name'] # i.e. 'v1.0'
|
||||
except: # fallback plan
|
||||
assets = ['yolov7.pt', 'yolov7-tiny.pt', 'yolov7x.pt', 'yolov7-d6.pt', 'yolov7-e6.pt',
|
||||
'yolov7-e6e.pt', 'yolov7-w6.pt']
|
||||
tag = subprocess.check_output('git tag', shell=True).decode().split()[-1]
|
||||
|
||||
name = file.name
|
||||
if name in assets:
|
||||
msg = f'{file} missing, try downloading from https://github.com/{repo}/releases/'
|
||||
redundant = False # second download option
|
||||
try: # GitHub
|
||||
url = f'https://github.com/{repo}/releases/download/{tag}/{name}'
|
||||
print(f'Downloading {url} to {file}...')
|
||||
torch.hub.download_url_to_file(url, file)
|
||||
assert file.exists() and file.stat().st_size > 1E6 # check
|
||||
except Exception as e: # GCP
|
||||
print(f'Download error: {e}')
|
||||
assert redundant, 'No secondary mirror'
|
||||
url = f'https://storage.googleapis.com/{repo}/ckpt/{name}'
|
||||
print(f'Downloading {url} to {file}...')
|
||||
os.system(f'curl -L {url} -o {file}') # torch.hub.download_url_to_file(url, weights)
|
||||
finally:
|
||||
if not file.exists() or file.stat().st_size < 1E6: # check
|
||||
file.unlink(missing_ok=True) # remove partial downloads
|
||||
print(f'ERROR: Download failure: {msg}')
|
||||
print('')
|
||||
return
|
||||
|
||||
|
||||
def gdrive_download(id='', file='tmp.zip'):
|
||||
# Downloads a file from Google Drive. from yolov7.utils.google_utils import *; gdrive_download()
|
||||
t = time.time()
|
||||
file = Path(file)
|
||||
cookie = Path('cookie') # gdrive cookie
|
||||
print(f'Downloading https://drive.google.com/uc?export=download&id={id} as {file}... ', end='')
|
||||
file.unlink(missing_ok=True) # remove existing file
|
||||
cookie.unlink(missing_ok=True) # remove existing cookie
|
||||
|
||||
# Attempt file download
|
||||
out = "NUL" if platform.system() == "Windows" else "/dev/null"
|
||||
os.system(f'curl -c ./cookie -s -L "drive.google.com/uc?export=download&id={id}" > {out}')
|
||||
if os.path.exists('cookie'): # large file
|
||||
s = f'curl -Lb ./cookie "drive.google.com/uc?export=download&confirm={get_token()}&id={id}" -o {file}'
|
||||
else: # small file
|
||||
s = f'curl -s -L -o {file} "drive.google.com/uc?export=download&id={id}"'
|
||||
r = os.system(s) # execute, capture return
|
||||
cookie.unlink(missing_ok=True) # remove existing cookie
|
||||
|
||||
# Error check
|
||||
if r != 0:
|
||||
file.unlink(missing_ok=True) # remove partial
|
||||
print('Download error ') # raise Exception('Download error')
|
||||
return r
|
||||
|
||||
# Unzip if archive
|
||||
if file.suffix == '.zip':
|
||||
print('unzipping... ', end='')
|
||||
os.system(f'unzip -q {file}') # unzip
|
||||
file.unlink() # remove zip to free space
|
||||
|
||||
print(f'Done ({time.time() - t:.1f}s)')
|
||||
return r
|
||||
|
||||
|
||||
def get_token(cookie="./cookie"):
|
||||
with open(cookie) as f:
|
||||
for line in f:
|
||||
if "download" in line:
|
||||
return line.split()[-1]
|
||||
return ""
|
||||
|
||||
# def upload_blob(bucket_name, source_file_name, destination_blob_name):
|
||||
# # Uploads a file to a bucket
|
||||
# # https://cloud.google.com/storage/docs/uploading-objects#storage-upload-object-python
|
||||
#
|
||||
# storage_client = storage.Client()
|
||||
# bucket = storage_client.get_bucket(bucket_name)
|
||||
# blob = bucket.blob(destination_blob_name)
|
||||
#
|
||||
# blob.upload_from_filename(source_file_name)
|
||||
#
|
||||
# print('File {} uploaded to {}.'.format(
|
||||
# source_file_name,
|
||||
# destination_blob_name))
|
||||
#
|
||||
#
|
||||
# def download_blob(bucket_name, source_blob_name, destination_file_name):
|
||||
# # Uploads a blob from a bucket
|
||||
# storage_client = storage.Client()
|
||||
# bucket = storage_client.get_bucket(bucket_name)
|
||||
# blob = bucket.blob(source_blob_name)
|
||||
#
|
||||
# blob.download_to_filename(destination_file_name)
|
||||
#
|
||||
# print('Blob {} downloaded to {}.'.format(
|
||||
# source_blob_name,
|
||||
# destination_file_name))
|
||||
File diff suppressed because it is too large
Load Diff
@@ -0,0 +1,227 @@
|
||||
# Model validation metrics
|
||||
|
||||
from pathlib import Path
|
||||
|
||||
import matplotlib.pyplot as plt
|
||||
import numpy as np
|
||||
import torch
|
||||
|
||||
from . import general
|
||||
|
||||
|
||||
def fitness(x):
|
||||
# Model fitness as a weighted combination of metrics
|
||||
w = [0.0, 0.0, 0.1, 0.9] # weights for [P, R, mAP@0.5, mAP@0.5:0.95]
|
||||
return (x[:, :4] * w).sum(1)
|
||||
|
||||
|
||||
def ap_per_class(tp, conf, pred_cls, target_cls, v5_metric=False, plot=False, save_dir='.', names=()):
|
||||
""" Compute the average precision, given the recall and precision curves.
|
||||
Source: https://github.com/rafaelpadilla/Object-Detection-Metrics.
|
||||
# Arguments
|
||||
tp: True positives (nparray, nx1 or nx10).
|
||||
conf: Objectness value from 0-1 (nparray).
|
||||
pred_cls: Predicted object classes (nparray).
|
||||
target_cls: True object classes (nparray).
|
||||
plot: Plot precision-recall curve at mAP@0.5
|
||||
save_dir: Plot save directory
|
||||
# Returns
|
||||
The average precision as computed in py-faster-rcnn.
|
||||
"""
|
||||
|
||||
# Sort by objectness
|
||||
i = np.argsort(-conf)
|
||||
tp, conf, pred_cls = tp[i], conf[i], pred_cls[i]
|
||||
|
||||
# Find unique classes
|
||||
unique_classes = np.unique(target_cls)
|
||||
nc = unique_classes.shape[0] # number of classes, number of detections
|
||||
|
||||
# Create Precision-Recall curve and compute AP for each class
|
||||
px, py = np.linspace(0, 1, 1000), [] # for plotting
|
||||
ap, p, r = np.zeros((nc, tp.shape[1])), np.zeros((nc, 1000)), np.zeros((nc, 1000))
|
||||
for ci, c in enumerate(unique_classes):
|
||||
i = pred_cls == c
|
||||
n_l = (target_cls == c).sum() # number of labels
|
||||
n_p = i.sum() # number of predictions
|
||||
|
||||
if n_p == 0 or n_l == 0:
|
||||
continue
|
||||
else:
|
||||
# Accumulate FPs and TPs
|
||||
fpc = (1 - tp[i]).cumsum(0)
|
||||
tpc = tp[i].cumsum(0)
|
||||
|
||||
# Recall
|
||||
recall = tpc / (n_l + 1e-16) # recall curve
|
||||
r[ci] = np.interp(-px, -conf[i], recall[:, 0], left=0) # negative x, xp because xp decreases
|
||||
|
||||
# Precision
|
||||
precision = tpc / (tpc + fpc) # precision curve
|
||||
p[ci] = np.interp(-px, -conf[i], precision[:, 0], left=1) # p at pr_score
|
||||
|
||||
# AP from recall-precision curve
|
||||
for j in range(tp.shape[1]):
|
||||
ap[ci, j], mpre, mrec = compute_ap(recall[:, j], precision[:, j], v5_metric=v5_metric)
|
||||
if plot and j == 0:
|
||||
py.append(np.interp(px, mrec, mpre)) # precision at mAP@0.5
|
||||
|
||||
# Compute F1 (harmonic mean of precision and recall)
|
||||
f1 = 2 * p * r / (p + r + 1e-16)
|
||||
if plot:
|
||||
plot_pr_curve(px, py, ap, Path(save_dir) / 'PR_curve.png', names)
|
||||
plot_mc_curve(px, f1, Path(save_dir) / 'F1_curve.png', names, ylabel='F1')
|
||||
plot_mc_curve(px, p, Path(save_dir) / 'P_curve.png', names, ylabel='Precision')
|
||||
plot_mc_curve(px, r, Path(save_dir) / 'R_curve.png', names, ylabel='Recall')
|
||||
|
||||
i = f1.mean(0).argmax() # max F1 index
|
||||
return p[:, i], r[:, i], ap, f1[:, i], unique_classes.astype('int32')
|
||||
|
||||
|
||||
def compute_ap(recall, precision, v5_metric=False):
|
||||
""" Compute the average precision, given the recall and precision curves
|
||||
# Arguments
|
||||
recall: The recall curve (list)
|
||||
precision: The precision curve (list)
|
||||
v5_metric: Assume maximum recall to be 1.0, as in YOLOv5, MMDetetion etc.
|
||||
# Returns
|
||||
Average precision, precision curve, recall curve
|
||||
"""
|
||||
|
||||
# Append sentinel values to beginning and end
|
||||
if v5_metric: # New YOLOv5 metric, same as MMDetection and Detectron2 repositories
|
||||
mrec = np.concatenate(([0.], recall, [1.0]))
|
||||
else: # Old YOLOv5 metric, i.e. default YOLOv7 metric
|
||||
mrec = np.concatenate(([0.], recall, [recall[-1] + 0.01]))
|
||||
mpre = np.concatenate(([1.], precision, [0.]))
|
||||
|
||||
# Compute the precision envelope
|
||||
mpre = np.flip(np.maximum.accumulate(np.flip(mpre)))
|
||||
|
||||
# Integrate area under curve
|
||||
method = 'interp' # methods: 'continuous', 'interp'
|
||||
if method == 'interp':
|
||||
x = np.linspace(0, 1, 101) # 101-point interp (COCO)
|
||||
ap = np.trapz(np.interp(x, mrec, mpre), x) # integrate
|
||||
else: # 'continuous'
|
||||
i = np.where(mrec[1:] != mrec[:-1])[0] # points where x axis (recall) changes
|
||||
ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1]) # area under curve
|
||||
|
||||
return ap, mpre, mrec
|
||||
|
||||
|
||||
class ConfusionMatrix:
|
||||
# Updated version of https://github.com/kaanakan/object_detection_confusion_matrix
|
||||
def __init__(self, nc, conf=0.25, iou_thres=0.45):
|
||||
self.matrix = np.zeros((nc + 1, nc + 1))
|
||||
self.nc = nc # number of classes
|
||||
self.conf = conf
|
||||
self.iou_thres = iou_thres
|
||||
|
||||
def process_batch(self, detections, labels):
|
||||
"""
|
||||
Return intersection-over-union (Jaccard index) of boxes.
|
||||
Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
|
||||
Arguments:
|
||||
detections (Array[N, 6]), x1, y1, x2, y2, conf, class
|
||||
labels (Array[M, 5]), class, x1, y1, x2, y2
|
||||
Returns:
|
||||
None, updates confusion matrix accordingly
|
||||
"""
|
||||
detections = detections[detections[:, 4] > self.conf]
|
||||
gt_classes = labels[:, 0].int()
|
||||
detection_classes = detections[:, 5].int()
|
||||
iou = general.box_iou(labels[:, 1:], detections[:, :4])
|
||||
|
||||
x = torch.where(iou > self.iou_thres)
|
||||
if x[0].shape[0]:
|
||||
matches = torch.cat((torch.stack(x, 1), iou[x[0], x[1]][:, None]), 1).cpu().numpy()
|
||||
if x[0].shape[0] > 1:
|
||||
matches = matches[matches[:, 2].argsort()[::-1]]
|
||||
matches = matches[np.unique(matches[:, 1], return_index=True)[1]]
|
||||
matches = matches[matches[:, 2].argsort()[::-1]]
|
||||
matches = matches[np.unique(matches[:, 0], return_index=True)[1]]
|
||||
else:
|
||||
matches = np.zeros((0, 3))
|
||||
|
||||
n = matches.shape[0] > 0
|
||||
m0, m1, _ = matches.transpose().astype(np.int16)
|
||||
for i, gc in enumerate(gt_classes):
|
||||
j = m0 == i
|
||||
if n and sum(j) == 1:
|
||||
self.matrix[gc, detection_classes[m1[j]]] += 1 # correct
|
||||
else:
|
||||
self.matrix[self.nc, gc] += 1 # background FP
|
||||
|
||||
if n:
|
||||
for i, dc in enumerate(detection_classes):
|
||||
if not any(m1 == i):
|
||||
self.matrix[dc, self.nc] += 1 # background FN
|
||||
|
||||
def matrix(self):
|
||||
return self.matrix
|
||||
|
||||
def plot(self, save_dir='', names=()):
|
||||
try:
|
||||
import seaborn as sn
|
||||
|
||||
array = self.matrix / (self.matrix.sum(0).reshape(1, self.nc + 1) + 1E-6) # normalize
|
||||
array[array < 0.005] = np.nan # don't annotate (would appear as 0.00)
|
||||
|
||||
fig = plt.figure(figsize=(12, 9), tight_layout=True)
|
||||
sn.set(font_scale=1.0 if self.nc < 50 else 0.8) # for label size
|
||||
labels = (0 < len(names) < 99) and len(names) == self.nc # apply names to ticklabels
|
||||
sn.heatmap(array, annot=self.nc < 30, annot_kws={"size": 8}, cmap='Blues', fmt='.2f', square=True,
|
||||
xticklabels=names + ['background FP'] if labels else "auto",
|
||||
yticklabels=names + ['background FN'] if labels else "auto").set_facecolor((1, 1, 1))
|
||||
fig.axes[0].set_xlabel('True')
|
||||
fig.axes[0].set_ylabel('Predicted')
|
||||
fig.savefig(Path(save_dir) / 'confusion_matrix.png', dpi=250)
|
||||
except Exception as e:
|
||||
pass
|
||||
|
||||
def print(self):
|
||||
for i in range(self.nc + 1):
|
||||
print(' '.join(map(str, self.matrix[i])))
|
||||
|
||||
|
||||
# Plots ----------------------------------------------------------------------------------------------------------------
|
||||
|
||||
def plot_pr_curve(px, py, ap, save_dir='pr_curve.png', names=()):
|
||||
# Precision-recall curve
|
||||
fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True)
|
||||
py = np.stack(py, axis=1)
|
||||
|
||||
if 0 < len(names) < 21: # display per-class legend if < 21 classes
|
||||
for i, y in enumerate(py.T):
|
||||
ax.plot(px, y, linewidth=1, label=f'{names[i]} {ap[i, 0]:.3f}') # plot(recall, precision)
|
||||
else:
|
||||
ax.plot(px, py, linewidth=1, color='grey') # plot(recall, precision)
|
||||
|
||||
ax.plot(px, py.mean(1), linewidth=3, color='blue', label='all classes %.3f mAP@0.5' % ap[:, 0].mean())
|
||||
ax.set_xlabel('Recall')
|
||||
ax.set_ylabel('Precision')
|
||||
ax.set_xlim(0, 1)
|
||||
ax.set_ylim(0, 1)
|
||||
plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left")
|
||||
fig.savefig(Path(save_dir), dpi=250)
|
||||
|
||||
|
||||
def plot_mc_curve(px, py, save_dir='mc_curve.png', names=(), xlabel='Confidence', ylabel='Metric'):
|
||||
# Metric-confidence curve
|
||||
fig, ax = plt.subplots(1, 1, figsize=(9, 6), tight_layout=True)
|
||||
|
||||
if 0 < len(names) < 21: # display per-class legend if < 21 classes
|
||||
for i, y in enumerate(py):
|
||||
ax.plot(px, y, linewidth=1, label=f'{names[i]}') # plot(confidence, metric)
|
||||
else:
|
||||
ax.plot(px, py.T, linewidth=1, color='grey') # plot(confidence, metric)
|
||||
|
||||
y = py.mean(0)
|
||||
ax.plot(px, y, linewidth=3, color='blue', label=f'all classes {y.max():.2f} at {px[y.argmax()]:.3f}')
|
||||
ax.set_xlabel(xlabel)
|
||||
ax.set_ylabel(ylabel)
|
||||
ax.set_xlim(0, 1)
|
||||
ax.set_ylim(0, 1)
|
||||
plt.legend(bbox_to_anchor=(1.04, 1), loc="upper left")
|
||||
fig.savefig(Path(save_dir), dpi=250)
|
||||
@@ -0,0 +1,489 @@
|
||||
# Plotting utils
|
||||
|
||||
import glob
|
||||
import math
|
||||
import os
|
||||
import random
|
||||
from copy import copy
|
||||
from pathlib import Path
|
||||
|
||||
import cv2
|
||||
import matplotlib
|
||||
import matplotlib.pyplot as plt
|
||||
import numpy as np
|
||||
import pandas as pd
|
||||
import seaborn as sns
|
||||
import torch
|
||||
import yaml
|
||||
from PIL import Image, ImageDraw, ImageFont
|
||||
from scipy.signal import butter, filtfilt
|
||||
|
||||
from utils.general import xywh2xyxy, xyxy2xywh
|
||||
from utils.metrics import fitness
|
||||
|
||||
# Settings
|
||||
matplotlib.rc('font', **{'size': 11})
|
||||
matplotlib.use('Agg') # for writing to files only
|
||||
|
||||
|
||||
def color_list():
|
||||
# Return first 10 plt colors as (r,g,b) https://stackoverflow.com/questions/51350872/python-from-color-name-to-rgb
|
||||
def hex2rgb(h):
|
||||
return tuple(int(h[1 + i:1 + i + 2], 16) for i in (0, 2, 4))
|
||||
|
||||
return [hex2rgb(h) for h in matplotlib.colors.TABLEAU_COLORS.values()] # or BASE_ (8), CSS4_ (148), XKCD_ (949)
|
||||
|
||||
|
||||
def hist2d(x, y, n=100):
|
||||
# 2d histogram used in labels.png and evolve.png
|
||||
xedges, yedges = np.linspace(x.min(), x.max(), n), np.linspace(y.min(), y.max(), n)
|
||||
hist, xedges, yedges = np.histogram2d(x, y, (xedges, yedges))
|
||||
xidx = np.clip(np.digitize(x, xedges) - 1, 0, hist.shape[0] - 1)
|
||||
yidx = np.clip(np.digitize(y, yedges) - 1, 0, hist.shape[1] - 1)
|
||||
return np.log(hist[xidx, yidx])
|
||||
|
||||
|
||||
def butter_lowpass_filtfilt(data, cutoff=1500, fs=50000, order=5):
|
||||
# https://stackoverflow.com/questions/28536191/how-to-filter-smooth-with-scipy-numpy
|
||||
def butter_lowpass(cutoff, fs, order):
|
||||
nyq = 0.5 * fs
|
||||
normal_cutoff = cutoff / nyq
|
||||
return butter(order, normal_cutoff, btype='low', analog=False)
|
||||
|
||||
b, a = butter_lowpass(cutoff, fs, order=order)
|
||||
return filtfilt(b, a, data) # forward-backward filter
|
||||
|
||||
|
||||
def plot_one_box(x, img, color=None, label=None, line_thickness=3):
|
||||
# Plots one bounding box on image img
|
||||
tl = line_thickness or round(0.002 * (img.shape[0] + img.shape[1]) / 2) + 1 # line/font thickness
|
||||
color = color or [random.randint(0, 255) for _ in range(3)]
|
||||
c1, c2 = (int(x[0]), int(x[1])), (int(x[2]), int(x[3]))
|
||||
cv2.rectangle(img, c1, c2, color, thickness=tl, lineType=cv2.LINE_AA)
|
||||
if label:
|
||||
tf = max(tl - 1, 1) # font thickness
|
||||
t_size = cv2.getTextSize(label, 0, fontScale=tl / 3, thickness=tf)[0]
|
||||
c2 = c1[0] + t_size[0], c1[1] - t_size[1] - 3
|
||||
cv2.rectangle(img, c1, c2, color, -1, cv2.LINE_AA) # filled
|
||||
cv2.putText(img, label, (c1[0], c1[1] - 2), 0, tl / 3, [225, 255, 255], thickness=tf, lineType=cv2.LINE_AA)
|
||||
|
||||
|
||||
def plot_one_box_PIL(box, img, color=None, label=None, line_thickness=None):
|
||||
img = Image.fromarray(img)
|
||||
draw = ImageDraw.Draw(img)
|
||||
line_thickness = line_thickness or max(int(min(img.size) / 200), 2)
|
||||
draw.rectangle(box, width=line_thickness, outline=tuple(color)) # plot
|
||||
if label:
|
||||
fontsize = max(round(max(img.size) / 40), 12)
|
||||
font = ImageFont.truetype("Arial.ttf", fontsize)
|
||||
txt_width, txt_height = font.getsize(label)
|
||||
draw.rectangle([box[0], box[1] - txt_height + 4, box[0] + txt_width, box[1]], fill=tuple(color))
|
||||
draw.text((box[0], box[1] - txt_height + 1), label, fill=(255, 255, 255), font=font)
|
||||
return np.asarray(img)
|
||||
|
||||
|
||||
def plot_wh_methods(): # from utils.plots import *; plot_wh_methods()
|
||||
# Compares the two methods for width-height anchor multiplication
|
||||
# https://github.com/ultralytics/yolov3/issues/168
|
||||
x = np.arange(-4.0, 4.0, .1)
|
||||
ya = np.exp(x)
|
||||
yb = torch.sigmoid(torch.from_numpy(x)).numpy() * 2
|
||||
|
||||
fig = plt.figure(figsize=(6, 3), tight_layout=True)
|
||||
plt.plot(x, ya, '.-', label='YOLOv3')
|
||||
plt.plot(x, yb ** 2, '.-', label='YOLOR ^2')
|
||||
plt.plot(x, yb ** 1.6, '.-', label='YOLOR ^1.6')
|
||||
plt.xlim(left=-4, right=4)
|
||||
plt.ylim(bottom=0, top=6)
|
||||
plt.xlabel('input')
|
||||
plt.ylabel('output')
|
||||
plt.grid()
|
||||
plt.legend()
|
||||
fig.savefig('comparison.png', dpi=200)
|
||||
|
||||
|
||||
def output_to_target(output):
|
||||
# Convert model output to target format [batch_id, class_id, x, y, w, h, conf]
|
||||
targets = []
|
||||
for i, o in enumerate(output):
|
||||
for *box, conf, cls in o.cpu().numpy():
|
||||
targets.append([i, cls, *list(*xyxy2xywh(np.array(box)[None])), conf])
|
||||
return np.array(targets)
|
||||
|
||||
|
||||
def plot_images(images, targets, paths=None, fname='images.jpg', names=None, max_size=640, max_subplots=16):
|
||||
# Plot image grid with labels
|
||||
|
||||
if isinstance(images, torch.Tensor):
|
||||
images = images.cpu().float().numpy()
|
||||
if isinstance(targets, torch.Tensor):
|
||||
targets = targets.cpu().numpy()
|
||||
|
||||
# un-normalise
|
||||
if np.max(images[0]) <= 1:
|
||||
images *= 255
|
||||
|
||||
tl = 3 # line thickness
|
||||
tf = max(tl - 1, 1) # font thickness
|
||||
bs, _, h, w = images.shape # batch size, _, height, width
|
||||
bs = min(bs, max_subplots) # limit plot images
|
||||
ns = np.ceil(bs ** 0.5) # number of subplots (square)
|
||||
|
||||
# Check if we should resize
|
||||
scale_factor = max_size / max(h, w)
|
||||
if scale_factor < 1:
|
||||
h = math.ceil(scale_factor * h)
|
||||
w = math.ceil(scale_factor * w)
|
||||
|
||||
colors = color_list() # list of colors
|
||||
mosaic = np.full((int(ns * h), int(ns * w), 3), 255, dtype=np.uint8) # init
|
||||
for i, img in enumerate(images):
|
||||
if i == max_subplots: # if last batch has fewer images than we expect
|
||||
break
|
||||
|
||||
block_x = int(w * (i // ns))
|
||||
block_y = int(h * (i % ns))
|
||||
|
||||
img = img.transpose(1, 2, 0)
|
||||
if scale_factor < 1:
|
||||
img = cv2.resize(img, (w, h))
|
||||
|
||||
mosaic[block_y:block_y + h, block_x:block_x + w, :] = img
|
||||
if len(targets) > 0:
|
||||
image_targets = targets[targets[:, 0] == i]
|
||||
boxes = xywh2xyxy(image_targets[:, 2:6]).T
|
||||
classes = image_targets[:, 1].astype('int')
|
||||
labels = image_targets.shape[1] == 6 # labels if no conf column
|
||||
conf = None if labels else image_targets[:, 6] # check for confidence presence (label vs pred)
|
||||
|
||||
if boxes.shape[1]:
|
||||
if boxes.max() <= 1.01: # if normalized with tolerance 0.01
|
||||
boxes[[0, 2]] *= w # scale to pixels
|
||||
boxes[[1, 3]] *= h
|
||||
elif scale_factor < 1: # absolute coords need scale if image scales
|
||||
boxes *= scale_factor
|
||||
boxes[[0, 2]] += block_x
|
||||
boxes[[1, 3]] += block_y
|
||||
for j, box in enumerate(boxes.T):
|
||||
cls = int(classes[j])
|
||||
color = colors[cls % len(colors)]
|
||||
cls = names[cls] if names else cls
|
||||
if labels or conf[j] > 0.25: # 0.25 conf thresh
|
||||
label = '%s' % cls if labels else '%s %.1f' % (cls, conf[j])
|
||||
plot_one_box(box, mosaic, label=label, color=color, line_thickness=tl)
|
||||
|
||||
# Draw image filename labels
|
||||
if paths:
|
||||
label = Path(paths[i]).name[:40] # trim to 40 char
|
||||
t_size = cv2.getTextSize(label, 0, fontScale=tl / 3, thickness=tf)[0]
|
||||
cv2.putText(mosaic, label, (block_x + 5, block_y + t_size[1] + 5), 0, tl / 3, [220, 220, 220], thickness=tf,
|
||||
lineType=cv2.LINE_AA)
|
||||
|
||||
# Image border
|
||||
cv2.rectangle(mosaic, (block_x, block_y), (block_x + w, block_y + h), (255, 255, 255), thickness=3)
|
||||
|
||||
if fname:
|
||||
r = min(1280. / max(h, w) / ns, 1.0) # ratio to limit image size
|
||||
mosaic = cv2.resize(mosaic, (int(ns * w * r), int(ns * h * r)), interpolation=cv2.INTER_AREA)
|
||||
# cv2.imwrite(fname, cv2.cvtColor(mosaic, cv2.COLOR_BGR2RGB)) # cv2 save
|
||||
Image.fromarray(mosaic).save(fname) # PIL save
|
||||
return mosaic
|
||||
|
||||
|
||||
def plot_lr_scheduler(optimizer, scheduler, epochs=300, save_dir=''):
|
||||
# Plot LR simulating training for full epochs
|
||||
optimizer, scheduler = copy(optimizer), copy(scheduler) # do not modify originals
|
||||
y = []
|
||||
for _ in range(epochs):
|
||||
scheduler.step()
|
||||
y.append(optimizer.param_groups[0]['lr'])
|
||||
plt.plot(y, '.-', label='LR')
|
||||
plt.xlabel('epoch')
|
||||
plt.ylabel('LR')
|
||||
plt.grid()
|
||||
plt.xlim(0, epochs)
|
||||
plt.ylim(0)
|
||||
plt.savefig(Path(save_dir) / 'LR.png', dpi=200)
|
||||
plt.close()
|
||||
|
||||
|
||||
def plot_test_txt(): # from utils.plots import *; plot_test()
|
||||
# Plot test.txt histograms
|
||||
x = np.loadtxt('test.txt', dtype=np.float32)
|
||||
box = xyxy2xywh(x[:, :4])
|
||||
cx, cy = box[:, 0], box[:, 1]
|
||||
|
||||
fig, ax = plt.subplots(1, 1, figsize=(6, 6), tight_layout=True)
|
||||
ax.hist2d(cx, cy, bins=600, cmax=10, cmin=0)
|
||||
ax.set_aspect('equal')
|
||||
plt.savefig('hist2d.png', dpi=300)
|
||||
|
||||
fig, ax = plt.subplots(1, 2, figsize=(12, 6), tight_layout=True)
|
||||
ax[0].hist(cx, bins=600)
|
||||
ax[1].hist(cy, bins=600)
|
||||
plt.savefig('hist1d.png', dpi=200)
|
||||
|
||||
|
||||
def plot_targets_txt(): # from utils.plots import *; plot_targets_txt()
|
||||
# Plot targets.txt histograms
|
||||
x = np.loadtxt('targets.txt', dtype=np.float32).T
|
||||
s = ['x targets', 'y targets', 'width targets', 'height targets']
|
||||
fig, ax = plt.subplots(2, 2, figsize=(8, 8), tight_layout=True)
|
||||
ax = ax.ravel()
|
||||
for i in range(4):
|
||||
ax[i].hist(x[i], bins=100, label='%.3g +/- %.3g' % (x[i].mean(), x[i].std()))
|
||||
ax[i].legend()
|
||||
ax[i].set_title(s[i])
|
||||
plt.savefig('targets.jpg', dpi=200)
|
||||
|
||||
|
||||
def plot_study_txt(path='', x=None): # from utils.plots import *; plot_study_txt()
|
||||
# Plot study.txt generated by test.py
|
||||
fig, ax = plt.subplots(2, 4, figsize=(10, 6), tight_layout=True)
|
||||
# ax = ax.ravel()
|
||||
|
||||
fig2, ax2 = plt.subplots(1, 1, figsize=(8, 4), tight_layout=True)
|
||||
# for f in [Path(path) / f'study_coco_{x}.txt' for x in ['yolor-p6', 'yolor-w6', 'yolor-e6', 'yolor-d6']]:
|
||||
for f in sorted(Path(path).glob('study*.txt')):
|
||||
y = np.loadtxt(f, dtype=np.float32, usecols=[0, 1, 2, 3, 7, 8, 9], ndmin=2).T
|
||||
x = np.arange(y.shape[1]) if x is None else np.array(x)
|
||||
s = ['P', 'R', 'mAP@.5', 'mAP@.5:.95', 't_inference (ms/img)', 't_NMS (ms/img)', 't_total (ms/img)']
|
||||
# for i in range(7):
|
||||
# ax[i].plot(x, y[i], '.-', linewidth=2, markersize=8)
|
||||
# ax[i].set_title(s[i])
|
||||
|
||||
j = y[3].argmax() + 1
|
||||
ax2.plot(y[6, 1:j], y[3, 1:j] * 1E2, '.-', linewidth=2, markersize=8,
|
||||
label=f.stem.replace('study_coco_', '').replace('yolo', 'YOLO'))
|
||||
|
||||
ax2.plot(1E3 / np.array([209, 140, 97, 58, 35, 18]), [34.6, 40.5, 43.0, 47.5, 49.7, 51.5],
|
||||
'k.-', linewidth=2, markersize=8, alpha=.25, label='EfficientDet')
|
||||
|
||||
ax2.grid(alpha=0.2)
|
||||
ax2.set_yticks(np.arange(20, 60, 5))
|
||||
ax2.set_xlim(0, 57)
|
||||
ax2.set_ylim(30, 55)
|
||||
ax2.set_xlabel('GPU Speed (ms/img)')
|
||||
ax2.set_ylabel('COCO AP val')
|
||||
ax2.legend(loc='lower right')
|
||||
plt.savefig(str(Path(path).name) + '.png', dpi=300)
|
||||
|
||||
|
||||
def plot_labels(labels, names=(), save_dir=Path(''), loggers=None):
|
||||
# plot dataset labels
|
||||
print('Plotting labels... ')
|
||||
c, b = labels[:, 0], labels[:, 1:].transpose() # classes, boxes
|
||||
nc = int(c.max() + 1) # number of classes
|
||||
colors = color_list()
|
||||
x = pd.DataFrame(b.transpose(), columns=['x', 'y', 'width', 'height'])
|
||||
|
||||
# seaborn correlogram
|
||||
sns.pairplot(x, corner=True, diag_kind='auto', kind='hist', diag_kws=dict(bins=50), plot_kws=dict(pmax=0.9))
|
||||
plt.savefig(save_dir / 'labels_correlogram.jpg', dpi=200)
|
||||
plt.close()
|
||||
|
||||
# matplotlib labels
|
||||
matplotlib.use('svg') # faster
|
||||
ax = plt.subplots(2, 2, figsize=(8, 8), tight_layout=True)[1].ravel()
|
||||
ax[0].hist(c, bins=np.linspace(0, nc, nc + 1) - 0.5, rwidth=0.8)
|
||||
ax[0].set_ylabel('instances')
|
||||
if 0 < len(names) < 30:
|
||||
ax[0].set_xticks(range(len(names)))
|
||||
ax[0].set_xticklabels(names, rotation=90, fontsize=10)
|
||||
else:
|
||||
ax[0].set_xlabel('classes')
|
||||
sns.histplot(x, x='x', y='y', ax=ax[2], bins=50, pmax=0.9)
|
||||
sns.histplot(x, x='width', y='height', ax=ax[3], bins=50, pmax=0.9)
|
||||
|
||||
# rectangles
|
||||
labels[:, 1:3] = 0.5 # center
|
||||
labels[:, 1:] = xywh2xyxy(labels[:, 1:]) * 2000
|
||||
img = Image.fromarray(np.ones((2000, 2000, 3), dtype=np.uint8) * 255)
|
||||
for cls, *box in labels[:1000]:
|
||||
ImageDraw.Draw(img).rectangle(box, width=1, outline=colors[int(cls) % 10]) # plot
|
||||
ax[1].imshow(img)
|
||||
ax[1].axis('off')
|
||||
|
||||
for a in [0, 1, 2, 3]:
|
||||
for s in ['top', 'right', 'left', 'bottom']:
|
||||
ax[a].spines[s].set_visible(False)
|
||||
|
||||
plt.savefig(save_dir / 'labels.jpg', dpi=200)
|
||||
matplotlib.use('Agg')
|
||||
plt.close()
|
||||
|
||||
# loggers
|
||||
for k, v in loggers.items() or {}:
|
||||
if k == 'wandb' and v:
|
||||
v.log({"Labels": [v.Image(str(x), caption=x.name) for x in save_dir.glob('*labels*.jpg')]}, commit=False)
|
||||
|
||||
|
||||
def plot_evolution(yaml_file='data/hyp.finetune.yaml'): # from utils.plots import *; plot_evolution()
|
||||
# Plot hyperparameter evolution results in evolve.txt
|
||||
with open(yaml_file) as f:
|
||||
hyp = yaml.load(f, Loader=yaml.SafeLoader)
|
||||
x = np.loadtxt('evolve.txt', ndmin=2)
|
||||
f = fitness(x)
|
||||
# weights = (f - f.min()) ** 2 # for weighted results
|
||||
plt.figure(figsize=(10, 12), tight_layout=True)
|
||||
matplotlib.rc('font', **{'size': 8})
|
||||
for i, (k, v) in enumerate(hyp.items()):
|
||||
y = x[:, i + 7]
|
||||
# mu = (y * weights).sum() / weights.sum() # best weighted result
|
||||
mu = y[f.argmax()] # best single result
|
||||
plt.subplot(6, 5, i + 1)
|
||||
plt.scatter(y, f, c=hist2d(y, f, 20), cmap='viridis', alpha=.8, edgecolors='none')
|
||||
plt.plot(mu, f.max(), 'k+', markersize=15)
|
||||
plt.title('%s = %.3g' % (k, mu), fontdict={'size': 9}) # limit to 40 characters
|
||||
if i % 5 != 0:
|
||||
plt.yticks([])
|
||||
print('%15s: %.3g' % (k, mu))
|
||||
plt.savefig('evolve.png', dpi=200)
|
||||
print('\nPlot saved as evolve.png')
|
||||
|
||||
|
||||
def profile_idetection(start=0, stop=0, labels=(), save_dir=''):
|
||||
# Plot iDetection '*.txt' per-image logs. from utils.plots import *; profile_idetection()
|
||||
ax = plt.subplots(2, 4, figsize=(12, 6), tight_layout=True)[1].ravel()
|
||||
s = ['Images', 'Free Storage (GB)', 'RAM Usage (GB)', 'Battery', 'dt_raw (ms)', 'dt_smooth (ms)', 'real-world FPS']
|
||||
files = list(Path(save_dir).glob('frames*.txt'))
|
||||
for fi, f in enumerate(files):
|
||||
try:
|
||||
results = np.loadtxt(f, ndmin=2).T[:, 90:-30] # clip first and last rows
|
||||
n = results.shape[1] # number of rows
|
||||
x = np.arange(start, min(stop, n) if stop else n)
|
||||
results = results[:, x]
|
||||
t = (results[0] - results[0].min()) # set t0=0s
|
||||
results[0] = x
|
||||
for i, a in enumerate(ax):
|
||||
if i < len(results):
|
||||
label = labels[fi] if len(labels) else f.stem.replace('frames_', '')
|
||||
a.plot(t, results[i], marker='.', label=label, linewidth=1, markersize=5)
|
||||
a.set_title(s[i])
|
||||
a.set_xlabel('time (s)')
|
||||
# if fi == len(files) - 1:
|
||||
# a.set_ylim(bottom=0)
|
||||
for side in ['top', 'right']:
|
||||
a.spines[side].set_visible(False)
|
||||
else:
|
||||
a.remove()
|
||||
except Exception as e:
|
||||
print('Warning: Plotting error for %s; %s' % (f, e))
|
||||
|
||||
ax[1].legend()
|
||||
plt.savefig(Path(save_dir) / 'idetection_profile.png', dpi=200)
|
||||
|
||||
|
||||
def plot_results_overlay(start=0, stop=0): # from utils.plots import *; plot_results_overlay()
|
||||
# Plot training 'results*.txt', overlaying train and val losses
|
||||
s = ['train', 'train', 'train', 'Precision', 'mAP@0.5', 'val', 'val', 'val', 'Recall', 'mAP@0.5:0.95'] # legends
|
||||
t = ['Box', 'Objectness', 'Classification', 'P-R', 'mAP-F1'] # titles
|
||||
for f in sorted(glob.glob('results*.txt') + glob.glob('../../Downloads/results*.txt')):
|
||||
results = np.loadtxt(f, usecols=[2, 3, 4, 8, 9, 12, 13, 14, 10, 11], ndmin=2).T
|
||||
n = results.shape[1] # number of rows
|
||||
x = range(start, min(stop, n) if stop else n)
|
||||
fig, ax = plt.subplots(1, 5, figsize=(14, 3.5), tight_layout=True)
|
||||
ax = ax.ravel()
|
||||
for i in range(5):
|
||||
for j in [i, i + 5]:
|
||||
y = results[j, x]
|
||||
ax[i].plot(x, y, marker='.', label=s[j])
|
||||
# y_smooth = butter_lowpass_filtfilt(y)
|
||||
# ax[i].plot(x, np.gradient(y_smooth), marker='.', label=s[j])
|
||||
|
||||
ax[i].set_title(t[i])
|
||||
ax[i].legend()
|
||||
ax[i].set_ylabel(f) if i == 0 else None # add filename
|
||||
fig.savefig(f.replace('.txt', '.png'), dpi=200)
|
||||
|
||||
|
||||
def plot_results(start=0, stop=0, bucket='', id=(), labels=(), save_dir=''):
|
||||
# Plot training 'results*.txt'. from utils.plots import *; plot_results(save_dir='runs/train/exp')
|
||||
fig, ax = plt.subplots(2, 5, figsize=(12, 6), tight_layout=True)
|
||||
ax = ax.ravel()
|
||||
s = ['Box', 'Objectness', 'Classification', 'Precision', 'Recall',
|
||||
'val Box', 'val Objectness', 'val Classification', 'mAP@0.5', 'mAP@0.5:0.95']
|
||||
if bucket:
|
||||
# files = ['https://storage.googleapis.com/%s/results%g.txt' % (bucket, x) for x in id]
|
||||
files = ['results%g.txt' % x for x in id]
|
||||
c = ('gsutil cp ' + '%s ' * len(files) + '.') % tuple('gs://%s/results%g.txt' % (bucket, x) for x in id)
|
||||
os.system(c)
|
||||
else:
|
||||
files = list(Path(save_dir).glob('results*.txt'))
|
||||
assert len(files), 'No results.txt files found in %s, nothing to plot.' % os.path.abspath(save_dir)
|
||||
for fi, f in enumerate(files):
|
||||
try:
|
||||
results = np.loadtxt(f, usecols=[2, 3, 4, 8, 9, 12, 13, 14, 10, 11], ndmin=2).T
|
||||
n = results.shape[1] # number of rows
|
||||
x = range(start, min(stop, n) if stop else n)
|
||||
for i in range(10):
|
||||
y = results[i, x]
|
||||
if i in [0, 1, 2, 5, 6, 7]:
|
||||
y[y == 0] = np.nan # don't show zero loss values
|
||||
# y /= y[0] # normalize
|
||||
label = labels[fi] if len(labels) else f.stem
|
||||
ax[i].plot(x, y, marker='.', label=label, linewidth=2, markersize=8)
|
||||
ax[i].set_title(s[i])
|
||||
# if i in [5, 6, 7]: # share train and val loss y axes
|
||||
# ax[i].get_shared_y_axes().join(ax[i], ax[i - 5])
|
||||
except Exception as e:
|
||||
print('Warning: Plotting error for %s; %s' % (f, e))
|
||||
|
||||
ax[1].legend()
|
||||
fig.savefig(Path(save_dir) / 'results.png', dpi=200)
|
||||
|
||||
|
||||
def output_to_keypoint(output):
|
||||
# Convert model output to target format [batch_id, class_id, x, y, w, h, conf]
|
||||
targets = []
|
||||
for i, o in enumerate(output):
|
||||
kpts = o[:,6:]
|
||||
o = o[:,:6]
|
||||
for index, (*box, conf, cls) in enumerate(o.detach().cpu().numpy()):
|
||||
targets.append([i, cls, *list(*xyxy2xywh(np.array(box)[None])), conf, *list(kpts.detach().cpu().numpy()[index])])
|
||||
return np.array(targets)
|
||||
|
||||
|
||||
def plot_skeleton_kpts(im, kpts, steps, orig_shape=None):
|
||||
#Plot the skeleton and keypointsfor coco datatset
|
||||
palette = np.array([[255, 128, 0], [255, 153, 51], [255, 178, 102],
|
||||
[230, 230, 0], [255, 153, 255], [153, 204, 255],
|
||||
[255, 102, 255], [255, 51, 255], [102, 178, 255],
|
||||
[51, 153, 255], [255, 153, 153], [255, 102, 102],
|
||||
[255, 51, 51], [153, 255, 153], [102, 255, 102],
|
||||
[51, 255, 51], [0, 255, 0], [0, 0, 255], [255, 0, 0],
|
||||
[255, 255, 255]])
|
||||
|
||||
skeleton = [[16, 14], [14, 12], [17, 15], [15, 13], [12, 13], [6, 12],
|
||||
[7, 13], [6, 7], [6, 8], [7, 9], [8, 10], [9, 11], [2, 3],
|
||||
[1, 2], [1, 3], [2, 4], [3, 5], [4, 6], [5, 7]]
|
||||
|
||||
pose_limb_color = palette[[9, 9, 9, 9, 7, 7, 7, 0, 0, 0, 0, 0, 16, 16, 16, 16, 16, 16, 16]]
|
||||
pose_kpt_color = palette[[16, 16, 16, 16, 16, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9]]
|
||||
radius = 5
|
||||
num_kpts = len(kpts) // steps
|
||||
|
||||
for kid in range(num_kpts):
|
||||
r, g, b = pose_kpt_color[kid]
|
||||
x_coord, y_coord = kpts[steps * kid], kpts[steps * kid + 1]
|
||||
if not (x_coord % 640 == 0 or y_coord % 640 == 0):
|
||||
if steps == 3:
|
||||
conf = kpts[steps * kid + 2]
|
||||
if conf < 0.5:
|
||||
continue
|
||||
cv2.circle(im, (int(x_coord), int(y_coord)), radius, (int(r), int(g), int(b)), -1)
|
||||
|
||||
for sk_id, sk in enumerate(skeleton):
|
||||
r, g, b = pose_limb_color[sk_id]
|
||||
pos1 = (int(kpts[(sk[0]-1)*steps]), int(kpts[(sk[0]-1)*steps+1]))
|
||||
pos2 = (int(kpts[(sk[1]-1)*steps]), int(kpts[(sk[1]-1)*steps+1]))
|
||||
if steps == 3:
|
||||
conf1 = kpts[(sk[0]-1)*steps+2]
|
||||
conf2 = kpts[(sk[1]-1)*steps+2]
|
||||
if conf1<0.5 or conf2<0.5:
|
||||
continue
|
||||
if pos1[0]%640 == 0 or pos1[1]%640==0 or pos1[0]<0 or pos1[1]<0:
|
||||
continue
|
||||
if pos2[0] % 640 == 0 or pos2[1] % 640 == 0 or pos2[0]<0 or pos2[1]<0:
|
||||
continue
|
||||
cv2.line(im, pos1, pos2, (int(r), int(g), int(b)), thickness=2)
|
||||
@@ -0,0 +1,374 @@
|
||||
# YOLOR PyTorch utils
|
||||
|
||||
import datetime
|
||||
import logging
|
||||
import math
|
||||
import os
|
||||
import platform
|
||||
import subprocess
|
||||
import time
|
||||
from contextlib import contextmanager
|
||||
from copy import deepcopy
|
||||
from pathlib import Path
|
||||
|
||||
import torch
|
||||
import torch.backends.cudnn as cudnn
|
||||
import torch.nn as nn
|
||||
import torch.nn.functional as F
|
||||
import torchvision
|
||||
|
||||
try:
|
||||
import thop # for FLOPS computation
|
||||
except ImportError:
|
||||
thop = None
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
@contextmanager
|
||||
def torch_distributed_zero_first(local_rank: int):
|
||||
"""
|
||||
Decorator to make all processes in distributed training wait for each local_master to do something.
|
||||
"""
|
||||
if local_rank not in [-1, 0]:
|
||||
torch.distributed.barrier()
|
||||
yield
|
||||
if local_rank == 0:
|
||||
torch.distributed.barrier()
|
||||
|
||||
|
||||
def init_torch_seeds(seed=0):
|
||||
# Speed-reproducibility tradeoff https://pytorch.org/docs/stable/notes/randomness.html
|
||||
torch.manual_seed(seed)
|
||||
if seed == 0: # slower, more reproducible
|
||||
cudnn.benchmark, cudnn.deterministic = False, True
|
||||
else: # faster, less reproducible
|
||||
cudnn.benchmark, cudnn.deterministic = True, False
|
||||
|
||||
|
||||
def date_modified(path=__file__):
|
||||
# return human-readable file modification date, i.e. '2021-3-26'
|
||||
t = datetime.datetime.fromtimestamp(Path(path).stat().st_mtime)
|
||||
return f'{t.year}-{t.month}-{t.day}'
|
||||
|
||||
|
||||
def git_describe(path=Path(__file__).parent): # path must be a directory
|
||||
# return human-readable git description, i.e. v5.0-5-g3e25f1e https://git-scm.com/docs/git-describe
|
||||
s = f'git -C {path} describe --tags --long --always'
|
||||
try:
|
||||
return subprocess.check_output(s, shell=True, stderr=subprocess.STDOUT).decode()[:-1]
|
||||
except subprocess.CalledProcessError as e:
|
||||
return '' # not a git repository
|
||||
|
||||
|
||||
def select_device(device='', batch_size=None):
|
||||
# device = 'cpu' or '0' or '0,1,2,3'
|
||||
s = f'YOLOR 🚀 {git_describe() or date_modified()} torch {torch.__version__} ' # string
|
||||
cpu = device.lower() == 'cpu'
|
||||
if cpu:
|
||||
os.environ['CUDA_VISIBLE_DEVICES'] = '-1' # force torch.cuda.is_available() = False
|
||||
elif device: # non-cpu device requested
|
||||
os.environ['CUDA_VISIBLE_DEVICES'] = device # set environment variable
|
||||
assert torch.cuda.is_available(), f'CUDA unavailable, invalid device {device} requested' # check availability
|
||||
|
||||
cuda = not cpu and torch.cuda.is_available()
|
||||
if cuda:
|
||||
n = torch.cuda.device_count()
|
||||
if n > 1 and batch_size: # check that batch_size is compatible with device_count
|
||||
assert batch_size % n == 0, f'batch-size {batch_size} not multiple of GPU count {n}'
|
||||
space = ' ' * len(s)
|
||||
for i, d in enumerate(device.split(',') if device else range(n)):
|
||||
p = torch.cuda.get_device_properties(i)
|
||||
s += f"{'' if i == 0 else space}CUDA:{d} ({p.name}, {p.total_memory / 1024 ** 2}MB)\n" # bytes to MB
|
||||
else:
|
||||
s += 'CPU\n'
|
||||
|
||||
logger.info(s.encode().decode('ascii', 'ignore') if platform.system() == 'Windows' else s) # emoji-safe
|
||||
return torch.device('cuda:0' if cuda else 'cpu')
|
||||
|
||||
|
||||
def time_synchronized():
|
||||
# pytorch-accurate time
|
||||
if torch.cuda.is_available():
|
||||
torch.cuda.synchronize()
|
||||
return time.time()
|
||||
|
||||
|
||||
def profile(x, ops, n=100, device=None):
|
||||
# profile a pytorch module or list of modules. Example usage:
|
||||
# x = torch.randn(16, 3, 640, 640) # input
|
||||
# m1 = lambda x: x * torch.sigmoid(x)
|
||||
# m2 = nn.SiLU()
|
||||
# profile(x, [m1, m2], n=100) # profile speed over 100 iterations
|
||||
|
||||
device = device or torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
|
||||
x = x.to(device)
|
||||
x.requires_grad = True
|
||||
print(torch.__version__, device.type, torch.cuda.get_device_properties(0) if device.type == 'cuda' else '')
|
||||
print(f"\n{'Params':>12s}{'GFLOPS':>12s}{'forward (ms)':>16s}{'backward (ms)':>16s}{'input':>24s}{'output':>24s}")
|
||||
for m in ops if isinstance(ops, list) else [ops]:
|
||||
m = m.to(device) if hasattr(m, 'to') else m # device
|
||||
m = m.half() if hasattr(m, 'half') and isinstance(x, torch.Tensor) and x.dtype is torch.float16 else m # type
|
||||
dtf, dtb, t = 0., 0., [0., 0., 0.] # dt forward, backward
|
||||
try:
|
||||
flops = thop.profile(m, inputs=(x,), verbose=False)[0] / 1E9 * 2 # GFLOPS
|
||||
except:
|
||||
flops = 0
|
||||
|
||||
for _ in range(n):
|
||||
t[0] = time_synchronized()
|
||||
y = m(x)
|
||||
t[1] = time_synchronized()
|
||||
try:
|
||||
_ = y.sum().backward()
|
||||
t[2] = time_synchronized()
|
||||
except: # no backward method
|
||||
t[2] = float('nan')
|
||||
dtf += (t[1] - t[0]) * 1000 / n # ms per op forward
|
||||
dtb += (t[2] - t[1]) * 1000 / n # ms per op backward
|
||||
|
||||
s_in = tuple(x.shape) if isinstance(x, torch.Tensor) else 'list'
|
||||
s_out = tuple(y.shape) if isinstance(y, torch.Tensor) else 'list'
|
||||
p = sum(list(x.numel() for x in m.parameters())) if isinstance(m, nn.Module) else 0 # parameters
|
||||
print(f'{p:12}{flops:12.4g}{dtf:16.4g}{dtb:16.4g}{str(s_in):>24s}{str(s_out):>24s}')
|
||||
|
||||
|
||||
def is_parallel(model):
|
||||
return type(model) in (nn.parallel.DataParallel, nn.parallel.DistributedDataParallel)
|
||||
|
||||
|
||||
def intersect_dicts(da, db, exclude=()):
|
||||
# Dictionary intersection of matching keys and shapes, omitting 'exclude' keys, using da values
|
||||
return {k: v for k, v in da.items() if k in db and not any(x in k for x in exclude) and v.shape == db[k].shape}
|
||||
|
||||
|
||||
def initialize_weights(model):
|
||||
for m in model.modules():
|
||||
t = type(m)
|
||||
if t is nn.Conv2d:
|
||||
pass # nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
|
||||
elif t is nn.BatchNorm2d:
|
||||
m.eps = 1e-3
|
||||
m.momentum = 0.03
|
||||
elif t in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6]:
|
||||
m.inplace = True
|
||||
|
||||
|
||||
def find_modules(model, mclass=nn.Conv2d):
|
||||
# Finds layer indices matching module class 'mclass'
|
||||
return [i for i, m in enumerate(model.module_list) if isinstance(m, mclass)]
|
||||
|
||||
|
||||
def sparsity(model):
|
||||
# Return global model sparsity
|
||||
a, b = 0., 0.
|
||||
for p in model.parameters():
|
||||
a += p.numel()
|
||||
b += (p == 0).sum()
|
||||
return b / a
|
||||
|
||||
|
||||
def prune(model, amount=0.3):
|
||||
# Prune model to requested global sparsity
|
||||
import torch.nn.utils.prune as prune
|
||||
print('Pruning model... ', end='')
|
||||
for name, m in model.named_modules():
|
||||
if isinstance(m, nn.Conv2d):
|
||||
prune.l1_unstructured(m, name='weight', amount=amount) # prune
|
||||
prune.remove(m, 'weight') # make permanent
|
||||
print(' %.3g global sparsity' % sparsity(model))
|
||||
|
||||
|
||||
def fuse_conv_and_bn(conv, bn):
|
||||
# Fuse convolution and batchnorm layers https://tehnokv.com/posts/fusing-batchnorm-and-conv/
|
||||
fusedconv = nn.Conv2d(conv.in_channels,
|
||||
conv.out_channels,
|
||||
kernel_size=conv.kernel_size,
|
||||
stride=conv.stride,
|
||||
padding=conv.padding,
|
||||
groups=conv.groups,
|
||||
bias=True).requires_grad_(False).to(conv.weight.device)
|
||||
|
||||
# prepare filters
|
||||
w_conv = conv.weight.clone().view(conv.out_channels, -1)
|
||||
w_bn = torch.diag(bn.weight.div(torch.sqrt(bn.eps + bn.running_var)))
|
||||
fusedconv.weight.copy_(torch.mm(w_bn, w_conv).view(fusedconv.weight.shape))
|
||||
|
||||
# prepare spatial bias
|
||||
b_conv = torch.zeros(conv.weight.size(0), device=conv.weight.device) if conv.bias is None else conv.bias
|
||||
b_bn = bn.bias - bn.weight.mul(bn.running_mean).div(torch.sqrt(bn.running_var + bn.eps))
|
||||
fusedconv.bias.copy_(torch.mm(w_bn, b_conv.reshape(-1, 1)).reshape(-1) + b_bn)
|
||||
|
||||
return fusedconv
|
||||
|
||||
|
||||
def model_info(model, verbose=False, img_size=640):
|
||||
# Model information. img_size may be int or list, i.e. img_size=640 or img_size=[640, 320]
|
||||
n_p = sum(x.numel() for x in model.parameters()) # number parameters
|
||||
n_g = sum(x.numel() for x in model.parameters() if x.requires_grad) # number gradients
|
||||
if verbose:
|
||||
print('%5s %40s %9s %12s %20s %10s %10s' % ('layer', 'name', 'gradient', 'parameters', 'shape', 'mu', 'sigma'))
|
||||
for i, (name, p) in enumerate(model.named_parameters()):
|
||||
name = name.replace('module_list.', '')
|
||||
print('%5g %40s %9s %12g %20s %10.3g %10.3g' %
|
||||
(i, name, p.requires_grad, p.numel(), list(p.shape), p.mean(), p.std()))
|
||||
|
||||
try: # FLOPS
|
||||
from thop import profile
|
||||
stride = max(int(model.stride.max()), 32) if hasattr(model, 'stride') else 32
|
||||
img = torch.zeros((1, model.yaml.get('ch', 3), stride, stride), device=next(model.parameters()).device) # input
|
||||
flops = profile(deepcopy(model), inputs=(img,), verbose=False)[0] / 1E9 * 2 # stride GFLOPS
|
||||
img_size = img_size if isinstance(img_size, list) else [img_size, img_size] # expand if int/float
|
||||
fs = ', %.1f GFLOPS' % (flops * img_size[0] / stride * img_size[1] / stride) # 640x640 GFLOPS
|
||||
except (ImportError, Exception):
|
||||
fs = ''
|
||||
|
||||
logger.info(f"Model Summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}")
|
||||
|
||||
|
||||
def load_classifier(name='resnet101', n=2):
|
||||
# Loads a pretrained model reshaped to n-class output
|
||||
model = torchvision.models.__dict__[name](pretrained=True)
|
||||
|
||||
# ResNet model properties
|
||||
# input_size = [3, 224, 224]
|
||||
# input_space = 'RGB'
|
||||
# input_range = [0, 1]
|
||||
# mean = [0.485, 0.456, 0.406]
|
||||
# std = [0.229, 0.224, 0.225]
|
||||
|
||||
# Reshape output to n classes
|
||||
filters = model.fc.weight.shape[1]
|
||||
model.fc.bias = nn.Parameter(torch.zeros(n), requires_grad=True)
|
||||
model.fc.weight = nn.Parameter(torch.zeros(n, filters), requires_grad=True)
|
||||
model.fc.out_features = n
|
||||
return model
|
||||
|
||||
|
||||
def scale_img(img, ratio=1.0, same_shape=False, gs=32): # img(16,3,256,416)
|
||||
# scales img(bs,3,y,x) by ratio constrained to gs-multiple
|
||||
if ratio == 1.0:
|
||||
return img
|
||||
else:
|
||||
h, w = img.shape[2:]
|
||||
s = (int(h * ratio), int(w * ratio)) # new size
|
||||
img = F.interpolate(img, size=s, mode='bilinear', align_corners=False) # resize
|
||||
if not same_shape: # pad/crop img
|
||||
h, w = [math.ceil(x * ratio / gs) * gs for x in (h, w)]
|
||||
return F.pad(img, [0, w - s[1], 0, h - s[0]], value=0.447) # value = imagenet mean
|
||||
|
||||
|
||||
def copy_attr(a, b, include=(), exclude=()):
|
||||
# Copy attributes from b to a, options to only include [...] and to exclude [...]
|
||||
for k, v in b.__dict__.items():
|
||||
if (len(include) and k not in include) or k.startswith('_') or k in exclude:
|
||||
continue
|
||||
else:
|
||||
setattr(a, k, v)
|
||||
|
||||
|
||||
class ModelEMA:
|
||||
""" Model Exponential Moving Average from https://github.com/rwightman/pytorch-image-models
|
||||
Keep a moving average of everything in the model state_dict (parameters and buffers).
|
||||
This is intended to allow functionality like
|
||||
https://www.tensorflow.org/api_docs/python/tf/train/ExponentialMovingAverage
|
||||
A smoothed version of the weights is necessary for some training schemes to perform well.
|
||||
This class is sensitive where it is initialized in the sequence of model init,
|
||||
GPU assignment and distributed training wrappers.
|
||||
"""
|
||||
|
||||
def __init__(self, model, decay=0.9999, updates=0):
|
||||
# Create EMA
|
||||
self.ema = deepcopy(model.module if is_parallel(model) else model).eval() # FP32 EMA
|
||||
# if next(model.parameters()).device.type != 'cpu':
|
||||
# self.ema.half() # FP16 EMA
|
||||
self.updates = updates # number of EMA updates
|
||||
self.decay = lambda x: decay * (1 - math.exp(-x / 2000)) # decay exponential ramp (to help early epochs)
|
||||
for p in self.ema.parameters():
|
||||
p.requires_grad_(False)
|
||||
|
||||
def update(self, model):
|
||||
# Update EMA parameters
|
||||
with torch.no_grad():
|
||||
self.updates += 1
|
||||
d = self.decay(self.updates)
|
||||
|
||||
msd = model.module.state_dict() if is_parallel(model) else model.state_dict() # model state_dict
|
||||
for k, v in self.ema.state_dict().items():
|
||||
if v.dtype.is_floating_point:
|
||||
v *= d
|
||||
v += (1. - d) * msd[k].detach()
|
||||
|
||||
def update_attr(self, model, include=(), exclude=('process_group', 'reducer')):
|
||||
# Update EMA attributes
|
||||
copy_attr(self.ema, model, include, exclude)
|
||||
|
||||
|
||||
class BatchNormXd(torch.nn.modules.batchnorm._BatchNorm):
|
||||
def _check_input_dim(self, input):
|
||||
# The only difference between BatchNorm1d, BatchNorm2d, BatchNorm3d, etc
|
||||
# is this method that is overwritten by the sub-class
|
||||
# This original goal of this method was for tensor sanity checks
|
||||
# If you're ok bypassing those sanity checks (eg. if you trust your inference
|
||||
# to provide the right dimensional inputs), then you can just use this method
|
||||
# for easy conversion from SyncBatchNorm
|
||||
# (unfortunately, SyncBatchNorm does not store the original class - if it did
|
||||
# we could return the one that was originally created)
|
||||
return
|
||||
|
||||
def revert_sync_batchnorm(module):
|
||||
# this is very similar to the function that it is trying to revert:
|
||||
# https://github.com/pytorch/pytorch/blob/c8b3686a3e4ba63dc59e5dcfe5db3430df256833/torch/nn/modules/batchnorm.py#L679
|
||||
module_output = module
|
||||
if isinstance(module, torch.nn.modules.batchnorm.SyncBatchNorm):
|
||||
new_cls = BatchNormXd
|
||||
module_output = BatchNormXd(module.num_features,
|
||||
module.eps, module.momentum,
|
||||
module.affine,
|
||||
module.track_running_stats)
|
||||
if module.affine:
|
||||
with torch.no_grad():
|
||||
module_output.weight = module.weight
|
||||
module_output.bias = module.bias
|
||||
module_output.running_mean = module.running_mean
|
||||
module_output.running_var = module.running_var
|
||||
module_output.num_batches_tracked = module.num_batches_tracked
|
||||
if hasattr(module, "qconfig"):
|
||||
module_output.qconfig = module.qconfig
|
||||
for name, child in module.named_children():
|
||||
module_output.add_module(name, revert_sync_batchnorm(child))
|
||||
del module
|
||||
return module_output
|
||||
|
||||
|
||||
class TracedModel(nn.Module):
|
||||
|
||||
def __init__(self, model=None, device=None, img_size=(640,640)):
|
||||
super(TracedModel, self).__init__()
|
||||
|
||||
print(" Convert model to Traced-model... ")
|
||||
self.stride = model.stride
|
||||
self.names = model.names
|
||||
self.model = model
|
||||
|
||||
self.model = revert_sync_batchnorm(self.model)
|
||||
self.model.to('cpu')
|
||||
self.model.eval()
|
||||
|
||||
self.detect_layer = self.model.model[-1]
|
||||
self.model.traced = True
|
||||
|
||||
rand_example = torch.rand(1, 3, img_size, img_size)
|
||||
|
||||
traced_script_module = torch.jit.trace(self.model, rand_example, strict=False)
|
||||
#traced_script_module = torch.jit.script(self.model)
|
||||
traced_script_module.save("traced_model.pt")
|
||||
print(" traced_script_module saved! ")
|
||||
self.model = traced_script_module
|
||||
self.model.to(device)
|
||||
self.detect_layer.to(device)
|
||||
print(" model is traced! \n")
|
||||
|
||||
def forward(self, x, augment=False, profile=False):
|
||||
out = self.model(x)
|
||||
out = self.detect_layer(out)
|
||||
return out
|
||||
@@ -0,0 +1 @@
|
||||
# init
|
||||
@@ -0,0 +1,24 @@
|
||||
import argparse
|
||||
|
||||
import yaml
|
||||
|
||||
from wandb_utils import WandbLogger
|
||||
|
||||
WANDB_ARTIFACT_PREFIX = 'wandb-artifact://'
|
||||
|
||||
|
||||
def create_dataset_artifact(opt):
|
||||
with open(opt.data) as f:
|
||||
data = yaml.load(f, Loader=yaml.SafeLoader) # data dict
|
||||
logger = WandbLogger(opt, '', None, data, job_type='Dataset Creation')
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument('--data', type=str, default='data/coco.yaml', help='data.yaml path')
|
||||
parser.add_argument('--single-cls', action='store_true', help='train as single-class dataset')
|
||||
parser.add_argument('--project', type=str, default='YOLOR', help='name of W&B Project')
|
||||
opt = parser.parse_args()
|
||||
opt.resume = False # Explicitly disallow resume check for dataset upload job
|
||||
|
||||
create_dataset_artifact(opt)
|
||||
@@ -0,0 +1,306 @@
|
||||
import json
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
import torch
|
||||
import yaml
|
||||
from tqdm import tqdm
|
||||
|
||||
sys.path.append(str(Path(__file__).parent.parent.parent)) # add utils/ to path
|
||||
from utils.datasets import LoadImagesAndLabels
|
||||
from utils.datasets import img2label_paths
|
||||
from utils.general import colorstr, xywh2xyxy, check_dataset
|
||||
|
||||
try:
|
||||
import wandb
|
||||
from wandb import init, finish
|
||||
except ImportError:
|
||||
wandb = None
|
||||
|
||||
WANDB_ARTIFACT_PREFIX = 'wandb-artifact://'
|
||||
|
||||
|
||||
def remove_prefix(from_string, prefix=WANDB_ARTIFACT_PREFIX):
|
||||
return from_string[len(prefix):]
|
||||
|
||||
|
||||
def check_wandb_config_file(data_config_file):
|
||||
wandb_config = '_wandb.'.join(data_config_file.rsplit('.', 1)) # updated data.yaml path
|
||||
if Path(wandb_config).is_file():
|
||||
return wandb_config
|
||||
return data_config_file
|
||||
|
||||
|
||||
def get_run_info(run_path):
|
||||
run_path = Path(remove_prefix(run_path, WANDB_ARTIFACT_PREFIX))
|
||||
run_id = run_path.stem
|
||||
project = run_path.parent.stem
|
||||
model_artifact_name = 'run_' + run_id + '_model'
|
||||
return run_id, project, model_artifact_name
|
||||
|
||||
|
||||
def check_wandb_resume(opt):
|
||||
process_wandb_config_ddp_mode(opt) if opt.global_rank not in [-1, 0] else None
|
||||
if isinstance(opt.resume, str):
|
||||
if opt.resume.startswith(WANDB_ARTIFACT_PREFIX):
|
||||
if opt.global_rank not in [-1, 0]: # For resuming DDP runs
|
||||
run_id, project, model_artifact_name = get_run_info(opt.resume)
|
||||
api = wandb.Api()
|
||||
artifact = api.artifact(project + '/' + model_artifact_name + ':latest')
|
||||
modeldir = artifact.download()
|
||||
opt.weights = str(Path(modeldir) / "last.pt")
|
||||
return True
|
||||
return None
|
||||
|
||||
|
||||
def process_wandb_config_ddp_mode(opt):
|
||||
with open(opt.data) as f:
|
||||
data_dict = yaml.load(f, Loader=yaml.SafeLoader) # data dict
|
||||
train_dir, val_dir = None, None
|
||||
if isinstance(data_dict['train'], str) and data_dict['train'].startswith(WANDB_ARTIFACT_PREFIX):
|
||||
api = wandb.Api()
|
||||
train_artifact = api.artifact(remove_prefix(data_dict['train']) + ':' + opt.artifact_alias)
|
||||
train_dir = train_artifact.download()
|
||||
train_path = Path(train_dir) / 'data/images/'
|
||||
data_dict['train'] = str(train_path)
|
||||
|
||||
if isinstance(data_dict['val'], str) and data_dict['val'].startswith(WANDB_ARTIFACT_PREFIX):
|
||||
api = wandb.Api()
|
||||
val_artifact = api.artifact(remove_prefix(data_dict['val']) + ':' + opt.artifact_alias)
|
||||
val_dir = val_artifact.download()
|
||||
val_path = Path(val_dir) / 'data/images/'
|
||||
data_dict['val'] = str(val_path)
|
||||
if train_dir or val_dir:
|
||||
ddp_data_path = str(Path(val_dir) / 'wandb_local_data.yaml')
|
||||
with open(ddp_data_path, 'w') as f:
|
||||
yaml.dump(data_dict, f)
|
||||
opt.data = ddp_data_path
|
||||
|
||||
|
||||
class WandbLogger():
|
||||
def __init__(self, opt, name, run_id, data_dict, job_type='Training'):
|
||||
# Pre-training routine --
|
||||
self.job_type = job_type
|
||||
self.wandb, self.wandb_run, self.data_dict = wandb, None if not wandb else wandb.run, data_dict
|
||||
# It's more elegant to stick to 1 wandb.init call, but useful config data is overwritten in the WandbLogger's wandb.init call
|
||||
if isinstance(opt.resume, str): # checks resume from artifact
|
||||
if opt.resume.startswith(WANDB_ARTIFACT_PREFIX):
|
||||
run_id, project, model_artifact_name = get_run_info(opt.resume)
|
||||
model_artifact_name = WANDB_ARTIFACT_PREFIX + model_artifact_name
|
||||
assert wandb, 'install wandb to resume wandb runs'
|
||||
# Resume wandb-artifact:// runs here| workaround for not overwriting wandb.config
|
||||
self.wandb_run = wandb.init(id=run_id, project=project, resume='allow')
|
||||
opt.resume = model_artifact_name
|
||||
elif self.wandb:
|
||||
self.wandb_run = wandb.init(config=opt,
|
||||
resume="allow",
|
||||
project='YOLOR' if opt.project == 'runs/train' else Path(opt.project).stem,
|
||||
name=name,
|
||||
job_type=job_type,
|
||||
id=run_id) if not wandb.run else wandb.run
|
||||
if self.wandb_run:
|
||||
if self.job_type == 'Training':
|
||||
if not opt.resume:
|
||||
wandb_data_dict = self.check_and_upload_dataset(opt) if opt.upload_dataset else data_dict
|
||||
# Info useful for resuming from artifacts
|
||||
self.wandb_run.config.opt = vars(opt)
|
||||
self.wandb_run.config.data_dict = wandb_data_dict
|
||||
self.data_dict = self.setup_training(opt, data_dict)
|
||||
if self.job_type == 'Dataset Creation':
|
||||
self.data_dict = self.check_and_upload_dataset(opt)
|
||||
else:
|
||||
prefix = colorstr('wandb: ')
|
||||
print(f"{prefix}Install Weights & Biases for YOLOR logging with 'pip install wandb' (recommended)")
|
||||
|
||||
def check_and_upload_dataset(self, opt):
|
||||
assert wandb, 'Install wandb to upload dataset'
|
||||
check_dataset(self.data_dict)
|
||||
config_path = self.log_dataset_artifact(opt.data,
|
||||
opt.single_cls,
|
||||
'YOLOR' if opt.project == 'runs/train' else Path(opt.project).stem)
|
||||
print("Created dataset config file ", config_path)
|
||||
with open(config_path) as f:
|
||||
wandb_data_dict = yaml.load(f, Loader=yaml.SafeLoader)
|
||||
return wandb_data_dict
|
||||
|
||||
def setup_training(self, opt, data_dict):
|
||||
self.log_dict, self.current_epoch, self.log_imgs = {}, 0, 16 # Logging Constants
|
||||
self.bbox_interval = opt.bbox_interval
|
||||
if isinstance(opt.resume, str):
|
||||
modeldir, _ = self.download_model_artifact(opt)
|
||||
if modeldir:
|
||||
self.weights = Path(modeldir) / "last.pt"
|
||||
config = self.wandb_run.config
|
||||
opt.weights, opt.save_period, opt.batch_size, opt.bbox_interval, opt.epochs, opt.hyp = str(
|
||||
self.weights), config.save_period, config.total_batch_size, config.bbox_interval, config.epochs, \
|
||||
config.opt['hyp']
|
||||
data_dict = dict(self.wandb_run.config.data_dict) # eliminates the need for config file to resume
|
||||
if 'val_artifact' not in self.__dict__: # If --upload_dataset is set, use the existing artifact, don't download
|
||||
self.train_artifact_path, self.train_artifact = self.download_dataset_artifact(data_dict.get('train'),
|
||||
opt.artifact_alias)
|
||||
self.val_artifact_path, self.val_artifact = self.download_dataset_artifact(data_dict.get('val'),
|
||||
opt.artifact_alias)
|
||||
self.result_artifact, self.result_table, self.val_table, self.weights = None, None, None, None
|
||||
if self.train_artifact_path is not None:
|
||||
train_path = Path(self.train_artifact_path) / 'data/images/'
|
||||
data_dict['train'] = str(train_path)
|
||||
if self.val_artifact_path is not None:
|
||||
val_path = Path(self.val_artifact_path) / 'data/images/'
|
||||
data_dict['val'] = str(val_path)
|
||||
self.val_table = self.val_artifact.get("val")
|
||||
self.map_val_table_path()
|
||||
if self.val_artifact is not None:
|
||||
self.result_artifact = wandb.Artifact("run_" + wandb.run.id + "_progress", "evaluation")
|
||||
self.result_table = wandb.Table(["epoch", "id", "prediction", "avg_confidence"])
|
||||
if opt.bbox_interval == -1:
|
||||
self.bbox_interval = opt.bbox_interval = (opt.epochs // 10) if opt.epochs > 10 else 1
|
||||
return data_dict
|
||||
|
||||
def download_dataset_artifact(self, path, alias):
|
||||
if isinstance(path, str) and path.startswith(WANDB_ARTIFACT_PREFIX):
|
||||
dataset_artifact = wandb.use_artifact(remove_prefix(path, WANDB_ARTIFACT_PREFIX) + ":" + alias)
|
||||
assert dataset_artifact is not None, "'Error: W&B dataset artifact doesn\'t exist'"
|
||||
datadir = dataset_artifact.download()
|
||||
return datadir, dataset_artifact
|
||||
return None, None
|
||||
|
||||
def download_model_artifact(self, opt):
|
||||
if opt.resume.startswith(WANDB_ARTIFACT_PREFIX):
|
||||
model_artifact = wandb.use_artifact(remove_prefix(opt.resume, WANDB_ARTIFACT_PREFIX) + ":latest")
|
||||
assert model_artifact is not None, 'Error: W&B model artifact doesn\'t exist'
|
||||
modeldir = model_artifact.download()
|
||||
epochs_trained = model_artifact.metadata.get('epochs_trained')
|
||||
total_epochs = model_artifact.metadata.get('total_epochs')
|
||||
assert epochs_trained < total_epochs, 'training to %g epochs is finished, nothing to resume.' % (
|
||||
total_epochs)
|
||||
return modeldir, model_artifact
|
||||
return None, None
|
||||
|
||||
def log_model(self, path, opt, epoch, fitness_score, best_model=False):
|
||||
model_artifact = wandb.Artifact('run_' + wandb.run.id + '_model', type='model', metadata={
|
||||
'original_url': str(path),
|
||||
'epochs_trained': epoch + 1,
|
||||
'save period': opt.save_period,
|
||||
'project': opt.project,
|
||||
'total_epochs': opt.epochs,
|
||||
'fitness_score': fitness_score
|
||||
})
|
||||
model_artifact.add_file(str(path / 'last.pt'), name='last.pt')
|
||||
wandb.log_artifact(model_artifact,
|
||||
aliases=['latest', 'epoch ' + str(self.current_epoch), 'best' if best_model else ''])
|
||||
print("Saving model artifact on epoch ", epoch + 1)
|
||||
|
||||
def log_dataset_artifact(self, data_file, single_cls, project, overwrite_config=False):
|
||||
with open(data_file) as f:
|
||||
data = yaml.load(f, Loader=yaml.SafeLoader) # data dict
|
||||
nc, names = (1, ['item']) if single_cls else (int(data['nc']), data['names'])
|
||||
names = {k: v for k, v in enumerate(names)} # to index dictionary
|
||||
self.train_artifact = self.create_dataset_table(LoadImagesAndLabels(
|
||||
data['train']), names, name='train') if data.get('train') else None
|
||||
self.val_artifact = self.create_dataset_table(LoadImagesAndLabels(
|
||||
data['val']), names, name='val') if data.get('val') else None
|
||||
if data.get('train'):
|
||||
data['train'] = WANDB_ARTIFACT_PREFIX + str(Path(project) / 'train')
|
||||
if data.get('val'):
|
||||
data['val'] = WANDB_ARTIFACT_PREFIX + str(Path(project) / 'val')
|
||||
path = data_file if overwrite_config else '_wandb.'.join(data_file.rsplit('.', 1)) # updated data.yaml path
|
||||
data.pop('download', None)
|
||||
with open(path, 'w') as f:
|
||||
yaml.dump(data, f)
|
||||
|
||||
if self.job_type == 'Training': # builds correct artifact pipeline graph
|
||||
self.wandb_run.use_artifact(self.val_artifact)
|
||||
self.wandb_run.use_artifact(self.train_artifact)
|
||||
self.val_artifact.wait()
|
||||
self.val_table = self.val_artifact.get('val')
|
||||
self.map_val_table_path()
|
||||
else:
|
||||
self.wandb_run.log_artifact(self.train_artifact)
|
||||
self.wandb_run.log_artifact(self.val_artifact)
|
||||
return path
|
||||
|
||||
def map_val_table_path(self):
|
||||
self.val_table_map = {}
|
||||
print("Mapping dataset")
|
||||
for i, data in enumerate(tqdm(self.val_table.data)):
|
||||
self.val_table_map[data[3]] = data[0]
|
||||
|
||||
def create_dataset_table(self, dataset, class_to_id, name='dataset'):
|
||||
# TODO: Explore multiprocessing to slpit this loop parallely| This is essential for speeding up the the logging
|
||||
artifact = wandb.Artifact(name=name, type="dataset")
|
||||
img_files = tqdm([dataset.path]) if isinstance(dataset.path, str) and Path(dataset.path).is_dir() else None
|
||||
img_files = tqdm(dataset.img_files) if not img_files else img_files
|
||||
for img_file in img_files:
|
||||
if Path(img_file).is_dir():
|
||||
artifact.add_dir(img_file, name='data/images')
|
||||
labels_path = 'labels'.join(dataset.path.rsplit('images', 1))
|
||||
artifact.add_dir(labels_path, name='data/labels')
|
||||
else:
|
||||
artifact.add_file(img_file, name='data/images/' + Path(img_file).name)
|
||||
label_file = Path(img2label_paths([img_file])[0])
|
||||
artifact.add_file(str(label_file),
|
||||
name='data/labels/' + label_file.name) if label_file.exists() else None
|
||||
table = wandb.Table(columns=["id", "train_image", "Classes", "name"])
|
||||
class_set = wandb.Classes([{'id': id, 'name': name} for id, name in class_to_id.items()])
|
||||
for si, (img, labels, paths, shapes) in enumerate(tqdm(dataset)):
|
||||
height, width = shapes[0]
|
||||
labels[:, 2:] = (xywh2xyxy(labels[:, 2:].view(-1, 4))) * torch.Tensor([width, height, width, height])
|
||||
box_data, img_classes = [], {}
|
||||
for cls, *xyxy in labels[:, 1:].tolist():
|
||||
cls = int(cls)
|
||||
box_data.append({"position": {"minX": xyxy[0], "minY": xyxy[1], "maxX": xyxy[2], "maxY": xyxy[3]},
|
||||
"class_id": cls,
|
||||
"box_caption": "%s" % (class_to_id[cls]),
|
||||
"scores": {"acc": 1},
|
||||
"domain": "pixel"})
|
||||
img_classes[cls] = class_to_id[cls]
|
||||
boxes = {"ground_truth": {"box_data": box_data, "class_labels": class_to_id}} # inference-space
|
||||
table.add_data(si, wandb.Image(paths, classes=class_set, boxes=boxes), json.dumps(img_classes),
|
||||
Path(paths).name)
|
||||
artifact.add(table, name)
|
||||
return artifact
|
||||
|
||||
def log_training_progress(self, predn, path, names):
|
||||
if self.val_table and self.result_table:
|
||||
class_set = wandb.Classes([{'id': id, 'name': name} for id, name in names.items()])
|
||||
box_data = []
|
||||
total_conf = 0
|
||||
for *xyxy, conf, cls in predn.tolist():
|
||||
if conf >= 0.25:
|
||||
box_data.append(
|
||||
{"position": {"minX": xyxy[0], "minY": xyxy[1], "maxX": xyxy[2], "maxY": xyxy[3]},
|
||||
"class_id": int(cls),
|
||||
"box_caption": "%s %.3f" % (names[cls], conf),
|
||||
"scores": {"class_score": conf},
|
||||
"domain": "pixel"})
|
||||
total_conf = total_conf + conf
|
||||
boxes = {"predictions": {"box_data": box_data, "class_labels": names}} # inference-space
|
||||
id = self.val_table_map[Path(path).name]
|
||||
self.result_table.add_data(self.current_epoch,
|
||||
id,
|
||||
wandb.Image(self.val_table.data[id][1], boxes=boxes, classes=class_set),
|
||||
total_conf / max(1, len(box_data))
|
||||
)
|
||||
|
||||
def log(self, log_dict):
|
||||
if self.wandb_run:
|
||||
for key, value in log_dict.items():
|
||||
self.log_dict[key] = value
|
||||
|
||||
def end_epoch(self, best_result=False):
|
||||
if self.wandb_run:
|
||||
wandb.log(self.log_dict)
|
||||
self.log_dict = {}
|
||||
if self.result_artifact:
|
||||
train_results = wandb.JoinedTable(self.val_table, self.result_table, "id")
|
||||
self.result_artifact.add(train_results, 'result')
|
||||
wandb.log_artifact(self.result_artifact, aliases=['latest', 'epoch ' + str(self.current_epoch),
|
||||
('best' if best_result else '')])
|
||||
self.result_table = wandb.Table(["epoch", "id", "prediction", "avg_confidence"])
|
||||
self.result_artifact = wandb.Artifact("run_" + wandb.run.id + "_progress", "evaluation")
|
||||
|
||||
def finish_run(self):
|
||||
if self.wandb_run:
|
||||
if self.log_dict:
|
||||
wandb.log(self.log_dict)
|
||||
wandb.run.finish()
|
||||
Reference in New Issue
Block a user