val benchmark==MOT20 BUG

[XDUbeginner]

Search before asking

• I have searched the Yolo Tracking issues and discussions and found no similar questions.

Yolo Tracking Component

Evaluation

Bug

when I run val.py , report this error

Environment

Python == 3.9
Package Version Editable project location

---

absl-py 1.4.0
addict 2.4.0
aliyun-python-sdk-core 2.14.0
aliyun-python-sdk-kms 2.16.2
annotated-types 0.6.0
appdirs 1.4.4
beautifulsoup4 4.12.2
boxmot 10.0.52 e:\yolo_tracking-master2024_3_4\yolo_tracking-master
brotlipy 0.7.0
cachetools 5.3.0
certifi 2023.5.7
cffi 1.15.1
cfgv 3.4.0
charset-normalizer 3.1.0
click 8.1.7
colorama 0.4.6
coloredlogs 15.0.1
contourpy 1.0.7
crcmod 1.7
cryptography 40.0.2
cycler 0.11.0
Cython 3.0.8
distlib 0.3.8
easydict 1.10
efficientnet-pytorch 0.7.1
einops 0.6.1
filelock 3.13.1
filterpy 1.4.5
flatbuffers 23.5.9
fonttools 4.39.3
fsspec 2023.6.0
ftfy 6.1.3
future 0.18.3
gdown 5.1.0
gitdb 4.0.10
GitPython 3.1.31
google-auth 2.18.0
google-auth-oauthlib 1.0.0
grad-cam 1.4.8
grpcio 1.54.0
huggingface-hub 0.15.1
humanfriendly 10.0
identify 2.5.35
idna 3.4
importlib-metadata 6.6.0
importlib-resources 5.12.0
Jinja2 3.1.3
jmespath 0.10.0
joblib 1.3.1
kiwisolver 1.4.4
lap 0.4.0
lapx 0.5.5
loguru 0.7.0
Markdown 3.4.3
markdown-it-py 3.0.0
MarkupSafe 2.1.2
matplotlib 3.7.1
mdurl 0.1.2
mkl-service 2.4.0
mmcv 2.1.0
mmengine 0.10.3
model-index 0.1.11
motmetrics 1.4.0
mpmath 1.3.0
networkx 3.2.1
nodeenv 1.8.0
numpy 1.24.4
nvidia-cublas-cu11 11.11.3.6
oauthlib 3.2.2
onnxruntime-gpu 1.14.1
opencv-python 4.7.0.72
opendatalab 0.0.10
openmim 0.3.9
openxlab 0.0.34
ordered-set 4.1.0
oss2 2.17.0
packaging 23.1
pandas 2.0.1
Pillow 9.5.0
pip 23.1.2
platformdirs 4.2.0
pre-commit 3.6.2
protobuf 4.23.0
psutil 5.9.5
py-cpuinfo 9.0.0
pyasn1 0.5.0
pyasn1-modules 0.3.0
pycocotools 2.0.7
pycparser 2.21
pycryptodome 3.20.0
pydantic 2.6.3
pydantic_core 2.16.3
Pygments 2.15.1
pyOpenSSL 23.1.1
pyparsing 3.1.0b1
pyreadline3 3.4.1
PySocks 1.7.1
python-dateutil 2.8.2
pytz 2023.3
pywin32 306
PyYAML 6.0
regex 2023.6.3
requests 2.28.2
requests-oauthlib 1.3.1
rich 13.4.2
rsa 4.9
safetensors 0.3.1
scikit-learn 1.3.0
scipy 1.10.1
seaborn 0.12.2
sentry-sdk 1.22.2
setuptools 60.2.0
six 1.16.0
smmap 5.0.0
soupsieve 2.4.1
sympy 1.12
tabulate 0.9.0
tensorboard 2.13.0
tensorboard-data-server 0.7.0
termcolor 2.3.0
thop 0.1.1.post2209072238
threadpoolctl 3.2.0
tidecv 1.0.1
timm 0.6.13
tomli 2.0.1
torch 1.13.1+cu117
torchaudio 0.13.1+cu117
torchvision 0.14.1+cu117
tqdm 4.65.0
ttach 0.0.3
typing_extensions 4.10.0
tzdata 2023.3
ultralytics 8.0.228
urllib3 1.26.15
virtualenv 20.25.1
wcwidth 0.2.13
Werkzeug 2.3.4
wheel 0.40.0
win-inet-pton 1.1.0
win32-setctime 1.1.0
xmltodict 0.13.0
yacs 0.1.8
yapf 0.40.0
zipp 3.15.0

Minimal Reproducible Example

import re
import sys
import argparse
import subprocess
from boxmot.utils import EXAMPLES, ROOT, WEIGHTS, EXPERIMENTATION
from pathlib import Path
from tracking.utils import (
download_mot_eval_tools,
download_mot_dataset,
unzip_mot_dataset,
eval_setup
)
from ultralytics.utils.files import increment_path

def parse_mot_results(results):
"""Extract the COMBINED HOTA, MOTA, IDF1 from the results generate by the
run_mot_challenge.py script.

    Args:
        str: mot_results

    Returns:
        (dict): {'HOTA': x, 'MOTA':y, 'IDF1':z}
    """
    combined_results = results.split('COMBINED')[2:-1]
    # robust way of getting first ints/float in string
    combined_results = [float(re.findall("[-+]?(?:\d*\.*\d+)", f)[0]) for f in combined_results]
    # pack everything in dict
    combined_results = {key: value for key, value in zip(['HOTA', 'MOTA', 'IDF1'], combined_results)}
    return combined_results

def trackeval(
args,
seq_paths,
save_dir,
MOT_results_folder,
gt_folder,
metrics = ["HOTA", "CLEAR", "Identity"]
):
"""
Executes a Python script to evaluate MOT challenge tracking results using specified metrics.

Parameters:
    script_path (str): The path to the evaluation script to run.
    trackers_folder (str): The folder where tracker results are stored.
    metrics (list): A list of metrics to use for evaluation. Defaults to ["HOTA", "CLEAR", "Identity"].
    num_parallel_cores (int): The number of parallel cores to use for evaluation. Defaults to 4.

Outputs:
    Prints the standard output and standard error from the evaluation script.
"""
# Define paths
d = [seq_path.parent.name for seq_path in seq_paths]
# Prepare arguments for subprocess call
args = [
    sys.executable, EXPERIMENTATION / 'val_utils' / 'scripts' / 'run_mot_challenge.py',
    "--GT_FOLDER", str(gt_folder),
    "--BENCHMARK", "",
    "--TRACKERS_FOLDER", args.exp_folder_path,
    "--TRACKERS_TO_EVAL", "",
    "--SPLIT_TO_EVAL", "train",
    "--METRICS", *metrics,
    "--USE_PARALLEL", "True",
    "--TRACKER_SUB_FOLDER", "",
    "--NUM_PARALLEL_CORES", str(4),
    "--SKIP_SPLIT_FOL", "True",
    "--SEQ_INFO", *d
]

# Execute the evaluation script
p = subprocess.Popen(
    args=args,
    stdout=subprocess.PIPE,
    stderr=subprocess.PIPE,
    text=True
)

stdout, stderr = p.communicate()

# Output the results
print("Standard Output:\n", stdout)
if stderr:
    print("Standard Error:\n", stderr)
return stdout

def parse_opt():
parser = argparse.ArgumentParser()
parser.add_argument('--yolo-model', type=Path, default=WEIGHTS / 'yolov8_best.pt',
help='yolo model path')
parser.add_argument('--reid-model', type=Path, default=WEIGHTS / 'osnet_x1_0_msmt17.pt',
help='reid model path')
parser.add_argument('--tracking-method', type=str, default='botsort',
help='deepocsort, botsort, strongsort, ocsort, bytetrack')
parser.add_argument('--source', type=str, default='F:/all_dataset/VisDrone-Track/VisDrone2019-MOT-test-challenge/sequences/uav0000320_03910_v',
help='file/dir/URL/glob, 0 for webcam')
parser.add_argument('--imgsz', '--img', '--img-size', nargs='+', type=int, default=[640],
help='inference size h,w')
parser.add_argument('--conf', type=float, default=0.1,
help='confidence threshold')
parser.add_argument('--iou', type=float, default=0.1,
help='intersection over union (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('--show', action='store_true',
help='display tracking video results')
parser.add_argument('--save', action='store_true',
help='save video tracking results')
# class 0 is person, 1 is bycicle, 2 is car... 79 is oven
parser.add_argument('--classes', nargs='+', type=int,
help='filter by class: --classes 0, or --classes 0 2 3')
parser.add_argument('--project', default=ROOT / 'runs' / 'mot',
help='save results to project/name')
parser.add_argument('--name', default='yolov8_best_osnet_x1_0_msmt17',
help='save results to project/name')
parser.add_argument('--dets', type=str, default='yolov8_best',
help='the folder name under project to load the detections from')
parser.add_argument('--embs', type=str, default='osnet_x1_0_msmt17',
help='the folder name under project/dets to load the embeddings from')
parser.add_argument('--exist-ok', action='store_true',
help='existing project/name ok, do not increment')
parser.add_argument('--half', action='store_true',
help='use FP16 half-precision inference')
parser.add_argument('--vid-stride', type=int, default=1,
help='video frame-rate stride')
parser.add_argument('--show-labels', action='store_false',
help='either show all or only bboxes')
parser.add_argument('--show-conf', action='store_false',
help='hide confidences when show')
parser.add_argument('--save-txt', action='store_true',default = True,
help='save tracking results in a txt file')
parser.add_argument('--save-id-crops', action='store_true',
help='save each crop to its respective id folder')
parser.add_argument('--save-mot', action='store_true',
help='save tracking results in a single txt file')
parser.add_argument('--line-width', default=None, type=int,
help='The line width of the bounding boxes. If None, it is scaled to the image size.')
parser.add_argument('--per-class', default=False, action='store_true',
help='not mix up classes when tracking')
parser.add_argument('--verbose', default=True, action='store_true',
help='print results per frame')
parser.add_argument('--agnostic-nms', default=False, action='store_true',
help='class-agnostic NMS')
parser.add_argument('--benchmark', type=str, default='MOT20',
help='MOT16, MOT17, MOT20')
parser.add_argument('--split', type=str, default='train',
help='existing project/name ok, do not increment')

opt = parser.parse_args()
return opt

def run_trackeval(opt):
if opt is None:
opt = parse_opt()
exp_folder_path = opt.project / (str(opt.dets) + "" + str(opt.embs) + "" + str(opt.tracking_method))
exp_folder_path = increment_path(path=exp_folder_path,mkdir=True, sep="_", exist_ok=opt.exist_ok)
opt.exp_folder_path = exp_folder_path
else:
opt = opt
opt.exist_ok = False

val_tools_path = EXPERIMENTATION / 'val_utils'
download_mot_eval_tools(val_tools_path)
zip_path = download_mot_dataset(val_tools_path, opt.benchmark)
unzip_mot_dataset(zip_path, val_tools_path, opt.benchmark)
seq_paths, save_dir, MOT_results_folder, gt_folder = eval_setup(opt, val_tools_path)
results = trackeval(opt, seq_paths, save_dir, MOT_results_folder, gt_folder)
combined_results = parse_mot_results(results)
print(combined_results)
return combined_results

if name == "main":
run_trackeval(None)

[XDUbeginner]

Pictures of the above bugs

[XDUbeginner]

[github-actions]

👋 Hello, this issue has been automatically marked as stale because it has not had recent activity. Please note it will be closed if no further activity occurs.
Feel free to inform us of any other issues you discover or feature requests that come to mind in the future. Pull Requests (PRs) are also always welcomed!