pmj110119 / yolox_deepsort_tracker Goto Github PK

img_visual, bbox = tracker.update(img) # feed one frame and get result ， img should be im。

Thanks a lot for your great work, really appreciate. I'm looking for a script which supports yolov4mish,csp,swish weights, Can you please do this if it's possible for you? A not single script is available on google/GitHub for yolov4xmish+deepsort.
Thanks looking forward to your reply.

(YOLOX_deepsort_tracker) sumyatnoe@MSI:/mnt/d/PycharmProjects/YOLOX_deepsort_tracker$ python demo.py --path=test.mp4
/mnt/d/PycharmProjects/YOLOX_deepsort_tracker/deep_sort/utils/parser.py:23: YAMLLoadWarning: calling yaml.load() without Loader=... is deprecated, as the default Loader is unsafe. Please
read https://msg.pyyaml.org/load for full details.
self.update(yaml.load(fo.read()))
/home/sumyatnoe/.local/lib/python3.8/site-packages/torch/functional.py:445: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Trigge
red internally at ../aten/src/ATen/native/TensorShape.cpp:2157.)
return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]

Traceback (most recent call last):
File "/home/下载/Pytorch_YOLOx_Deepsort-main/track.py", line 113, in
vdo_trk.run()
File "/home//下载/Pytorch_YOLOx_Deepsort-main/track.py", line 64, in run
info = self.detector.detect(im, visual=False)
File "/home/下载/Pytorch_YOLOx_Deepsort-main/YOLOX/detector.py", line 53, in detect
info['boxes'] = outputs[:, 0:4]/ratio
TypeError: 'NoneType' object is not subscriptable

May I know about how to train Deep Sort with Custom Dataset?

Hello, I have this error, could you please help me to find out the problem ?

File "track.py", line 33, in init
self.detector=build_detector(cfg,use_cuda=use_cuda)
File "\yolox-pytorch\detector.py", line 71, in build_detector
model= Detector(cfg.YOLOX.MODEL, cfg.YOLOX.WEIGHT)
File "\yolox-pytorch\detector.py", line 31, in init
self.exp = get_exp_by_name(model)
File "\yolox-pytorch\exp\build.py", line 36, in get_exp_by_name
return get_exp_by_file(exp_path)
File "\yolox-pytorch\exp\build.py", line 17, in get_exp_by_file
raise ImportError("{} doesn't contains class named 'Exp'".format(exp_file))
ImportError: \yolox-pytorch\exp\custom.py doesn't contains class named 'Exp

'
Thank you

visualize的时候我额外输出了一些信息，但是deepsort和yolox的代码我没改，我是每秒track一次，视频fps是25，可以看
到以下连续两次track（frame475和frame500），第二次画出来的框其实是前一次那辆车所在的位置，请问为什么会这样呢？有没有什么方法可以修正？谢谢。

hi , please can you help me
I was doing every step , but i face this problem in detector.py :
YOLOX_deepsort_tracker/detector.py", line 51, in detect
outputs = postprocess(
AttributeError: 'NoneType' object has no attribute 'cpu'

the code in detecto.py is :
with torch.no_grad():
outputs = self.model(img)
outputs = postprocess(
outputs, self.exp.num_classes, self.exp.test_conf, self.exp.nmsthre # TODO:用户可更改
)[0].cpu().numpy()

Thank you

When I run the demo on CPU only runs slower than using tinyYoloV4+DeepSort why is that? Is this repo configured to run optimized on GPU only?

non_max_suppression in deep_sort/sort/preprocessing behave strangely than conventional NMS

Yolox has a Apache 2.0 license. What is this ? Could it be used for commercial ?

Traceback (most recent call last):
File "demo.py", line 52, in
track_cap(args.path)
File "demo.py", line 34, in track_cap
image,_ = tracker.update(im)
File "G:\YOLOX_deepsort_tracker-master\tracker.py", line 27, in update
info = self.detector.detect(image, visual=False)
File "G:\YOLOX_deepsort_tracker-master\detector.py", line 55, in detect
info['boxes'] = outputs[:, 0:4]/ratio
TypeError: 'NoneType' object is not subscriptable

I create a new anaconda environment based on python 3.9.6.

Then I simply run pip install -r requirements.txt
I got the error message:

ERROR: Could not find a version that satisfies the requirement torchvision==0.10.0+cu111 (from versions: 0.1.6, 0.1.7, 0.1.8, 0.1.9, 0.2.0, 0.2.1, 0.2.2, 0.2.2.post2, 0.2.2.post3, 0.8.2, 0.9.0, 0.9.1, 0.10.0)
ERROR: No matching distribution found for torchvision==0.10.0+cu111

My system has the following specs:
CPU: Core I5, 11500
GPU: GTX1080
Nvidia driver version: 470.57.02,
CUDA version: 11.4

My task is use yolox to run object detect with YoloX, instead of tracking.
I trained yolox on many images with bboxes of car, then run car detection on a video.
But some of the frames are missing, may I use deepsort method to improve(postprocessing or offline or online?) the result of detection?
For example, predict the bbox when Yolox could not detect any car bbox in a frame (actually there is car)
Thanks.

File "/YOLOX_deepsort_tracker/detector.py", line 53, in detect
)[0].cpu().numpy()
AttributeError: 'NoneType' object has no attribute 'cpu'

.\YOLOX_deepsort_tracker\detector.py", line 24, in init
super(Detector, self).init()

TypeError: super(type, obj): obj must be an instance or subtype of type

How can i fix this error ?

Traceback (most recent call last):
File "/home/kerwin/KeTrack/demo.py", line 59, in
track_images(args.path)
File "/home/kerwin/KeTrack/demo.py", line 10, in track_images
tracker = Tracker(model='yolox-s', ckpt='weights/yolox_s.pth',filter_class=['person'])
File "/home/kerwin/KeTrack/tracker.py", line 18, in init
cfg.merge_from_file("deep_sort/configs/deep_sort.yaml")
File "/home/kerwin/KeTrack/deep_sort/utils/parser.py", line 23, in merge_from_file
self.update(yaml.load(fo.read()))
TypeError: load() missing 1 required positional argument: 'Loader'

Appreciate for your work!

did you test benchmarks(MOT15, MOT17...)

If the frame have no object, the code will fall.Please take care of it, thanks a lot