Jinkun Cao

@hhhhhhogan I have merged the oc-sort support into the master branch now. Welcome to have a try!

Hi @ZihaoZhao @corrado113 , I tried to reproduce your issue. But the same command runs smoothly on my machine with: CUDA: 11.2 torch: 1.6.0 torchvision: 0.7.0 This issue is usually...

@AndrewGuo0930 To evaluate on multiple classes, you have to customize the `CLASS_TO_EVAL` in TrackEval as [here](https://github.com/JonathonLuiten/TrackEval/blob/bcd03a6cc5f4fa0074da89d95e859fe77e264c3e/trackeval/datasets/mot_challenge_2d_box.py#L24). The corresponding line to edit is in the `mmtrack/datasets/mot_challenge_dataset.py` as [here](https://github.com/open-mmlab/mmtracking/blob/059dc9988791d3afdc65b8c291ae17ee9631b723/mmtrack/datasets/mot_challenge_dataset.py#L278). For other problems,...

Moreover, StrongSORT emphasizes on advanced post-processing methods, which can be developed as a shared module for general MOT algorithms here.

@austinmw Also, you can post this expectation in [this roadmap discussion issue](https://github.com/open-mmlab/mmtracking/issues/73).

hi @austinmw , I believe the author of StrongSORT is contributing to this repo here (https://github.com/open-mmlab/mmtracking/pull/630). You may want to keep an eye on its progress.

I'm still receiving this error. Hope someone could help to fix it.

The easiest way might be aligning your annotation format with MOT and then defining your own dataset like what Dancetrack dataset does upon MOT.

We didn't make that as we originally target for the pedestrian-only tracking datasets, i.e. MOT datasets. But that might be trivial to support with minor modification: 1. Following the DETR...

You can do this by using the detection read from existing txt files instead of from the detector. I don't the current ByteTrack supports this but that would be trivial...