Pengxiang Li

Hi, You can choose the category you need, the current `qdtrack` also selected 482 categories. But there are bugs in the training of qdtrack on TAO. I will fix them...

I will paste the PR link here if I fix it.

https://github.com/open-mmlab/mmtracking/pull/609

You only need to modify JSON annotations. Filter the categories and corresponding annotations you need. You can refer to here. https://github.com/open-mmlab/mmtracking/blob/master/tools/convert_datasets/tao/tao2coco.py

Hi, lijinrun I think there is something wrong with your **data configuration file**. Please use the following configuration, which is equivalent to `mmtracking/configs/_base_/datasets/mot_challenge_det.py`. When we train `bytetrack`, we don't need...

> > > > Hi, do you use the latest version of mmdet? > > > > > > > > > My mmdet's version is 2.24. Maybe not all...

Hi, [moey920](https://github.com/moey920), In my opinion, there should be something wrong with your **annotation file**. I suggest you check the **"categories"** of the annotation file. In addition, you should also check...

Well, according to my understanding, you want to track UAVs, that is, there is only one type of target. Do I understand correctly？

I see you use `bytetrack`, so I don't recommend using **datatype** of `CocoVideoDataset`. You can take a look here https://github.com/open-mmlab/mmtracking/blob/2542848f4b441e92b5b7c54c32285d0e675dac1a/configs/mot/bytetrack/bytetrack_yolox_x_crowdhuman_mot17-private-half.py#L90 to change your own configuration file, because for bytetrack, we...

Of course, you can take a look at the issue🚀. https://github.com/open-mmlab/mmtracking/issues/564