Raif Olson

Hi @SkalskiP, I would be interested in working on this. I have a question about possible implementation, though. Would we want to keep the tracking algorithm the same and just...

Nvm, I think the best option is to only return tracks that have valid tracker ID's so we don't have to change how the annotators work.

Hi @SkalskiP Based on the warning at the beginning of the console output: ``` /Users/tobieabel/PycharmProjects/Demo_4_Fight_and_Fall/venv/lib/python3.11/site-packages/supervision/detection/utils.py:56: RuntimeWarning: invalid value encountered in divide return area_inter / (area_true[:, None] + area_detection - area_inter)...

@SkalskiP Sure!

There was a second bug that I found. In my original testing I was accidentally using model.track(frame) to get the detections from ultralytics. Then when I called sv.from_ultralytics each detection...

Oh, that's a good point. It would fix both the issue of returning detections that never actually are tracked, and the minimum_consecutive_frames arguement. Would you like me to add this...

Hi @SkalskiP I agree with both of those, I updated my [colab notebook](https://colab.research.google.com/drive/1azgc_A-divM3Z5l0TyHW2SRHnw6EUlbO?usp=sharing) to have more comprehensive tests, which should cover all of the different scenarios of detection objects being...

Also, I am wondering if we should add some unit tests for ByteTrack. I have some rudimentary one in a [colab notebook](https://colab.research.google.com/drive/1azgc_A-divM3Z5l0TyHW2SRHnw6EUlbO?usp=sharing). It would help when people need to test...

This is actually wrong. I forgot to test with low-confidence detections, and when that happens, the tracker_ids are -1, but the size of tracks returned from update_with_detections is none-zero, so...

Hi @Burhan-Q, I have some experience adding ReID to object trackers, and would be interested in giving this a try. Are you thinking of using a YOLO backbone for the...