search

ifzhang / ByteTrack

[ECCV 2022] ByteTrack: Multi-Object Tracking by Associating Every Detection Box
MIT License
4.8k stars 906 forks source link

Losing the Tracks even after the detection score is 0.9 #216

Closed PranvQT closed 2 years ago

PranvQT commented 2 years ago

I have a detector which works quite well for the videos that I have.

I am passing these detection to the BYTEtracker as mentioned here in the readme "Combining BYTE with other detectors" section.

from yolox.tracker.byte_tracker import BYTETracker tracker = BYTETracker(args) for image in images: dets = detector(image) online_targets = tracker.update(dets, info_imgs, img_size)

The detection scores are in the range of 0.8 to 0.95 but still when the human in the video moves , the tracker looses the track and assigns new id to that person.

Can you please tell me what I can do to stop this?

danial880 commented 1 year ago

@PranvQT how did you solve this issue?

PranvQT commented 1 year ago

I tuned the hyper-parameters and that helped..I don't remember exactly which one but there are about 5-6 parameters try changing them ..that should help

@PranvQT how did you solve this issue?

pax7 commented 1 year ago

@PranvQT - can you please paste the numbers you're using right now for those hyper-params?