search

ifzhang / ByteTrack

[ECCV 2022] ByteTrack: Multi-Object Tracking by Associating Every Detection Box
MIT License
4.69k stars 891 forks source link

I've encountered while using the YOLOX trained model for object tracking. #378

Open tericalpha opened 8 months ago

tericalpha commented 8 months ago

@ifzhang @Johnqczhang i had trained the head model on YOLOX and added that model in ByteTracker pretraied model folder and ran the below command !python3 tools/demo_track.py video -f exps/default/yolox_s.py -c /content/drive/MyDrive/epoch_99_ckpt.pth --fp16 --fuse --path /content/ByteTrack/videos/palace.mp4 --save_result &> log.txt Not getting the bounding boxes on head. But when i ran the source YOLOX object detection demo , am getting the bounding boxes with the same model. Any idea , what am missing here.

yaoshanliang commented 4 months ago

@tericalpha same question. How did you solve that?