search

ymzis69 / HybridSORT

[AAAI2024]Hybrid-SORT: Weak Cues Matter for Online Multi-Object Tracking
MIT License
148 stars 21 forks source link

How can we use official YOLOX as the detector? #30

Open yaoshanliang opened 1 month ago

yaoshanliang commented 1 month ago

I trained a weight from the official YOLOX (https://github.com/Megvii-BaseDetection/YOLOX), and put it in the self.ckpt. However, when I run the evaluation code using run_hybrid_sort_dance.py, the generated detection results are empty. It works well when I used the trained weight using tools/train.py from HybridSORT.

Can we use the trained weight from official YOLOX? Or some other modifications should be made.

ymzis69 commented 1 month ago

Hello, our weights are derived from ByteTrack and OC-SORT, trained on a pedestrian dataset with only one class, whereas the official YOLOX weights are trained on 80 classes. Our code currently does not support multi-class tracking. If you need to perform multi-class tracking, perhaps you can refer to this repository: https://github.com/CaptainEven/MCMOT-ByteTrack.