H-TransTrack

Modified files

To support hybrid branch

• models/deformable_detrtrack_train_hybrid_branch.py

• models/deformable_transformer_track_hybrid_branch.py

• engine_track.py

• main_track.py

MODEL ZOO

CrowdHuman Pre-training

pretrained-model

MOT17 Validation

MOT17 Test

Requirements

• Linux, CUDA>=9.2, GCC>=5.4
• Python>=3.7
• PyTorch ≥ 1.5 and torchvision that matches the PyTorch installation. You can install them together at pytorch.org to make sure of this
• OpenCV is optional and needed by demo and visualization

2. Prepare datasets and annotations

   mkdir crowdhuman
   cp -r /path_to_crowdhuman_dataset/CrowdHuman_train crowdhuman/CrowdHuman_train
   cp -r /path_to_crowdhuman_dataset/CrowdHuman_val crowdhuman/CrowdHuman_val
   mkdir mot
   cp -r /path_to_mot_dataset/train mot/train
   cp -r /path_to_mot_dataset/test mot/test

   CrowdHuman dataset is available in CrowdHuman.

   python3 track_tools/convert_crowdhuman_to_coco.py

   MOT dataset is available in MOT.

   python3 track_tools/convert_mot_to_coco.py

3. Pre-train on crowdhuman

sh configs/.sh

4. Train H-TransTrack

sh configs/.sh

5. Evaluate TransTrack

sh configs/.sh

6. Visualize TransTrack

   python3 track_tools/txt2video.py

Citation

@article{jia2022detrs,
  title={DETRs with Hybrid Matching},
  author={Jia, Ding and Yuan, Yuhui and He, Haodi and Wu, Xiaopei and Yu, Haojun and Lin, Weihong and Sun, Lei and Zhang, Chao and Hu, Han},
  journal={arXiv preprint arXiv:2207.13080},
  year={2022}
}

@article{sun2020transtrack,
  title={Transtrack: Multiple object tracking with transformer},
  author={Sun, Peize and Cao, Jinkun and Jiang, Yi and Zhang, Rufeng and Xie, Enze and Yuan, Zehuan and Wang, Changhu and Luo, Ping},
  journal={arXiv preprint arXiv:2012.15460},
  year={2020}
}