BiCnet-TKS: Learning Efficient Spatial-Temporal Representation for Video Person Re-Identification

Requirements: Python=3.8 and Pytorch=1.7.0

Training and test

# For MARS
python train.py -d mars --root your path to MARS --save_dir log-mars #
python test.py -d mars --root your path to MARS --save_dir log-mars #

Citation

If you use this code for your research, please cite our paper:

@inproceedings{BiCnet-TKS,
  title={BiCnet-TKS: Learning Efficient Spatial-Temporal Representation for Video Person Re-Identification},
  author={Ruibing Hou and Hong Chang and Bingpeng Ma and  Rui Huang and Shiguang Shan},
  booktitle={CVPR},
  year={2021}
}