Camera-aware Label Refinement for Unsupervised Person Re-Identification

The official repository for Camera-aware Label Refinement for Unsupervised Person Re-Identification. We achieve state-of-the-art performances on purly unsupervised tasks for object re-ID.

Our unified framework

Requirements

Installation

git clone https://github.com/leeBooMla/CALR
cd CALR
python setup.py install

Prepare Datasets

Download the person datasets Market-1501,MSMT17,DukeMTMC-reID and the vehicle datasets VeRi-776. Then put them under a foler such as '/xx/xx/dataset' Then unzip them under the directory like

/xx/xx/dataset
├── market1501
│   └── Market-1501-v15.09.15
├── msmt17
│   └── MSMT17_V1
├── dukemtmcreid
│   └── DukeMTMC-reID
└── veri
    └── VeRi

Prepare ImageNet Pre-trained Models for IBN-Net

When training with the backbone of IBN-ResNet, you need to download the ImageNet-pretrained model from this link and save it under the path of examples/pretrained/.

ImageNet-pretrained models for ResNet-50 will be automatically downloaded in the python script.

Training

We utilize 4 GTX-2080TI GPUs for training.

• The training have two stages, we first conduct intra-camera training to save the local_clusters. Or you could directly our local results in Baidu Yun(password:2qwm);

• use --iters 200 (default) for Market1501, and --iters 400 for other datasets;

• use --width 128 --height 256 (default) for person datasets, and --height 224 --width 224 for vehicle datasets;

• use -a resnet50 (default) for the backbone of ResNet-50, and -a resnet_ibn50a for the backbone of IBN-ResNet.

examples:

inter-camera training

CUDA_VISIBLE_DEVICES=0,1,2,3 python intra_camera_training.py -b 256 -a resnet50 -d veri --iters 100 --epochs 20

inter-camera training

CUDA_VISIBLE_DEVICES=0,1,2,3 python inter_camera_training.py -b 256 -a resnet50 -d veri --iters 400 --epochs 50

Evaluation

We utilize 1 GTX-2080TI GPU for testing. Note that

To evaluate the model, run:

CUDA_VISIBLE_DEVICES=0 \
python test.py \
  -d $DATASET --resume $PATH

Results

You can download the above models in the paper from Baidu Yun (password: vc3h)

Citation

If you find this code useful for your research, please cite our paper

@misc{li2024cameraaware,
      title={Camera-aware Label Refinement for Unsupervised Person Re-identification}, 
      author={Pengna Li and Kangyi Wu and Wenli Huang and Sanping Zhou and Jinjun Wang},
      year={2024},
      eprint={2403.16450},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Acknowledgements

Thanks to Yixiao Ge for opening source of her excellent works SpCL.