[Paper] [Project]

This repository contains the implementation of Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID, which provides state-of-the-art performances on both unsupervised domain adaptation tasks and unsupervised learning tasks for object re-ID, including person re-ID and vehicle re-ID.

[2020-07-01] We did the code refactoring to support distributed training, stronger performances and more features. Please see OpenUnReID.

git clone https://github.com/yxgeee/SpCL.git
cd SpCL
python setup.py install

cd examples && mkdir data

Download the person datasets DukeMTMC-reID, Market-1501, MSMT17, PersonX, and the vehicle datasets VehicleID, VeRi-776, VehicleX. Then unzip them under the directory like

SpCL/examples/data
├── dukemtmc
│   └── DukeMTMC-reID
├── market1501
│   └── Market-1501-v15.09.15
├── msmt17
│   └── MSMT17_V1
├── personx
│   └── PersonX
├── vehicleid
│   └── VehicleID -> VehicleID_V1.0
├── vehiclex
│   └── AIC20_ReID_Simulation -> AIC20_track2/AIC20_ReID_Simulation
└── veri
    └── VeRi -> VeRi_with_plate

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 logs/pretrained/.

mkdir logs && cd logs
mkdir pretrained

The file tree should be

SpCL/logs
└── pretrained
    └── resnet50_ibn_a.pth.tar

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

We utilize 4 GTX-1080TI GPUs for training. Note that

• use --iters 400 (default) for DukeMTMC-reID, Market-1501 and PersonX datasets, and --iters 800 for MSMT17, VeRi-776, VehicleID and VehicleX 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.

To train the model(s) in the paper, run this command:

CUDA_VISIBLE_DEVICES=0,1,2,3 python examples/spcl_train_uda.py -ds $SOURCE_DATASET -dt $TARGET_DATASET --logs-dir $PATH_LOGS

Example #1: DukeMTMC-reID -> Market-1501

CUDA_VISIBLE_DEVICES=0,1,2,3 python examples/spcl_train_uda.py -ds dukemtmc -dt market1501 --logs-dir logs/spcl_uda/duke2market_resnet50

Example #2: DukeMTMC-reID -> MSMT17

CUDA_VISIBLE_DEVICES=0,1,2,3 python examples/spcl_train_uda.py -ds dukemtmc -dt msmt17 --iters 800 --logs-dir logs/spcl_uda/duke2msmt_resnet50

Example #3: VehicleID -> VeRi

CUDA_VISIBLE_DEVICES=0,1,2,3 python examples/spcl_train_uda.py -ds vehicleid -dt veri --iters 800 --height 224 --width 224 --logs-dir logs/spcl_uda/vehicleid2veri_resnet50

To train the model(s) in the paper, run this command:

CUDA_VISIBLE_DEVICES=0,1,2,3 python examples/spcl_train_usl.py -d $DATASET --logs-dir $PATH_LOGS

Example #1: DukeMTMC-reID

CUDA_VISIBLE_DEVICES=0,1,2,3 python examples/spcl_train_usl.py -d dukemtmc --logs-dir logs/spcl_usl/duke_resnet50

We utilize 1 GTX-1080TI GPU for testing. Note that

• use --width 128 --height 256 (default) for person datasets, and --height 224 --width 224 for vehicle datasets;
• use --dsbn for domain adaptive models, and add --test-source if you want to test on the source domain;
• use -a resnet50 (default) for the backbone of ResNet-50, and -a resnet_ibn50a for the backbone of IBN-ResNet.

To evaluate the model on the target-domain dataset, run:

CUDA_VISIBLE_DEVICES=0 python examples/test.py --dsbn -d $DATASET --resume $PATH_MODEL

To evaluate the model on the source-domain dataset, run:

CUDA_VISIBLE_DEVICES=0 python examples/test.py --dsbn --test-source -d $DATASET --resume $PATH_MODEL

Example #1: DukeMTMC-reID -> Market-1501

# test on the target domain
CUDA_VISIBLE_DEVICES=0 python examples/test.py --dsbn -d market1501 --resume logs/spcl_uda/duke2market_resnet50/model_best.pth.tar
# test on the source domain
CUDA_VISIBLE_DEVICES=0 python examples/test.py --dsbn --test-source -d dukemtmc --resume logs/spcl_uda/duke2market_resnet50/model_best.pth.tar

To evaluate the model, run:

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

Example #1: DukeMTMC-reID

CUDA_VISIBLE_DEVICES=0 python examples/test.py -d dukemtmc --resume logs/spcl_usl/duke_resnet50/model_best.pth.tar

You can download the above models in the paper from Google Drive.

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

@misc{ge2020selfpaced,
    title={Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID},
    author={Yixiao Ge and Dapeng Chen and Feng Zhu and Rui Zhao and Hongsheng Li},
    year={2020},
    eprint={2006.02713},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}