CPM-Triplet-loss

• Introduction: This project implements the CPM-Triplet loss for person re-identification. Our work is based on : Bag of Tricks (paper and official code) and light-reid (official code).

Dependencies

• Anaconda (Python 3.7)
• PyTorch 1.1.0
• PrettyTable (pip install prettytable)
• GPU Memory >= 10G
• Memory >= 10G

Dataset Preparation

• Market-1501 (Project, Google Drive)
• DukeMTMC-reID (Project, Google Drive)
• MSMT17 (Project, Paper)

Run

Train on Market-1501/DukeMTMC-reID/MTMC17

python3 main.py --mode train \
    --train_dataset market --test_dataset market \
    --market_path /path/to/market/dataset/ \
    --output_path ./results/market/ 
python3 main.py --mode train \
    --train_dataset duke --test_dataset duke \
    --duke_path /path/to/duke/dataset/ \
    --output_path ./results/duke/
python3 main.py --mode train \
    --train_dataset msmt --test_dataset msmt --steps 400 --pid_num 1041 \
    --duke_path /path/to/msmt/dataset/ \
    --output_path ./results/msmt/

Train with ResNet50-IBNa backbone

# download model to ./core/nets/models/ from https://drive.google.com/file/d/1_r4wp14hEMkABVow58Xr4mPg7gvgOMto/view
python3 main.py --mode train -cnnbackbone res50ibna \
    --train_dataset market --test_dataset market \
    --market_path /path/to/market/dataset/ \
    --output_path ./results/market/ 

Train with OSNet-AIN backbone

# download model to ./core/nets/models/ from https://mega.nz/#!YTZFnSJY!wlbo_5oa2TpDAGyWCTKTX1hh4d6DvJhh_RUA2z6i_so
python3 main.py --mode train -cnnbackbone osnetain \
    --train_dataset market --test_dataset market \
    --market_path /path/to/market/dataset/ \
    --output_path ./results/market/ 

Test on Market-1501/DukeMTMC-reID/MTMC-17

python3 main.py --mode test \
    --train_dataset market --test_dataset market \
    --market_path /path/to/market/dataset/ \
    --resume_test_model /path/to/trained/model.pkl \ 
    --output_path ./results/test-on-market/
python3 main.py --mode test \
    --train_dataset duke --test_dataset duke \
    --market_path /path/to/duke/dataset/ \
    --resume_test_model /path/to/trained/model.pkl \ 
    --output_path ./results/test-on-duke/
python3 main.py --mode test \
    --train_dataset msmt --test_dataset msmt --pid_num 1041 \
    --market_path /path/to/msmt/dataset/ \
    --resume_test_model /path/to/trained/model.pkl \ 
    --output_path ./results/test-on-msmt/

Visualize Market-1501/DukeMTMC-reID

python3 main.py --mode visualize --visualize_mode inter-camera \
    --train_dataset market --visualize_dataset market \
    --market_path /path/to/market/dataset/ \
    --resume_visualize_model /path/to/trained/model.pkl \ 
    --visualize_output_path ./results/vis-on-market/ 
python3 main.py --mode visualize --visualize_mode inter-camera \
    --train_dataset duke --visualize_dataset duke \
    --market_path /path/to/duke/dataset/ \
    --resume_visualize_model /path/to/trained/model.pkl \ 
    --visualize_output_path ./results/vis-on-duke/ 

Visualize Customed Dataset with Trained Model

# customed dataset structure
|____ data_path/
     |____ person_id_1/
          |____ pid_1_imgid_1.jpg
          |____ pid_1_imgid_2.jpg
          |____ ......
     |____ person_id_2/
     |____ person_id_3/
     |____ ......

python3 demo.py \
    --resume_visualize_model /path/to/pretrained/model.pkl \
    --query_path /path/to/query/dataset/ --gallery_path /path/to/gallery/dataset/ \
    --visualize_output_path ./results/vis-on-cus/

Experiments

Please select CPM-Triplet loss and Triplet loss in core/base.py.

Contacts

If you have any question about the project, please contact with me.

E-mail: weiyu_zeng@foxmail.com