This work introduces the new benchmark dataset, Café, and a new model for group activity detection (GAD).

• Ubuntu 20.04
• Python 3.8.5
• CUDA 11.0
• PyTorch 1.7.1

conda env create --file environment.yml

conda activate gad

pip install torch==1.7.1 cu110 torchvision==0.8.2 cu110 -f https://download.pytorch.org/whl/torch_stable.html

sh scripts/setup.sh

Download Café dataset from:

https://cvlab.postech.ac.kr/research/CAFE/

sh scripts/download_checkpoints.sh

or from:

https://drive.google.com/file/d/1W_2gkzARCzSdK8Db4G4pkzN3GrJTYo8R/view?usp=drive_link

• Café dataset (split by view)

    sh scripts/test_cafe_view.sh
  

• Café dataset (split by place)

    sh scripts/test_cafe_place.sh
  

• Café dataset (split by view)

    sh scripts/train_cafe_view.sh
  

• Café dataset (split by place)

    sh scripts/train_cafe_place.sh
  

├── Dataset/
│     └── cafe/
│           └── gt_tracks.pkl
├── dataloader/
├── evaluation/
│     └── gt_tracks.txt
├── label_map/
├── models/
├── scripts/
└── util/
train.py
test.py
environment.yml
README.md

If you find our work useful, please consider citing our paper:

@article{kim2023towards,
  title={Towards More Practical Group Activity Detection: A New Benchmark and Model},
  author={Kim, Dongkeun and Song, Youngkil and Cho, Minsu and Kwak, Suha},
  journal={arXiv preprint arXiv:2312.02878},
  year={2023}
}

This work was supported by the NRF grant and the IITP grant funded by Ministry of Science and ICT, Korea (RS-2019-II191906, IITP-2020-0-00842, NRF-2021R1A2C3012728, RS-2022-II220264).