ST-GDance: Long-Term and Collision-Free Group Choreography from Music

Official implementation of ST-GDance
📢 Our work has been accepted at BMVC 2025 (Oral)

[arXiv] [Project Page]

First of all, please setup the virtual environment with Anaconda:

conda create -n stgdance python=3.7.12
conda activate stgdance
conda install pytorch=2.0.0 torchvision=0.15.0 torchaudio=2.0.0 pytorch-cuda=11.8 -c pytorch -c nvidia
pip install -r requirements.txt

We use the GDANCE dataset to train and evaluate our models. Please download and extract the data into ./data/gdance/. If you extract it to a different location, you will need to update the data path in the configuration accordingly.

Our model takes music features as input to generate corresponding dance motions. We provide pre-extracted Jukebox features derived from the GDANCE music sequences for your convenience.

If you find this work useful, please consider citing:

@inproceedings{Xu_2025_BMVC,
  author    = {Jing Xu and Weiqiang Wang and Cunjian Chen and Jun Liu and Qiuhong Ke},
  title     = {ST-GDance: Long-Term and Collision-Free Group Choreography from Music},
  booktitle = {36th British Machine Vision Conference 2025, {BMVC} 2025, Sheffield, UK, November 24-27, 2025},
  publisher = {BMVA},
  year      = {2025},
  url       = {https://bmva-archive.org.uk/bmvc/2025/assets/papers/Paper_66/paper.pdf}
}