[ECCV 2024] MacDiff: Unified Skeleton Modeling with Masked Conditional Diffusion

[📝 Paper] [📄 Supplementary] [🚀 Website]

🎉 Accepted by ECCV 2024

Installation

conda create -n macdiff python=3.8

conda activate macdiff

pip install -r requirements.txt

Data Preparation

Please follow the data preparation process of MAMP.

Training and Testing

Please refer to the bash scripts. Before running the scripts, you may:

1. Check the default configurations in the yaml files. You can also overwrite them in the bash scripts.

2. Change the paths for saving your checkpoints and logs.

If you find any problems with the code, please feel free to open an issue or contact us by sending an email to aladonwlh[AT]stu.pku.edu.cn.

Citation

If you find this work useful for your research, please consider citing our work:

@inproceedings{wu2024macdiff,
  title={MacDiff: Unified Skeleton Modeling with Masked Conditional Diffusion},
  author={Wu, Lehong and Lin, Lilang and Zhang, Jiahang and Ma, Yiyang and Liu, Jiaying},
  booktitle={European Conference on Computer Vision (ECCV)},
  year={2024}
}

Acknowledgment

The framework of our code is based on MAE, MAMP.