Introduction

This repo is official PyTorch implementation of Disco4D: Disentangled 4D Human Generation and Animation from a Single Image (CVPR2025).

Installation

Refer to install.md

Train

Prepare the dataset:

python lib/dataset/process_4ddress.py

Image-to-3D:

### Generate avatar from image
sh run_img_synthesis.sh 
# ### Animate avatar
sh run_animate.sh 
# ### Edit avatar
sh run_edit.sh 

Video-to-4D:

### Generate avatar from 4D-Dress clips
sh run_video_synthesis.sh 

Acknowledgement

This work is built on many amazing research works and open-source projects, thanks a lot to all the authors for sharing!

• splattingavatar
• gaussian-grouping
• dreamgaussian
• gaussian-splatting and diff-gaussian-rasterization
• threestudio
• nvdiffrast
• dearpygui

Citation

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

@inproceedings{
  title={Disco4D: Disentangled 4D Human Generation and Animation from a Single Image},
  author={Pang, Hui En and Liu, Shuai and Cai, Zhongang and Yang, Lei and Zhang, Tianwei and Liu, Ziwei},
  booktitle={CVPR},
  year={2025}
}

License

Distributed under the S-Lab License. See LICENSE for more information.

Acknowledgements

This research/project is supported by the National Research Foundation, Singapore under its AI Singapore Programme. This study is also supported by the Ministry of Education, Singapore, under its MOE AcRF Tier 2 (MOE-T2EP20221-0012), NTU NAP, and under the RIE2020 Industry Alignment Fund – Industry Collaboration Projects (IAF-ICP) Funding Initiative, as well as cash and in-kind contribution from the industry partner(s). We sincerely thank the anonymous reviewers for their valuable comments on this paper.