StdGEN

📃 Paper • 💻 [Github Repo] • 🌐 Project Page

StdGEN is an innovative pipeline for generating semantically decomposed high-quality 3D characters from single images.

All Models

We open-sourced the following list of models:

Model	Description
StdGEN-canonicalize-1024	Convert reference character image to A-pose
StdGEN-multiview-1024	Convert A-pose image to multi-view images and normals
StdGEN-mesh-slrm	Semantic-aware Large Reconstruction Model

Citation

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

@article{he2024stdgen,
  title={StdGEN: Semantic-Decomposed 3D Character Generation from Single Images},
  author={He, Yuze and Zhou, Yanning and Zhao, Wang and Wu, Zhongkai and Xiao, Kaiwen and Yang, Wei and Liu, Yong-Jin and Han, Xiao},
  journal={arXiv preprint arXiv:2411.05738},
  year={2024}
}