3DHumanGAN: Towards Photo-realistic 3D-Aware Human Image Generation

Zhuoqian Yang, Shikai Li, Wayne Wu, Bo Dai
[Video Demo] | [Project Page] | [Technical Report]

Abstract: We present 3DHumanGAN, a 3D-aware generative adversarial network (GAN) that synthesizes images of full-body humans with consistent appearances under different view-angles and body-poses. To tackle the representational and computational challenges in synthesizing the articulated structure of human bodies, we propose a novel generator architecture in which a 2D convolutional backbone is modulated by a 3D pose mapping network. The 3D pose mapping network is formulated as a renderable implicit function conditioned on a posed 3D human mesh. This design has several merits: i) it allows us to harness the power of 2D GANs to generate photo-realistic images; ii) it generates consistent images under varying view-angles and specifiable poses; iii) the model can benefit from the 3D human prior. Our model is adversarially learned from a collection of web images needless of manual annotation.

TODOs

• Release technical report.
• Release code and pretrained models.
• Implement huggingface demos.
• Implement EG3D backend.

Citation

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

@article{yang20223dhumangan,
    title={3DHumanGAN: Towards Photo-realistic 3D-Aware Human Image Generation}, 
    author={Yang, Zhuoqian and Li, Shikai and Wu, Wayne and Dai, Bo},
    journal   = {arXiv preprint},
    volume    = {arXiv:2212.07378},
    year    = {2022}
}