Loading ...
60%

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

1Shanghai AI Laboratory, 2SenseTime Research
Note: Please use a browser other than Safari for HTML5 videos to play properly.

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.

Video Demo



Pose Interpolation

Our 3D-aware GAN can be used to render simple animation of generated humans by interpolating between poses.






BibTeX

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}
}