Weichuang Li

I am currently a Ph.D student at HKUST(GZ) , where I am supervised by Prof. Yuxuan Liang. I was an research assistant at Shanghai Artificial Intelligence Laboratory, where I work on computer vision and neural rendering.

I got my Master degree at City Univerisity of Hong Kong, studying Computer Science.

Email  /  Github

profile photo
Research

My research interests include computer vision, neural rendering, and video processing.

Real3D-Portrait: One-shot Realistic 3D Talking Portrait Synthesis
Zhenhui Ye, Tianyun Zhong, Yi Ren, Jiaqi Yang, Weichuang Li, Jiawei Huang, Ziyue Jiang, Jinzheng He, Rongjie Huang, Jinglin Liu, Chen Zhang, Xiang Yin, Zejun MA, Zhou Zhao
ICLR, 2024(Spotlight)  
Project page / Paper

Real3D-Portrait is a framework that enhances one-shot 3D face reconstruction, accurate motion-conditioned animation, and audio-driven talking face generation, producing more realistic and versatile talking portrait videos than prior methods.

Affordance-Driven Next-Best-View Planning for Robotic Grasping
Xuechao Zhang, Dong Wang, Sun Han, Weichuang Li, Bin Zhao, Zhigang Wang, Xiaoming Duan, Chongrong Fang, Xuelong Li, Jianping He
CoRL, 2023  
Project page / Paper

We introduce an AffordanCE-driven Next-Best-View planning policy (ACE-NBV) that tries to find a feasible grasp for target object via continuously observing scenes from new viewpoints.

One-Shot High-Fidelity Talking-Head Synthesis with Deformable Neural Radiance Field
Weichaung Li, Longhao Zhang, Dong Wang, Bin Zhao, Zhigang Wang, Mulin Chen, Bang Zhang, Zhongjian Wang, Liefeng Bo, Xuelong Li
CVPR, 2023  
Project page / arXiv

We propose a deformable NeRF that is capable of generating high-fidelity talking portraits with single image input.

Detection of GAN-Generated Images by Estimating Artifact Similarity
Weichuang Li, Peisong He, Haoliang Li, Hongxia Wang, Ruimei Zhang
IEEE Signal Processing Letters (SPL), 2021
Poster / PDF / Code

We formulate the task of GAN-generated image detection as a problem of estimating how artifacts of the suspicious image are similar to GAN-generated images.


Thank JonBarron for this awesome template.