GaussiAnimate: Reconstruct and Rig Animatable Categories with Level of Dynamics
A scaffold–skin rigging system that compresses dynamic 4D shapes into controllable, expressive skelebones and reanimates them through Partwise Motion Matching.
Computer Graphics · 3D Vision
Jiaxin Wang
王佳鑫
PhD Student Westlake University
I am a PhD student at Westlake University, advised by Prof. Yuliang Xiu. I studied Computer Science at Hangzhou Dianzi University (2020–2024), where I began my research journey with Prof. Weichen Dai, before joining Westlake as a Research Assistant (2024–2026) under the joint mentorship of Yuliang Xiu, Anpei Chen, Cheng Lin, and Zhiyang Dou.
My work focuses on representations and reconstruction for 3D and 4D worlds. I am always open to new collaborations—and equally happy to exchange ideas about cooking.
01
Latest
News
GaussiAnimate has been accepted to the SIGGRAPH Asia 2026 Conference Track.
HashRF was published in Applied Sciences.
GGAvatar appeared at Pacific Graphics 2024.
02
Selected work
Publications
GGAvatar: Dynamic Facial Geometric Adjustment for Gaussian Head Avatar
A coarse-to-fine Gaussian avatar representation with neutral geometric initialization and a low-dimensional Geometry Morph Adjuster for detailed facial deformation.
Neural Radiance Fields with Hash-Low-Rank Decomposition
A compact radiance field that combines low-rank tensor decomposition with multi-resolution hash encoding to preserve fine detail using fewer parameters.
Open to collaboration