My name is Jingyuan Li (李靖远). I am currently a Master’s student in Computer Science at ETH Zürich, specializing in Visual and Interactive Computing, where I work under the supervision of Prof. Siyu Tang and Prof. Olga Sorkine-Hornung.
My research focuses on deep learning and computer graphics, particularly on physics-based and kinematic character animation, motion generation, and controllable generative models.
Before joining ETH, I obtained my BSc in Computer Science from Harbin Institute of Technology, Shenzhen.
Stylized phase manifolds disentangle motion content, temporal structure, and style, enabling controllable and high-quality text-driven motion generation through a diffusion-based …
We propose CFDNet, a contrastive feature decomposition network for single image layer separation, which outperforms existing methods in reflection separation and intrinsic image …
