Tzu-Mao Li

Tzu-Mao Li is an Assistant Professor in the Department of Computer Science and Engineering (CSE) at the University of California, San Diego (UCSD), affiliated with the Center for Visual Computing. His research focuses on differentiable graphics algorithms, combining classical visual computing with modern machine learning techniques. He holds a Ph.D. from MIT CSAIL under Frédo Durand and a postdoc at MIT and UC Berkeley with Jonathan Ragan-Kelley. His work spans rendering, programming languages for graphics, Monte Carlo methods, and inverse problems.

Education: B.S. and M.S. from National Taiwan University (2011-2013), advised by Yung-Yu Chuang. Ph.D. from MIT CSAIL (Computer Graphics Group), advised by Frédo Durand. Postdoctoral research at MIT and UC Berkeley with Jonathan Ragan-Kelley.

Research Interests: Differentiable rendering, Monte Carlo integration, programming language design for visual computing, physical simulation, adversarial machine learning, and applications in computer vision and robotics. Key areas include rendering algorithms (path tracing, bidirectional methods), optimization techniques (MCMC, gradient-based), and neural representations (SDFs, neural fields).

Publications focus on advancing rendering algorithms, differentiable systems, and applications in inverse problems. Notable contributions include edge sampling for differentiable rendering, warped-area sampling, and diffusion models for BSDF sampling.

Awards: ACM SIGGRAPH 2020 Outstanding Doctoral Dissertation Award, multiple Best Paper Awards at SIGGRAPH, and oral presentations at ICCV.

Teaching: Courses include CSE 167 (Computer Graphics), CSE 168 (Rendering), and CSE 272 (Advanced Image Synthesis), emphasizing physically-based methods and programming.