Abstract
Forward rendering, a central topic in computer graphics, focuses on synthesizing physically accurate images of a digital scene by simulating light transport in the scene. Inverse rendering, on the contrary, is concerned with inferring scene data (such as the shape and optical property of an object) using images of the scene as input. With many real-world applications in VR/AR, computer graphics & vision, computational design, and computational fabrication, inverse rendering has been an active topic in various research communities.
Physics-based inverse rendering aims to recover scene parameters with the presence of complex light transport effects such as soft shadow and inter-reflection. Solving these problems using analysis-by-synthesis requires the use of differentiable rendering methods that estimate how the rendering of a scene alters with respect to differential changes of the scene.
In recent years, our group has made significant contributions to physics-based differentiable and inverse rendering. In this talk, I will present some of our works and discuss many remaining challenges.
This event will also be streamed and recorded here: https://go.illinois.edu/immerseseminar100225
Shuang Zhao (Associate Professor, Computer Science)
I am an Associate Professor at the Siebel School of Computing and Data Science of the University of Illinois Urbana-Champaign (UIUC). Before joining UIUC, I was an associate professor in the Department of Computer Science at the University of California, Irvine (UCI). I obtained my Ph.D. in Computer Science from Cornell University in 2014, under the advisement of Kavita Bala.
I received the NSF CAREER Award and a SIGGRAPH Asia Best Papers Award in 2023.