Zoom: https://illinois.zoom.us/j/86207829593?pwd=ci9BZU9xbms3ZnZmOTVFZGFoNG5Ldz09
Abstract:
Over the past decade, the emergence and development of spatial computing devices (e.g., Apple Vision Pro and Meta Quest 3) has been on the path of shaping and defining the future of human-computer interaction. The quest for passing the visual Turing test, where human observers cannot reliably distinguish between real and computer-generated content, has become a primary focus of R&D, recognized as one of the 10 grand challenges by the National Academy of Engineering. In this talk, I'll explore how physics-guided machine learning, innovative display design, differentiable rendering, and generative models can address key obstacles in the current experience. I will delve into creating true 3D displays that go beyond 200-year-old stereoscopic techniques, eliminating the visual discomfort that haunts existing devices. Then, I'll discuss leveraging these displays to offer a universal perception experience for various vision conditions, eliminating the need for extra prescription glass and enabling eyeglass-form-factor devices. Lastly, I will discuss expertise-free modeling of real-world appearances in digital environments at high quality. Throughout, I'll highlight the broad applications of the developed technologies, from biomedical to manufacturing, sensing, and scientific imaging, ushering in new possibilities for science and engineering innovation.
Bio:
Liang Shi is a final year PhD at MIT Computer Science and Artificial Intelligence Laboratory (CSAIL), working with Professor Wojciech Matusik. He obtained his M.S. in Electrical Engineering at Stanford University working with Prof. Gordon Wetzstein. His research interest lies in computer graphics and computational displays/imaging, with a focus on improving 3D user experiences in next-generation spatial computing devices. He is a recipient of the Snap Fellowship, Meta Fellowship, Rising Star of Light by Nature Light, and Sony Focused Research Award.
Faculty Host: Shenlong Wang
Meeting ID: 862 0782 9593; Password: csillinois
