Towards Efficient, Dynamic, Portable Immersive Media

Date: Feb 14, 11:45am-12:45pm

Abstract: In the realm of immersive media, neural field-based methods have rapidly advanced in tasks such as 3D reconstruction and free-viewpoint rendering, achieving remarkable success in the fields of computer vision and graphics. Notably, volumetric scene representations—led by Neural Radiance Fields (NeRF) and Gaussian Splatting—have enabled high-quality immersive rendering of natural scenes. In this talk, we will explore three critical dimensions of immersive rendering: rendering efficiency, dynamic scene handling, and portability. Additionally, we will share our explorations in these areas and offer insights into the future development of immersive images and videos.

Biographical Sketch: Dr. Zhong Li is a Senior Research Engineer at Apple Inc., focusing on Generative AI techniques. He holds a Ph.D. from the University of Delaware and previously served as a Senior Staff Research Scientist at OPPO US Research Center, where he advanced mobile AR/VR technologies. His research spans computer graphics, 3D vision, and computational imaging, with expertise in neural rendering, 3D Generative AI, image-based rendering, relighting, and modeling 2D/3D human faces and bodies. His work has been published in top venues such as TPAMI, TVCG, SIGGRAPH, CVPR, ICCV, ECCV, and NeurIPS. He received the Best Paper Award at SIGGRAPH VRCAI in 2019.

Location and Zoom link: Zoom only at https://fsu.zoom.us/j/97062665937