Dr. Changyin (CY) Zhou, co-founder and CTO of Visbit, is a tech visionary and entrepreneur with more than a decade of experience in envisioning and developing new cameras and vision technology. He led the R&D effort of the Visbit View Optimized Streaming (VVOS) technology. Prior to Visbit, he worked at Google X Lab as a Senior Software Engineer and at Google Research as a Senior Scientist. At Google X Lab, he was one of the founding members of the Gcam team and worked on projects including the Google Glass camera, Android camera HDR+ mode, and other still-confidential computational camera projects. Prior to Google, he worked at Microsoft Research and NVIDIA Research. Dr. Zhou holds a PhD in Computer Science from Columbia University and holds over 30 patents in his field. He has also published articles in leading academic journals and has given speeches at some of the most renowned computer vision/graphic conferences, including Siggraph, ICCV, CVPR and ICCP.
8K resolution in VR videos yields a clarify of 20 pixels per degree for end users, which is often referred to as the bottom line of VR experience. We witness the announcements of various 8K VR solutions in 2019. Yet 8K VR videos pose significant challenges to not only video encoding on camera side, video transcoding on cloud, CDN, but also to video decoding and rendering on client side. In this talk, we will discuss three questions that are critical to user experience: 1) How to evaluate the final clarity that a system delivers? 2) How to measure Motion-to-photon latency, which is the #1 quality metric in VR? 3) As foveated streaming and rendering being one key solution to numerous VR challenges, how should we evaluate its efficiency?