日期：2023/12/08 - 2023/12/08
主讲人：Yifei Zhu, Assistant Professor, UM-SJTU Joint Institute, Shanghai Jiao Tong University时间：2023年12月8日（周五）上午10:00-11:30
The advancement of machine vision algorithms and the widespread deployment of network cameras have given rise to a significant demand for video analytics. Video analytics finds extensive applications in various domains such as entertainment, smart cities, transportation, and security. However, designing an efficient video analytics system still faces technological challenges such as dynamic networks, diverse tasks, and high computational resource consumption.
The conventional design patterns of single-node pure edge computing or cloud computing often fall short in meeting the diverse requirements of video analytics systems. This seminar will explore the significant advantages of collaborative video analytics and present three essential collaborative video analytics frameworks, cross-edge collaborative analysis, cloud-edge collaborative analysis, and cloud-side collaborative analysis. The goal is to elucidate the broad design space in collaborative video analytics and elaborate on its substantial benefits. We will complete the seminar by summarizing some common design principles and possible future directions.
Yifei Zhu received the B.E. degree from Xi’an Jiaotong University, Xian, China, in 2012, and the M.Phil. degree from The Hong Kong University of Science and Technology, Hong Kong, China, in 2015, and the Ph. D. degree in Computer Science from the Simon Fraser University, BC, Canada, in 2020. He is currently an assistant professor at UM-SJTU Joint Institute, Shanghai Jiao Tong University. He is also an affiliate assistant professor with Department of Electronic Engineering at Shanghai Jiao Tong University. His research interests include edge computing, multimedia networking, distributed machine learning system. He has published more than 50 papers at top conferences and journals including SIGCOMM, INFOCOM, Multimedia, ToN, TMC, etc. He served as a guest editor for IEEE Network and a TPC member in multiple major networking conferences, including IWQoS, ICC, etc.