Recent years have seen tremendous iterations on autonomous driving technologies, pushing the deployment of self-driving cars closer to its realization. As the experimental deployments scale, more challenging and less frequent corner cases surface to stress-test the reliability of the autonomous driving system. Examples of these corner cases include limited visibility due to occlusion, degraded perception at long range, transient reflection and so on. To address the limited visibility issue, in particular, cooperative perception has been proposed to leverage vehicle-to-everything (V2X) communication to share perception data with nearby vehicles to fill in the invisible area. In this talk, I will present a line of cooperative perception system research from its initial prototyping, and scaling up, to its expansion from perception to end-to-end driving behaviors.

Hang Qiu is an assistant professor at the Department of Electrical and Computer Engineering at the University of California, Riverside. Previously, he was a postdoctoral scholar in the Platform Lab at Stanford University, and a software engineer at Waymo LLC. He received his Ph.D. from the Department of Electrical and Computer Engineering at the University of Southern California and his Bachelor's degree from Shanghai Jiao Tong University. His research focus is on networked cyber-physical systems. His work draws upon theories and methods from machine learning, wireless networking, computer vision, and robotics to build robust and cooperative intelligence in edge autonomous systems. He is a recipient of the MLSys Outstanding Paper Award, ACM Mobisys Best Paper Runner-up Award, and COMAP ICM Outstanding Winner Award.