ML Techniques for Autonomous Driving

SystemX Affiliates: login to view related content.

Topic: 
ML Techniques for Autonomous Driving
Thursday, May 12, 2022 - 5:30pm to 6:30pm
Venue: 
Shriram 104
Speaker: 
Chen Wu - Waymo
Abstract / Description: 

*To receive email announcements and live stream information for upcoming seminars, please subscribe to the SystemX Seminar/EE310 Mailing list here.

Autonomous driving is a dynamically evolving field with rich ML research and development. It poses a variety of real world challenges to the effectiveness and scalability of ML algorithms. In this talk, I will give a quick overview of the ML techniques used to build a highly performing autonomous system, from modern model architectures that solve Perception problems across different sensor modalities, to behavior prediction, motion planning, and simulation.

However, we cannot stop at building the system once. As a real product and to serve many more people in many more geographies, the Waymo driver needs to learn quickly from new data and continuously expand its capability to drive in more environments safely. I will touch upon some ML techniques we develop at Waymo that enable scalable learning, including autolabeling, data augmentation, and sensor simulation.

Bio: 

Chen Wu is an Engineering Director and Head of Perception at Waymo, a self-driving technology company with a mission to make it safe and easy for people and things to move around. The Perception team is a group of talented engineers and researchers who develop state-of-the-art ML techniques and turn them into a real-time Perception system that runs on the Waymo vehicles. Prior to Waymo, Chen worked on the cameras on Google Glass. Before that, Chen was at YouTube where she used machine learning to enable 2D videos to be viewed in 3D. Chen holds a Ph.D. and M.S. in electrical engineering from Stanford University, and a B.S. in control theory from Tsinghua University in China.