Robotic Hearing Systems for Autonomous Vehicles

How do we teach an autonomous vehicle to hear? This is an interesting problem both because we expect our cars to hear as well as we do and keep us all safe. We must do it in busy city soundscapes and at 60 mph. All the while preserving the privacy of people not connected with the car.

Dr. Xuan Zhong is trained in hearing science and has been working for self-driving car companies for several years.

Who: Dr. Xuan Zhong
What: Building Ears for Robots: Machine Hearing in the Age of Autonomy
When: Friday October 27 at 10:30AM
Where: CCRMA Seminar Room
Why: We will have self-driving cars someday, and we want them to hear us

Whether you walk, bike or drive to CCRMA, we want our cars to hear us. How do we do that?

 

Title: Building Ears for Robots: Machine Hearing in the Age of Autonomy

Abstract:
What does sounds mean to robots? How do we build ears for them? The field of robot hearing is increasingly important due to the deployment of service robots that can operate in uncertain environments. In this seminar, we first discuss the hardware design considerations of robot hearing systems in the context of robotaxi. Then a preliminary framework for robot hearing software is introduced based on the theory of probabilistic robotics.

Bio:

Xuan Zhong is an engineering manager of audio signal processing on autonomous vehicles at Zoox. In 2015, he finished his PhD in hearing science at Arizona State University. The topic of his dissertation was "Dynamic spatial hearing by human and robot listeners".