Location: Zoom (online) https://purdue-edu.zoom.us/j/99326083199?pwd=TjhqaE1TYkZ2RG1RMkt6S1JZbGk1UT09

Agenda: 3:30pm-4:45pm (Seminar + Q&A)

Automatic Parameters Tuning for Autonomous Driving System

Abstract

Thousands of parameters need to be tuned across the autonomy stack and beyond for the autonomous driving systems. It is challenging to balance various requirements such as safety, efficiency, and comfort of autonomous driving. In this talk, we will share an automatic parameters tuning system for autonomous driving systems with Bayesian optimization. Particularly, a generic framework is proposed for parameter tuning of different use cases, which is scalable with a large number of scenes, such as prediction and planning, to conduct the auto-tuning and validation. In addition, a modularized architecture is developed with a user-friendly end-to-end interface for experiments, visualization, and analysis. In this talk, we will also share the information of the purpose-built, fully autonomous, all-electric passenger vehicle at Zoox.