Efficient Learning of Urban Driving Policies Using Bird'View State Representations

Raphael Trumpp, Martin Buchner, Abhinav Valada, Marco Caccamo

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Autonomous driving involves complex decision-making in highly interactive environments, requiring thoughtful negotiation with other traffic participants. While reinforcement learning provides a way to learn such interaction behavior, efficient learning critically depends on scalable state representations. Contrary to imitation learning methods, high-dimensional state representations still constitute a major bottleneck for deep reinforcement learning methods in autonomous driving. In this paper, we study the challenges of constructing bird's-eye-view representations for autonomous driving and propose a recurrent learning architecture for long-horizon driving. Our PPO-based approach, called RecurrDriveNet, is demonstrated on a simulated autonomous driving task in CARLA, where it outperforms traditional frame-stacking methods while only requiring one million experiences for efficient training. RecurrDriveNet causes less than one infraction per driven kilometer by interacting safely with other road users.

Original languageEnglish
Title of host publication2023 IEEE 26th International Conference on Intelligent Transportation Systems, ITSC 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages4181-4186
Number of pages6
ISBN (Electronic)9798350399462
DOIs
StatePublished - 2023
Event26th IEEE International Conference on Intelligent Transportation Systems, ITSC 2023 - Bilbao, Spain
Duration: 24 Sep 202328 Sep 2023

Publication series

NameIEEE Conference on Intelligent Transportation Systems, Proceedings, ITSC
ISSN (Print)2153-0009
ISSN (Electronic)2153-0017

Conference

Conference26th IEEE International Conference on Intelligent Transportation Systems, ITSC 2023
Country/TerritorySpain
CityBilbao
Period24/09/2328/09/23

Fingerprint

Dive into the research topics of 'Efficient Learning of Urban Driving Policies Using Bird'View State Representations'. Together they form a unique fingerprint.

Cite this