DeFIX: Detecting and Fixing Failure Scenarios with Reinforcement Learning in Imitation Learning Based Autonomous Driving


Creative Commons License

Dagdanov R., Eksen F., Durmuş H., Yurdakul F., Üre N. K.

25th IEEE International Conference on Intelligent Transportation Systems, ITSC 2022, Macau, China, 8 - 12 October 2022, vol.2022-October, pp.4215-4220 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 2022-October
  • Doi Number: 10.1109/itsc55140.2022.9922209
  • City: Macau
  • Country: China
  • Page Numbers: pp.4215-4220
  • Keywords: Autonomous Driving, Imitation Learning, Reinforcement Learning
  • Istanbul Technical University Affiliated: Yes

Abstract

© 2022 IEEE.Safely navigating through an urban environment without violating any traffic rules is a crucial performance target for reliable autonomous driving. In this paper, we present a Reinforcement Learning (RL) based methodology to DEtect and FIX (DeFIX) failures of an Imitation Learning (IL) agent by extracting infraction spots and re-constructing mini -scenarios on these infraction areas to train an RL agent for fixing the shortcomings of the IL approach. DeFIX is a continuous learning framework, where extraction of failure scenarios and training of RL agents are executed in an infinite loop. After each new policy is trained and added to the library of policies, a policy classifier method effectively decides on which policy to activate at each step during the evaluation. It is demonstrated that even with only one RL agent trained on failure scenario of an IL agent, DeFIX method is either competitive or does out-perform state-of-the-art IL and RL based autonomous urban driving benchmarks. We trained and validated our approach on the most challenging map (Town05) of CARLA simulator which involves complex, realistic, and adversarial driving scenarios. The source code is publicly available at https://github.com/data-and-decision-lab/DeFIX