Urban Driving with Multi-Objective Deep Reinforcement Learning

By Krzysztof Czarnecki. 2018

86 citations recorded by the source corpus.

Autonomous driving is a challenging domain that entails multiple aspects: a vehicle should be able to drive to its destination as fast as possible while avoiding collision, obeying traffic rules and ensuring the comfort of passengers. In this paper, we present a deep learning variant of thresholded lexicographic Q-learning for the task of urban driving. Our multi-objective DQN agent learns to drive on multi-lane roads and intersections, yielding and changing lanes according to traffic rules. We also propose an extension for factored Markov Decision Processes to the DQN architecture that provides auxiliary features for the Q function. This is shown to significantly improve data efficiency. \footnoteData efficiency as measured by the number of training steps required to achieve similar performance. We then show that the learned policy is able to zero-shot transfer to a ring road without sacrificing performance.

Subfields: Artificial Intelligence, Automotive Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Automotive Engineering

Citations in the corpus

Deep Reinforcement Learning for Autonomous Driving: A Survey (2020, background): “As AD is essentially a multi-objective problem, methods from the ﬁeld multi-objective RL such as thresholded lexicographic ordering may be easily applied and have been demonstrated to work well (see e.g. [80]).”
Reward (Mis)design for Autonomous Driving (2021, background): “Application to AD Acknowledging the subjectivity of classifying whether reward is shaped when shaping is not explicitly discussed, we confidently judge that, of the 19 publications we surveyed, 13 included reward shaping via one or more attributes of their reward functions [13,29,24,32,53,9,21,48,37”