Learning to Drive using Inverse Reinforcement Learning and Deep Q-Networks

By Sahand Sharifzadeh, Ioannis Chiotellis, Rudolph Triebel, Daniel Cremers. 2016

96 citations recorded by the source corpus.

We propose an inverse reinforcement learning (IRL) approach using Deep Q-Networks to extract the rewards in problems with large state spaces. We evaluate the performance of this approach in a simulation-based autonomous driving scenario. Our results resemble the intuitive relation between the reward function and readings of distance sensors mounted at different poses on the car. We also show that, after a few learning rounds, our simulated agent generates collision-free motions and performs human-like lane change behaviour.

Subfields: Artificial Intelligence, Automotive Engineering, Cognitive Neuroscience

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

In Cognitive Neuroscience

Citations in the corpus

Deep Reinforcement Learning for Autonomous Driving: A Survey (2020, methodology): “Authors of [116] proposed to learn comfortable driving trajectories optimization using expert demonstration from human drivers using Maximum Entropy Inverse RL. Authors of [117] used DQN as the reﬁnement step in IRL to extract the rewards, in an effort learn human-like lane change behavior.”
Deep Reinforcement Learning for Intelligent Transportation Systems: A Survey (2020, background): “[114] DQN Autonomous driving Lane changing Personal simulator Expert driver”
Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art (2017, methodology): “Sharifzadeh et al. (2016) present a different paradigm based on Markov Decision Processes, for more promising performance when facing scenarios that are very different from the ones in the training data.”
A Survey of State-Action Representations for Autonomous Driving (2018, methodology): “This representation is used in (Hadsell et al., 2009; Cardamone et al., 2009; Sharifzadeh et al., 2016; Kue(cid:29)er et al., 2017; Pfei(cid:27)er et al., 2017a; Plessen, 2017; Bhattacharyya et al., 2018) 1.2.3 Lane-centric coordinates But one can observe that the tactical-level decision making…”
Distributed Multiagent Coordinated Learning for Autonomous Driving in Highways Based on Dynamic Coordination Graphs (2020, unknown): “behavior learning [15], [16].”
Deep Reinforcement Learning for Autonomous Driving (2018, methodology): “[17] achieve collision-free motion and human-like lane change behavior by using an inverse reinforcement learning approach.”
Continuous control for robot based on deep reinforcement learning (2019, methodology): “[53] adopted a DQN-based inverse reinforcement learning to acquire the rewards for the autonomous driving scenario.”
Exploring applications of deep reinforcement learning for real-world autonomous driving systems (2019, methodology): “Using Deep Q-Networks as the reﬁnement step in IRL is proposed in (Sharifzadeh et al., 2016) to extract the rewards.”
A Survey on Trajectory-Prediction Methods for Autonomous Driving (2022, background): “Sharifzadeh et al. [173] leverage IRL with Deep Q-Networks (DQN) to extract the rewards with large state spaces.”