Increasing the Efficiency of Policy Learning for Autonomous Vehicles by Multi-Task Representation Learning

TL;DR

This paper introduces a multi-task representation learning approach that creates low-dimensional, informative environment representations for autonomous vehicle policy learning, leading to faster and more efficient reinforcement learning in complex urban scenarios.

Contribution

It proposes a multi-head encoder-decoder neural network to learn rich latent representations and a hazard signal to improve reinforcement learning efficiency for autonomous driving.

Findings

Faster policy learning with less data.

Enhanced performance over baseline methods.

More informative environment representations.

Abstract

Driving in a dynamic, multi-agent, and complex urban environment is a difficult task requiring a complex decision-making policy. The learning of such a policy requires a state representation that can encode the entire environment. Mid-level representations that encode a vehicle's environment as images have become a popular choice. Still, they are quite high-dimensional, limiting their use in data-hungry approaches such as reinforcement learning. In this article, we propose to learn a low-dimensional and rich latent representation of the environment by leveraging the knowledge of relevant semantic factors. To do this, we train an encoder-decoder deep neural network to predict multiple application-relevant factors such as the trajectories of other agents and the ego car. Furthermore, we propose a hazard signal based on other vehicles' future trajectories and the planned route which is used in conjunction with the learned latent representation as input to a down-stream policy. We demonstrate that using the multi-head encoder-decoder neural network results in a more informative representation than a standard single-head model. In particular, the proposed representation learning and the hazard signal help reinforcement learning to learn faster, with increased performance and less data than baseline methods.