Learning to Drive in a Day

TL;DR

This paper presents a pioneering application of deep reinforcement learning enabling a vehicle to learn lane following from scratch within a few episodes using only monocular images, without predefined rules or supervision.

Contribution

It introduces a novel on-vehicle deep reinforcement learning framework for autonomous driving that learns a driving policy quickly with minimal prior information.

Findings

Successfully learned lane following in a few episodes

Used a simple reward based on distance traveled without intervention

Demonstrated feasibility of RL for autonomous driving without explicit rules

Abstract

We demonstrate the first application of deep reinforcement learning to autonomous driving. From randomly initialised parameters, our model is able to learn a policy for lane following in a handful of training episodes using a single monocular image as input. We provide a general and easy to obtain reward: the distance travelled by the vehicle without the safety driver taking control. We use a continuous, model-free deep reinforcement learning algorithm, with all exploration and optimisation performed on-vehicle. This demonstrates a new framework for autonomous driving which moves away from reliance on defined logical rules, mapping, and direct supervision. We discuss the challenges and opportunities to scale this approach to a broader range of autonomous driving tasks.