Learning Interactive Driving Policies via Data-driven Simulation

TL;DR

This paper introduces a data-driven simulation method using in-painted adversarial vehicles to enhance the learning of robust interactive driving policies, enabling direct transfer to real autonomous vehicles without traditional sim-to-real techniques.

Contribution

The paper proposes a novel simulation approach with in-painted adversarial vehicles to improve multi-agent interaction learning in driving policies.

Findings

Policies trained with the method transfer directly to real vehicles.

The approach enhances robustness in interactive driving scenarios.

No traditional sim-to-real transfer methods are needed.

Abstract

Data-driven simulators promise high data-efficiency for driving policy learning. When used for modelling interactions, this data-efficiency becomes a bottleneck: Small underlying datasets often lack interesting and challenging edge cases for learning interactive driving. We address this challenge by proposing a simulation method that uses in-painted ado vehicles for learning robust driving policies. Thus, our approach can be used to learn policies that involve multi-agent interactions and allows for training via state-of-the-art policy learning methods. We evaluate the approach for learning standard interaction scenarios in driving. In extensive experiments, our work demonstrates that the resulting policies can be directly transferred to a full-scale autonomous vehicle without making use of any traditional sim-to-real transfer techniques such as domain randomization.