Toward Efficient and Robust Behavior Models for Multi-Agent Driving Simulation

TL;DR

This paper presents a scalable, efficient, and robust behavior model for multi-agent driving simulation using instance-centric representations and adversarial inverse reinforcement learning.

Contribution

It introduces an instance-centric scene representation and a query-centric encoder, improving efficiency and robustness in multi-agent driving simulations.

Findings

Reduces training and inference times significantly.

Outperforms agent-centric baselines in accuracy and robustness.

Scales efficiently with the number of tokens.

Abstract

Scalable multi-agent driving simulation requires behavior models that are both realistic and computationally efficient. We address this by optimizing the behavior model that controls individual traffic participants. To improve efficiency, we adopt an instance-centric scene representation, where each traffic participant and map element is modeled in its own local coordinate frame. This design enables efficient, viewpoint-invariant scene encoding and allows static map tokens to be reused across simulation steps. To model interactions, we employ a query-centric symmetric context encoder with relative positional encodings between local frames. We use Adversarial Inverse Reinforcement Learning to learn the behavior model and propose an adaptive reward transformation that automatically balances robustness and realism during training. Experiments demonstrate that our approach scales efficiently with the number of tokens, significantly reducing training and inference times, while outperforming several agent-centric baselines in terms of positional accuracy and robustness.