🤖 AI Summary
Autonomous driving systems underperform in safety-critical scenarios due to data scarcity and the complexity of modeling multi-vehicle interactions. To address this, we propose the first game-theoretic simulation framework tailored for multi-agent safety-critical traffic scenarios. Our approach models road driving as a risk-controllable multi-vehicle game, integrating approximate Nash equilibrium computation with risk-sensitive constrained optimization, and introducing distribution-alignment regularization to enhance scenario authenticity and adversarial exploitability. Unlike conventional data-driven methods, our framework overcomes sparsity and staticity limitations, enabling high-fidelity, diverse, and dynamically adjustable-risk scenario generation. Extensive evaluation on multiple real-world datasets demonstrates superior fidelity, strong adversarial robustness (i.e., “breakability”), and cross-scenario generalization. We further release an interactive demonstration platform to facilitate reproducibility and community adoption.
📝 Abstract
While modern Autonomous Vehicle (AV) systems can develop reliable driving policies under regular traffic conditions, they frequently struggle with safety-critical traffic scenarios. This difficulty primarily arises from the rarity of such scenarios in driving datasets and the complexities associated with predictive modeling among multiple vehicles. To support the testing and refinement of AV policies, simulating safety-critical traffic events is an essential challenge to be addressed. In this work, we introduce TrafficGamer, which facilitates game-theoretic traffic simulation by viewing common road driving as a multi-agent game. In evaluating the empirical performance across various real-world datasets, TrafficGamer ensures both fidelity and exploitability of the simulated scenarios, guaranteeing that they not only statically align with real-world traffic distribution but also efficiently capture equilibriums for representing safety-critical scenarios involving multiple agents. Additionally, the results demonstrate that TrafficGamer exhibits highly flexible simulation across various contexts. Specifically, we demonstrate that the generated scenarios can dynamically adapt to equilibriums of varying tightness by configuring risk-sensitive constraints during optimization. To the best of our knowledge, TrafficGamer is the first simulator capable of generating diverse traffic scenarios involving multiple agents. We have provided a demo webpage for the project at https://qiaoguanren.github.io/trafficgamer-demo/.