Not All Agents Matter: From Global Attention Dilution to Risk-Prioritized Game Planning

TL;DR

This paper introduces GameAD, a risk-aware game-theoretic framework for autonomous driving that prioritizes threat interactions, improving safety over existing models.

Contribution

It proposes a novel risk-prioritized game planning approach with new modules and metrics, advancing end-to-end autonomous driving safety.

Findings

Outperforms state-of-the-art methods on nuScenes and Bench2Drive datasets.

Significantly improves trajectory safety in autonomous driving scenarios.

Introduces the Planning Risk Exposure metric for long-term risk assessment.

Abstract

End-to-end autonomous driving resides not in the integration of perception and planning, but rather in the dynamic multi-agent game within a unified representation space. Most existing end-to-end models treat all agents equally, hindering the decoupling of real collision threats from complex backgrounds. To address this issue, We introduce the concept of Risk-Prioritized Game Planning, and propose GameAD, a novel framework that models end-to-end autonomous driving as a risk-aware game problem. The GameAD integrates Risk-Aware Topology Anchoring, Strategic Payload Adapter, Minimax Risk-Aware Sparse Attention, and Risk Consistent Equilibrium Stabilization to enable game theoretic decision making with risk prioritized interactions. We also present the Planning Risk Exposure metric, which quantifies the cumulative risk intensity of planned trajectories over a long horizon for safe autonomous driving. Extensive experiments on the nuScenes and Bench2Drive datasets show that our approach significantly outperforms state-of-the-art methods, especially in terms of trajectory safety.