Interaction-Dynamics-Aware Perception Zones for Obstacle Detection Safety Evaluation

TL;DR

This paper proposes a novel safety zone metric for obstacle detection in autonomous vehicles that accounts for dynamic interactions and safety-criticality, improving safety evaluation accuracy.

Contribution

It introduces an interaction-dynamics-aware safety zone based on Hamilton-Jacobi reachability, enhancing perception evaluation by considering agent interactions.

Findings

The safety zone better captures safety-critical perception errors.

The approach is computationally lightweight.

It outperforms baseline safety evaluation methods.

Abstract

To enable safe autonomous vehicle (AV) operations, it is critical that an AV's obstacle detection module can reliably detect obstacles that pose a safety threat (i.e., are safety-critical). It is therefore desirable that the evaluation metric for the perception system captures the safety-criticality of objects. Unfortunately, existing perception evaluation metrics tend to make strong assumptions about the objects and ignore the dynamic interactions between agents, and thus do not accurately capture the safety risks in reality. To address these shortcomings, we introduce an interaction-dynamics-aware obstacle detection evaluation metric by accounting for closed-loop dynamic interactions between an ego vehicle and obstacles in the scene. By borrowing existing theory from optimal control theory, namely Hamilton-Jacobi reachability, we present a computationally tractable method for constructing a ``safety zone'': a region in state space that defines where safety-critical obstacles lie for the purpose of defining safety metrics. Our proposed safety zone is mathematically complete, and can be easily computed to reflect a variety of safety requirements. Using an off-the-shelf detection algorithm from the nuScenes detection challenge leaderboard, we demonstrate that our approach is computationally lightweight, and can better capture safety-critical perception errors than a baseline approach.