LMM-enhanced Safety-Critical Scenario Generation for Autonomous Driving System Testing From Non-Accident Traffic Videos

TL;DR

This paper introduces LEADE, a novel method leveraging Large Multimodal Models to generate and optimize safety-critical scenarios from real traffic videos, aiming to identify safety violations in autonomous driving systems during routine traffic conditions.

Contribution

LEADE automatically creates and refines traffic scenarios from videos to discover safety violations in ADSs, highlighting the importance of non-crash scenarios for safety testing.

Findings

LEADE effectively generates realistic safety-critical scenarios from traffic videos.

It uncovers safety violations in Level-4 ADS during routine traffic scenarios.

LEADE enhances scenario diversity for comprehensive ADS safety evaluation.

Abstract

Safety testing serves as the fundamental pillar for the development of autonomous driving systems (ADSs). To ensure the safety of ADSs, it is paramount to generate a diverse range of safety-critical test scenarios. While existing ADS practitioners primarily focus on reproducing real-world traffic accidents in simulation environments to create test scenarios, it's essential to highlight that many of these accidents do not directly result in safety violations for ADSs due to the differences between human driving and autonomous driving. More importantly, we observe that some accident-free real-world scenarios can not only lead to misbehaviors in ADSs but also be leveraged for the generation of ADS violations during simulation testing. Therefore, it is of significant importance to discover safety violations of ADSs from routine traffic scenarios (i.e., non-crash scenarios). We introduce LEADE, a novel methodology to achieve the above goal. It automatically generates abstract and concrete scenarios from real-traffic videos. Then it optimizes these scenarios to search for safety violations of the ADS in semantically consistent scenarios where human-driving worked safely. Specifically, LEADE enhances the ability of Large Multimodal Models (LMMs) to accurately construct abstract scenarios from traffic videos and generate concrete scenarios by multi-modal few-shot Chain of Thought (CoT). Based on them, LEADE assesses and increases the behavior differences between the ego vehicle and human-driving in semantic equivalent scenarios (here equivalent semantics means that each participant in test scenarios has the same behaviors as those observed in the original real traffic scenarios). We implement and evaluate LEADE on the industrial-grade Level-4 ADS, Apollo.