Domain-Enhanced Dual-Branch Model for Efficient and Interpretable Accident Anticipation

TL;DR

This paper presents a dual-branch accident anticipation model that combines visual dashcam data with textual reports, leveraging large multimodal models and prompt engineering to improve accuracy, efficiency, and interpretability in autonomous driving systems.

Contribution

It introduces a novel multimodal dual-branch framework with feature aggregation and prompt strategies, advancing accident prediction performance and interpretability.

Findings

Achieves superior accuracy on benchmark datasets

Reduces computational overhead compared to existing methods

Enhances interpretability of accident predictions

Abstract

Developing precise and computationally efficient traffic accident anticipation system is crucial for contemporary autonomous driving technologies, enabling timely intervention and loss prevention. In this paper, we propose an accident anticipation framework employing a dual-branch architecture that effectively integrates visual information from dashcam videos with structured textual data derived from accident reports. Furthermore, we introduce a feature aggregation method that facilitates seamless integration of multimodal inputs through large models (GPT-4o, Long-CLIP), complemented by targeted prompt engineering strategies to produce actionable feedback and standardized accident archives. Comprehensive evaluations conducted on benchmark datasets (DAD, CCD, and A3D) validate the superior predictive accuracy, enhanced responsiveness, reduced computational overhead, and improved interpretability of our approach, thus establishing a new benchmark for state-of-the-art performance in traffic accident anticipation.