Real-time Traffic Accident Anticipation with Feature Reuse

TL;DR

This paper introduces RARE, a lightweight, real-time traffic accident anticipation framework that leverages pre-trained object detector features and a novel loss to improve speed, accuracy, and interpretability for autonomous driving safety.

Contribution

RARE eliminates heavy feature extraction modules, reducing latency, and introduces an Attention Score Ranking Loss to enhance accident anticipation accuracy and interpretability.

Findings

RARE achieves 4-8x speedup over existing methods.

It attains state-of-the-art accuracy on DAD and CCD benchmarks.

RARE operates at 73.3 FPS with 13.6ms latency.

Abstract

This paper addresses the problem of anticipating traffic accidents, which aims to forecast potential accidents before they happen. Real-time anticipation is crucial for safe autonomous driving, yet most methods rely on computationally heavy modules like optical flow and intermediate feature extractors, making real-world deployment challenging. In this paper, we thus introduce RARE (Real-time Accident anticipation with Reused Embeddings), a lightweight framework that capitalizes on intermediate features from a single pre-trained object detector. By eliminating additional feature-extraction pipelines, RARE significantly reduces latency. Furthermore, we introduce a novel Attention Score Ranking Loss, which prioritizes higher attention on accident-related objects over non-relevant ones. This loss enhances both accuracy and interpretability. RARE demonstrates a 4-8 times speedup over existing approaches on the DAD and CCD benchmarks, achieving a latency of 13.6ms per frame (73.3 FPS) on an RTX 6000. Moreover, despite its reduced complexity, it attains state-of-the-art Average Precision and reliably anticipates imminent collisions in real time. These results highlight RARE's potential for safety-critical applications where timely and explainable anticipation is essential.