Learning from the Unseen: Generative Data Augmentation for Geometric-Semantic Accident Anticipation

TL;DR

This paper introduces a dual-path framework combining synthetic data generation and graph neural networks to improve accident anticipation in autonomous driving, validated on a new comprehensive benchmark dataset.

Contribution

It presents a novel generative data augmentation method and a semantic-aware graph neural network for better accident prediction in autonomous driving.

Findings

Significant accuracy improvements on existing datasets.

Extended anticipation lead time in accident prediction.

Effective mitigation of data scarcity issues.

Abstract

Anticipating traffic accidents is a critical yet unresolved problem for autonomous driving, hindered by the inherent complexity of modeling interactions between road users and the limited availability of diverse, large-scale datasets. To address these issues, we propose a dual-path framework. On the one hand, we employ a video synthesis pipeline that, guided by structured prompts, derives feature distributions from existing corpora and produces high-fidelity synthetic driving scenes consistent with the statistical patterns of real data. On the other hand, we design a graph neural network enriched with semantic cues, enabling dynamic reasoning over both spatial and semantic relations among participants. To validate the effectiveness of our approach, we release a new benchmark dataset containing standardized, finely annotated video sequences that cover a broad spectrum of regions, weather, and traffic conditions. Evaluations across existing datasets and our new benchmark confirm notable gains in both accuracy and anticipation lead time, highlighting the capacity of the proposed framework to mitigate current data bottlenecks and enhance the reliability of autonomous driving systems.