AccidentGPT: A V2X Environmental Perception Multi-modal Large Model for Accident Analysis and Prevention

Traffic accidents are a significant factor leading to injuries and property losses, prompting extensive research in the field of traffic safety. However, previous studies, whether focused on static environment assessment, dynamic driving analysis, pre-accident prediction, or post-accident rule checks, have often been conducted independently. Our introduces V2X Environmental Perception Multi-modal Large Model AccidentGPT for accident analysis and prevention. AccidentGPT establishes a multi-modal information interaction framework based on multisensory perception. It adopts a holistic approach to address traffic safety issues, providing environmental perception for autonomous vehicles to avoid collisions and maintain control. In human-driven vehicles, it offers proactive safety warnings, blind spot alerts, and driving suggestions through human-machine dialogue. Additionally, it aids traffic police and management agencies in considering factors such as pedestrians, vehicles, roads, and the environment for intelligent real-time analysis of traffic safety. The system also conducts a thorough analysis of accident causes and post-accident liabilities, making it the first large-scale model to integrate comprehensive scene understanding into traffic safety research. Project page: https://accidentgpt.github.io

I. INTRODUCTION

Exploring traffic safety research with a primary focus on accident analysis and prevention is crucial for enhancing overall safety within the entire transportation system and reducing casualties and property losses [1], [2], [3]. Scholars have delved into this subject from various perspectives, categorizing studies based on traffic scenarios, events, and levels of traffic intelligence. Investigations cover static environment analysis [4], [5], [6], dynamic object state checks [7], [8], accident prediction, human driving safety analysis [9], [10], [11], [12], and comprehensive safety analysis for autonomous driving [13], [14]. This multifaceted approach reflects the complexity of traffic safety research, involving broader impacts, collision prevention, driving behavior, and vehicle dynamics. Proactive collision prediction and in-depth causal analysis aim to reduce the frequency and severity of accidents. As research in autonomous driving progresses, exploring human-machine hybrid driving [15], [16] and the coexistence of mixed vehicle types becomes crucial to achieving precise safety and control mechanisms in fully autonomous vehicles.