TrafficRAG: A Multimodal RAG Framework for Traffic Accident Liability Determination
2026-06-01 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors developed TrafficRAG, a system that helps analyze who is responsible in traffic accidents by combining video information with legal knowledge. It first turns videos into clear text descriptions, then searches for relevant laws and past similar cases using smart search methods. Finally, it uses a language model to consider all this information and create detailed, legally accurate reports. Their tests show that TrafficRAG is more accurate and reliable than previous methods. This approach helps make accident liability decisions less subjective and more consistent.
traffic accident liabilitymultimodal retrievalvision-language modelBM25 retrievaldense embedding retrievallarge language modellegal domain knowledgereport generationfactual faithfulnessliability analysis
Authors
Xu Li, Zedong Fu, Xinyi Li, Xun Han
Abstract
Traffic accident liability analysis is a critical yet challenging task in intelligent transportation and legal assistance. Existing methods often suffer from low efficiency, subjective judgment, and inconsistent analysis results. Meanwhile, large language models are constrained by noisy video inputs and insufficient legal domain knowledge. To address these issues, this work presents TrafficRAG, a multimodal retrieval-augmented framework for automated traffic accident analysis and report generation. Specifically, the proposed framework first adopts a vision-language model to produce structured textual descriptions of accident scenarios, which serve as accurate retrieval queries. Based on these textual queries, a hybrid retrieval strategy integrating BM25 sparse retrieval and dense embedding retrieval is employed to fetch relevant traffic regulations and similar historical cases. Finally, the large language model incorporates retrieved legal knowledge and multimodal accident evidence for comprehensive reasoning, and generates standardized, legally grounded liability analysis reports. Extensive experiments show that TrafficRAG consistently outperforms baseline methods, achieving 77.32% Legal Norm Adaptation Accuracy, 81.71% Factual Faithfulness, and a Liability Ratio MAE of 5.48%. The results validate that integrating multimodal factual evidence with legal clauses via retrieval augmentation can effectively improve the reliability and accuracy of traffic accident liability determination.