Structured Prompting and Multi-Agent Knowledge Distillation for Traffic Video Interpretation and Risk Inference
By: Yunxiang Yang, Ningning Xu, Jidong J. Yang
Potential Business Impact:
Helps cars understand roads and dangers better.
Comprehensive highway scene understanding and robust traffic risk inference are vital for advancing Intelligent Transportation Systems (ITS) and autonomous driving. Traditional approaches often struggle with scalability and generalization, particularly under the complex and dynamic conditions of real-world environments. To address these challenges, we introduce a novel structured prompting and knowledge distillation framework that enables automatic generation of high-quality traffic scene annotations and contextual risk assessments. Our framework orchestrates two large Vision-Language Models (VLMs): GPT-4o and o3-mini, using a structured Chain-of-Thought (CoT) strategy to produce rich, multi-perspective outputs. These outputs serve as knowledge-enriched pseudo-annotations for supervised fine-tuning of a much smaller student VLM. The resulting compact 3B-scale model, named VISTA (Vision for Intelligent Scene and Traffic Analysis), is capable of understanding low-resolution traffic videos and generating semantically faithful, risk-aware captions. Despite its significantly reduced parameter count, VISTA achieves strong performance across established captioning metrics (BLEU-4, METEOR, ROUGE-L, and CIDEr) when benchmarked against its teacher models. This demonstrates that effective knowledge distillation and structured multi-agent supervision can empower lightweight VLMs to capture complex reasoning capabilities. The compact architecture of VISTA facilitates efficient deployment on edge devices, enabling real-time risk monitoring without requiring extensive infrastructure upgrades.
Similar Papers
Multi-Agent Visual-Language Reasoning for Comprehensive Highway Scene Understanding
CV and Pattern Recognition
Helps cameras see road dangers and warn drivers.
VISTA: A Vision and Intent-Aware Social Attention Framework for Multi-Agent Trajectory Prediction
CV and Pattern Recognition
Helps self-driving cars avoid crashing into each other.
VISTA: Vision-Language Imitation of Situational Thinking and Attention for Human-Like Driver Focus in Dynamic Environments
CV and Pattern Recognition
Predicts where drivers look using words.