面向交通视频解读与风险推理的结构化提示与多智能体知识蒸馏
Structured Prompting and Multi-Agent Knowledge Distillation for Traffic Video Interpretation and Risk Inference
AI总结：
本文提出结构化提示与多智能体知识蒸馏框架，利用GPT-4o和o3-mini生成伪标注，训练3B级轻量模型VISTA，实现交通视频风险感知描述与边缘端实时监测。
摘要：
全面的高速公路场景理解与稳健的交通风险推理对于推进智能交通系统（ITS）和自动驾驶至关重要。传统方法在真实世界复杂动态条件下往往难以实现良好的可扩展性和泛化能力。为应对这些挑战，我们提出了一种新颖的结构化提示与知识蒸馏框架，能够自动生成高质量的交通场景标注和上下文风险评估。该框架协调两个大型视觉-语言模型（VLMs）：GPT-4o 和 o3-mini，采用结构化思维链（Chain-of-Thought, CoT）策略生成丰富的多视角输出。这些输出作为知识增强的伪标注，用于对一个规模小得多的学生视觉-语言模型进行监督微调。由此得到的紧凑型 3B 规模模型被命名为 VISTA（Vision for Intelligent Scene and Traffic Analysis，智能场景与交通分析视觉模型），能够理解低分辨率交通视频并生成语义忠实、具有风险感知能力的描述。尽管参数量显著减少，VISTA 在与教师模型对比时，在既有描述生成指标（BLEU-4、METEOR、ROUGE-L 和 CIDEr）上仍取得了强劲性能。这表明有效的知识蒸馏和结构化多智能体监督能够赋能轻量级视觉-语言模型，使其获得复杂推理能力。VISTA 的紧凑架构便于在边缘设备上高效部署，无需大规模基础设施升级即可实现实时风险监测。
英文摘要：
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.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
