Bridging the Modality Gap in Roadside LiDAR: A Training-Free Vision-Language Model Framework for Vehicle Classification
弥合道路激光雷达的模态差距:一种无需训练的视觉-语言模型框架用于车辆分类
机构 * Department of Civil Engineering, City College of New York(城市学院土木工程系) ; Department of Computer Science, City College of New York(城市学院计算机科学系)
专题命中 VLM训练与架构 :vision-language model(title,abstract);VLM(abstract);分类 cs.CV、cs.LG
AI总结 本文提出一种无需训练的视觉-语言模型框架,用于解决道路激光雷达中稀疏点云与密集图像之间的模态差距问题,实现细粒度车辆分类。
Comments 12 pages, 10 figures, 4 tables