GraphBEV++: Multi-Modal Feature Alignment for Autonomous Driving
GraphBEV++: 自动驾驶中的多模态特征对齐
机构 * Beijing Key Laboratory of Traffic Data Mining and Embodied Intelligence, School of Computer Science and Technology, Beijing Jiaotong University(北京交通大学计算机科学与技术学院,交通数据挖掘与具身智能北京市重点实验室) ; School of Artificial Intelligence (School of Software), Yanshan University(燕山大学人工智能学院(软件学院)) ; University of Macau(澳门大学) ; Nanyang Technological University(南洋理工大学) ; The University of Queensland(昆士兰大学)
专题命中 BEV与占用 :BEV(summary_cn,abstract);autonomous driving(title,abstract);LiDAR(abstract);occupancy(abstract)
AI总结 针对自动驾驶中BEV感知的特征未对齐问题,提出GraphBEV++框架,通过局部对齐(LocalAlign-v2)和全局对齐(GlobalAlign-v2)模块,利用图匹配、可变形偏移和扩散去噪方法,在多种基准上实现最优性能。
Comments 30 pages, 7 figures