M2H: Multi-Task Learning with Efficient Window-Based Cross-Task Attention for Monocular Spatial Perception
M2H:基于高效窗口交叉任务注意力的多任务学习用于单目空间感知
机构 * Department of Earth Observation Science(地球观测科学系)
AI总结 本文提出M2H框架,通过高效的窗口交叉任务注意力模块,实现单目图像上的语义分割、深度估计、边缘检测和表面法线估计,同时在计算效率上优于现有方法。
Comments Accepted to the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025). 8 pages, 7 figures