Radar-Camera BEV Multi-Task Learning with Cross-Task Attention Bridge for Joint 3D Detection and Segmentation 事件

Radar-Camera BEV Multi-Task Learning with Cross-Task Attention Bridge for Joint 3D Detection and Segmentation arXiv:2604.12918v2 Announce Type: replace Abstract: Bird's-eye-view (BEV) representations are the dominant paradigm for 3D perception in autonomous driving, providing a unified spatial canvas where detection and segmentation features are geometrically registered to the same physical coordinate system. However, existing radar-camera fusion methods treat these tasks in isolation, missing