DFormerv2: Geometry Self-Attention for RGBD Semantic Segmentation.

Bowen Yin, Jiao-Long Cao, Ming-Ming Cheng, Qibin Hou

Browse the full CVPR paper archive.