Prototype-Voxel Contrastive Learning for LiDAR Point Cloud Panoptic Segmentation
Minzhe Liu,Qiang Zhou,Hengshuang Zhao,Jianing Li,Yuan Du,Kurt Keutzer,Li Du,Shanghang Zhang,Minzhe Liu,Qiang Zhou,Hengshuang Zhao,Jianing Li,Yuan Du,Kurt Keutzer,Li Du,Shanghang Zhang
LiDAR point cloud panoptic segmentation, including both semantic and instance segmentation, plays a critical role in meticulous scene understanding for autonomous driving. Existing 3D voxelized approaches either utilize 3D sparse convolution that only focuses on local scene understanding, or add extra and time-consuming PointNet branch to capture global feature structures. To address these limitat...
