1 paper
Shaobo Xia, Jun Yue, Kacper Kania +4
We propose a weakly supervised semantic segmentation method for point clouds that predicts "per-point" labels from just "whole-scene" annotations. The key challenge here is the dis…