GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in Point Cloud

CVPR 2019 Li YiWang ZhaoHe WangMinhyuk SungLeonidas Guibas

We introduce a novel 3D object proposal approach named Generative Shape Proposal Network (GSPN) for instance segmentation in point cloud data. Instead of treating object proposal as a direct bounding box regression problem, we take an analysis-by-synthesis strategy and generate proposals by reconstructing shapes from noisy observations in a scene... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT LEADERBOARD
3D Instance Segmentation ScanNet(v2) R-PointNet Mean AP @ 0.5 30.6 # 8
3D Object Detection ScanNetV2 GSPN [email protected] 30.6 # 4
[email protected] 17.7 # 4