SGPN: Similarity Group Proposal Network for 3D Point Cloud Instance Segmentation

TLDR

The paper proposes SGPN, a deep learning framework for 3D point‑cloud instance segmentation. SGPN predicts point‑grouping proposals and semantic labels with a single network, using a similarity matrix of point embeddings to generate accurate grouping proposals. Experiments show SGPN improves 3D instance segmentation, object detection, and semantic segmentation, and can incorporate 2D CNN features to further boost performance.

Abstract

We introduce Similarity Group Proposal Network (SGPN), a simple and intuitive deep learning framework for 3D object instance segmentation on point clouds. SGPN uses a single network to predict point grouping proposals and a corresponding semantic class for each proposal, from which we can directly extract instance segmentation results. Important to the effectiveness of SGPN is its novel representation of 3D instance segmentation results in the form of a similarity matrix that indicates the similarity between each pair of points in embedded feature space, thus producing an accurate grouping proposal for each point. Experimental results on various 3D scenes show the effectiveness of our method on 3D instance segmentation, and we also evaluate the capability of SGPN to improve 3D object detection and semantic segmentation results. We also demonstrate its flexibility by seamlessly incorporating 2D CNN features into the framework to boost performance.