Point Cloud Oversegmentation with Graph-Structured Deep Metric Learning

TL;DR

This paper introduces a supervised deep metric learning framework for oversegmenting 3D point clouds into superpoints, significantly improving over previous methods and enhancing semantic segmentation accuracy.

Contribution

The paper presents a novel graph-structured deep metric learning approach for point cloud oversegmentation, achieving state-of-the-art results with fewer superpoints and improving semantic segmentation.

Findings

Achieved new state-of-the-art oversegmentation performance on S3DIS and vKITTI datasets.

Required over five times fewer superpoints to reach similar performance on S3DIS.

Enhanced superpoint-based semantic segmentation results.

Abstract

We propose a new supervized learning framework for oversegmenting 3D point clouds into superpoints. We cast this problem as learning deep embeddings of the local geometry and radiometry of 3D points, such that the border of objects presents high contrasts. The embeddings are computed using a lightweight neural network operating on the points' local neighborhood. Finally, we formulate point cloud oversegmentation as a graph partition problem with respect to the learned embeddings. This new approach allows us to set a new state-of-the-art in point cloud oversegmentation by a significant margin, on a dense indoor dataset (S3DIS) and a sparse outdoor one (vKITTI). Our best solution requires over five times fewer superpoints to reach similar performance than previously published methods on S3DIS. Furthermore, we show that our framework can be used to improve superpoint-based semantic segmentation algorithms, setting a new state-of-the-art for this task as well.