Local Spectral Graph Convolution for Point Set Feature Learning

TL;DR

This paper introduces a local spectral graph convolution method with a novel pooling strategy for point cloud feature learning, improving classification and segmentation performance by capturing local geometric relationships.

Contribution

It proposes a spectral graph convolution approach with recursive clustering pooling to better encode local point relationships in point cloud analysis.

Findings

Consistent improvement in point set classification accuracy.

Enhanced segmentation performance on multiple datasets.

Richer feature descriptors from spectral clustering pooling.

Abstract

Feature learning on point clouds has shown great promise, with the introduction of effective and generalizable deep learning frameworks such as pointnet++. Thus far, however, point features have been abstracted in an independent and isolated manner, ignoring the relative layout of neighboring points as well as their features. In the present article, we propose to overcome this limitation by using spectral graph convolution on a local graph, combined with a novel graph pooling strategy. In our approach, graph convolution is carried out on a nearest neighbor graph constructed from a point's neighborhood, such that features are jointly learned. We replace the standard max pooling step with a recursive clustering and pooling strategy, devised to aggregate information from within clusters of nodes that are close to one another in their spectral coordinates, leading to richer overall feature descriptors. Through extensive experiments on diverse datasets, we show a consistent demonstrable advantage for the tasks of both point set classification and segmentation.