A Graph-CNN for 3D Point Cloud Classification

TL;DR

This paper introduces PointGCN, a graph convolutional neural network designed for classifying 3D point cloud data, effectively capturing local structures and demonstrating competitive performance on benchmark datasets.

Contribution

The paper presents a novel Graph-CNN architecture with localized convolutions and graph pooling for 3D point cloud classification, improving stability and performance.

Findings

Achieves competitive accuracy on ModelNet benchmark.

More stable than existing methods.

Effectively captures local point cloud structures.

Abstract

Graph convolutional neural networks (Graph-CNNs) extend traditional CNNs to handle data that is supported on a graph. Major challenges when working with data on graphs are that the support set (the vertices of the graph) do not typically have a natural ordering, and in general, the topology of the graph is not regular (i.e., vertices do not all have the same number of neighbors). Thus, Graph-CNNs have huge potential to deal with 3D point cloud data which has been obtained from sampling a manifold. In this paper, we develop a Graph-CNN for classifying 3D point cloud data, called PointGCN. The architecture combines localized graph convolutions with two types of graph downsampling operations (also known as pooling). By the effective exploration of the point cloud local structure using the Graph-CNN, the proposed architecture achieves competitive performance on the 3D object classification benchmark ModelNet, and our architecture is more stable than competing schemes.