ShellNet: Efficient Point Cloud Convolutional Neural Networks using Concentric Shells Statistics

TL;DR

ShellNet introduces ShellConv, a permutation-invariant convolution using concentric shells, enabling efficient, fast training of deep neural networks directly on point cloud data for various 3D scene understanding tasks.

Contribution

The paper proposes ShellConv, a novel convolution operator based on concentric shells, and ShellNet, an efficient neural network architecture for point cloud analysis.

Findings

Achieves state-of-the-art results in object classification and segmentation.

Maintains fast training speed and simple architecture.

Effectively handles large receptive fields on point clouds.

Abstract

Deep learning with 3D data has progressed significantly since the introduction of convolutional neural networks that can handle point order ambiguity in point cloud data. While being able to achieve good accuracies in various scene understanding tasks, previous methods often have low training speed and complex network architecture. In this paper, we address these problems by proposing an efficient end-to-end permutation invariant convolution for point cloud deep learning. Our simple yet effective convolution operator named ShellConv uses statistics from concentric spherical shells to define representative features and resolve the point order ambiguity, allowing traditional convolution to perform on such features. Based on ShellConv we further build an efficient neural network named ShellNet to directly consume the point clouds with larger receptive fields while maintaining less layers. We demonstrate the efficacy of ShellNet by producing state-of-the-art results on object classification, object part segmentation, and semantic scene segmentation while keeping the network very fast to train.