Mining Point Cloud Local Structures by Kernel Correlation and Graph Pooling

TL;DR

This paper introduces novel operations for deep learning on 3D point clouds that enhance local structure exploitation, leading to improved semantic understanding and performance on major datasets.

Contribution

It proposes two new operations—kernel correlation for local geometric structures and recursive feature aggregation for high-dimensional features—to improve PointNet's ability to learn from local neighborhoods.

Findings

Achieves better performance on major point cloud datasets.

Efficiently captures local geometric and feature structures.

Outperforms existing methods in semantic learning tasks.

Abstract

Unlike on images, semantic learning on 3D point clouds using a deep network is challenging due to the naturally unordered data structure. Among existing works, PointNet has achieved promising results by directly learning on point sets. However, it does not take full advantage of a point's local neighborhood that contains fine-grained structural information which turns out to be helpful towards better semantic learning. In this regard, we present two new operations to improve PointNet with a more efficient exploitation of local structures. The first one focuses on local 3D geometric structures. In analogy to a convolution kernel for images, we define a point-set kernel as a set of learnable 3D points that jointly respond to a set of neighboring data points according to their geometric affinities measured by kernel correlation, adapted from a similar technique for point cloud registration. The second one exploits local high-dimensional feature structures by recursive feature aggregation on a nearest-neighbor-graph computed from 3D positions. Experiments show that our network can efficiently capture local information and robustly achieve better performances on major datasets. Our code is available at http://www.merl.com/research/license#KCNet