MKConv: Multidimensional Feature Representation for Point Cloud Analysis

TL;DR

This paper introduces MKConv, a novel multidimensional convolution operator for point cloud analysis that enhances feature representation by transforming point features into multidimensional matrices, improving local structure understanding.

Contribution

The paper proposes MKConv, which transforms point features into multidimensional matrices and applies discrete convolutions, avoiding information loss from voxelization, and introduces a spatial attention module for better structure awareness.

Findings

MKConv outperforms existing methods in object classification.

MKConv improves accuracy in object part segmentation.

MKConv achieves superior results in scene semantic segmentation.

Abstract

Despite the remarkable success of deep learning, an optimal convolution operation on point clouds remains elusive owing to their irregular data structure. Existing methods mainly focus on designing an effective continuous kernel function that can handle an arbitrary point in continuous space. Various approaches exhibiting high performance have been proposed, but we observe that the standard pointwise feature is represented by 1D channels and can become more informative when its representation involves additional spatial feature dimensions. In this paper, we present Multidimensional Kernel Convolution (MKConv), a novel convolution operator that learns to transform the point feature representation from a vector to a multidimensional matrix. Unlike standard point convolution, MKConv proceeds via two steps. (i) It first activates the spatial dimensions of local feature representation by exploiting multidimensional kernel weights. These spatially expanded features can represent their embedded information through spatial correlation as well as channel correlation in feature space, carrying more detailed local structure information. (ii) Then, discrete convolutions are applied to the multidimensional features which can be regarded as a grid-structured matrix. In this way, we can utilize the discrete convolutions for point cloud data without voxelization that suffers from information loss. Furthermore, we propose a spatial attention module, Multidimensional Local Attention (MLA), to provide comprehensive structure awareness within the local point set by reweighting the spatial feature dimensions. We demonstrate that MKConv has excellent applicability to point cloud processing tasks including object classification, object part segmentation, and scene semantic segmentation with superior results.