Learning Geometry-Disentangled Representation for Complementary Understanding of 3D Object Point Cloud
Mutian Xu 1 2, Junhao Zhang , Zhipeng Zhou 1, Mingye Xu 1 3, Xiaojuan Qi 2, Yu Qiao 1 M.Xu and J.Zhang contribute equally.Corresponding author.
Abstract
In 2D image processing, some attempts decompose images into high and low frequency components for describing edge and smooth parts respectively. Similarly, the contour and flat area of 3D objects, such as the boundary and seat area of a chair, describe different but also complementary geometries. However, such investigation is lost in previous deep networks that understand point clouds by directly treating all points or local patches equally. To solve this problem, we propose Geometry-Disentangled Attention Network (GDANet). GDANet introduces Geometry-Disentangle Module to dynamically disentangle point clouds into the contour and flat part of 3D objects, respectively denoted by sharp and gentle variation components. Then GDANet exploits Sharp-Gentle Complementary Attention Module that regards the features from sharp and gentle variation components as two holistic representations, and pays different attentions to them while fusing them respectively with original point cloud features. In this way, our method captures and refines the holistic and complementary 3D geometric semantics from two distinct disentangled components to supplement the local information. Extensive experiments on
中文速览
点云（point cloud）三维理解一直面临一个被忽视的问题：物体的轮廓边缘区域和平坦表面区域在几何上截然不同，却被以往网络一视同仁地处理，导致关键几何信息被冗余信息淹没。受二维图像高低频分解思路的启发，作者提出了几何解耦注意力网络（Geometry-Disentangled Attention Network，GDANet），核心是借助图信号处理中的高通滤波器，将点云动态分解为"锐变分量"（对应轮廓）和"缓变分量"（对应平坦区域）两种互补的整体表示，再通过设计的锐-缓互补注意力模块，根据几何相关性为两类分量分配不同的注意力权重，并分别与原始点特征融合，从而在局部信息之外补充全局互补的几何语义。在三维目标分类和分割基准上的大量实验表明，GDANet 以更少的参数量达到了当时最优性能，并对点云密度变化、旋转和噪声表现出更强的鲁棒性，证明了几何解耦这一思路在三维深度学习中的有效价值。
原文 arXiv:2012.10921;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2012.10921v3