Skip to main navigation Skip to search Skip to main content

ROTATION-EQUIVARIANT GRAPH CONVOLUTIONAL NETWORKS FOR SPHERICAL DATA VIA GLOBAL-LOCAL ATTENTION

  • Jiayi Xu
  • , Qin Yang
  • , Chenglin Li*
  • , Junni Zou*
  • , Hongkai Xiong
  • , Xinlong Pan*
  • , Haipeng Wang
  • *Corresponding author for this work
  • Shanghai Jiao Tong University
  • Naval Aviation University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Graph convolutional networks (GCNs) are widely adopted for spherical data processing, striking a balance between rotation equivariance and computation efficiency. However, current GCN-based approaches either apply radial filters with a fixed convolution kernel, or adopt attention schemes to learn a variable kernel based only on node features, which lack expressiveness and may suffer over-smoothing problem. To address this, in this paper, we propose a global-local attention-based spherical graph convolution (GlasGC) for spherical data representation. Specifically, the designed graph convolution includes a structure-enhanced local attention module to estimate the structure correlation efficiently based on topological characteristics of the given spherical graph. The structure correlation is then incorporated with the feature correlation to form the local attention that determines importance of a neighbor node to the center node. We also design a global attention module to obtain the importance scores of different nodes based on both node features and graph topology, which acts as a message-passing enhancer for informative nodes and can relieve the over-smoothing problem. Based on our proposed GlasGC, we design specific GlasGCNs for classification and semantic segmentation of spherical data. Empirical evaluations on these tasks demonstrate that our proposed GlasGCNs can generally achieve better classification and segmentation performance.

Original languageEnglish
Title of host publication2022 IEEE International Conference on Image Processing, ICIP 2022 - Proceedings
PublisherIEEE Computer Society
Pages2501-2505
Number of pages5
ISBN (Electronic)9781665496209
DOIs
StatePublished - 2022
Externally publishedYes
Event29th IEEE International Conference on Image Processing, ICIP 2022 - Bordeaux, France
Duration: 16 Oct 2022 → 19 Oct 2022

Publication series

NameProceedings - International Conference on Image Processing, ICIP
ISSN (Print)1522-4880

Conference

Conference29th IEEE International Conference on Image Processing, ICIP 2022
Country/TerritoryFrance
CityBordeaux
Period16/10/22 → 19/10/22

Keywords

  • attention mechanism
  • classification
  • graph convolutional networks
  • semantic segmentation
  • Spherical images

Fingerprint

Dive into the research topics of 'ROTATION-EQUIVARIANT GRAPH CONVOLUTIONAL NETWORKS FOR SPHERICAL DATA VIA GLOBAL-LOCAL ATTENTION'. Together they form a unique fingerprint.

Cite this