Simple and Effective Spatial-Attended Graph CNN with Learnable Aggregation for Classification in Point Clouds

Min Kuan Chang*, Yu Lin Wu*, Feng Tsun Chien, Guu Chang Yang*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

The main task of point cloud classification is to extract unique features from each point cloud and to accurately distinguish individual point clouds. How to extract meaningful features in point clouds to improve recognition accuracy has always been a challenge. In this work, the double-KNN dynamic graphs with different neighbor sets are proposed to process the input point cloud separately and to extract respective features. These two separate features are individually gone through the convolutional block attention module (CBAM) [1], followed by learnable aggression mechanisms to construct the final salient features for classification. The simulation shows that the proposed architecture has better performance in overall accuracy and average accuracy when testing in the ModelNet40 dataset.

Original languageEnglish
Title of host publicationGCCE 2023 - 2023 IEEE 12th Global Conference on Consumer Electronics
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages105-106
Number of pages2
ISBN (Electronic)9798350340181
DOIs
StatePublished - 2023
Event12th IEEE Global Conference on Consumer Electronics, GCCE 2023 - Nara, Japan
Duration: 10 Oct 202313 Oct 2023

Publication series

NameGCCE 2023 - 2023 IEEE 12th Global Conference on Consumer Electronics

Conference

Conference12th IEEE Global Conference on Consumer Electronics, GCCE 2023
Country/TerritoryJapan
CityNara
Period10/10/2313/10/23

Keywords

  • Convolution Network
  • Deep learning
  • Point cloud classification
  • self-Attention module

Fingerprint

Dive into the research topics of 'Simple and Effective Spatial-Attended Graph CNN with Learnable Aggregation for Classification in Point Clouds'. Together they form a unique fingerprint.

Cite this