Decomposed gate attention graph convolutional networks for skeleton-based action recognition

Dengdi Sun; Yuanyin Zhou; Bin Luo; Zhuanlian Ding

doi:10.1117/12.3033194

5 July 2024 Decomposed gate attention graph convolutional networks for skeleton-based action recognition

Dengdi Sun, Yuanyin Zhou, Bin Luo, Zhuanlian Ding

Proceedings Volume 13184, Third International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024); 131844V (2024) https://doi.org/10.1117/12.3033194
Event: 3rd International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024), 2024, Kuala Lumpur, Malaysia

Abstract

Graph convolutional networks (GCNs) have garnered significant interest in the field of skeleton-based action recognition in the past few years. In this study, we introduce a novel approach called the Decomposed Gate Excitation Graph Convolutional Network (DGA-GCN) that can extract the features of joint and bones in action sequences from different dimensions for skeleton action recognition. The proposed DGA-GCN comprises an Initial Feature Extraction(IFE)module, a Multi-granularity Timing Analysis (MTA) module, and a Gate Spatial Temporal Attention (GSTA) module. The IFE module is capable of extracting more comprehensive features from the joints, bones, and temporal aspects of the frame. MTC can model global and short-term timing relationships. GSTE calculates the attention rate of bones and joints and then uses it Excite motion-sensitive channels. These three modeling can be integrated into the existing skeleton-based GCN for modeling. Extensive experiments show that HEG-GCN has higher recognition accuracy on NTURGB+D, NTU RGB+D 120 and Northwestern-UCLA data sets, while ensuring better recognition performance.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Dengdi Sun, Yuanyin Zhou, Bin Luo, and Zhuanlian Ding "Decomposed gate attention graph convolutional networks for skeleton-based action recognition", Proc. SPIE 13184, Third International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024), 131844V (5 July 2024); https://doi.org/10.1117/12.3033194

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
7 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Action recognition

Bone

Feature extraction

Convolution

Data modeling

Motion models

Modeling

Show All Keywords

Keywords/Phrases

Search In:

Publication Years