Paper
10 October 2023 Hypergraph neural network for gait recognition based on event camera
Lihua Fu, Shaoxing Yan
Author Affiliations +
Proceedings Volume 12799, Third International Conference on Advanced Algorithms and Signal Image Processing (AASIP 2023); 127994R (2023) https://doi.org/10.1117/12.3006325
Event: 3rd International Conference on Advanced Algorithms and Signal Image Processing (AASIP 2023), 2023, Kuala Lumpur, Malaysia
Abstract
Gait recognition is a biometric technology that has higher accuracy compared to fingerprint recognition and facial recognition. Event Camera is a new type of camera with advantages such as high temporal resolution, high dynamic range. This paper proposes a hypergraph neural network for gait recognition based on event camera. By using event flow downsampling module for downsampling, the amount of data is reduced without affecting the discriminability. The function of event feature extraction module is to convert events into graph nodes. Spatiotemporal hypergraph convolution module can construct a hypergraph, extract spatiotemporal feature and obtain pedestrian gait feature. This paper constructs a gait recognition dataset CASIA-B-EV based on event camera. Finally, the effectiveness of the proposed model is verified by experiments.
(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.
Lihua Fu and Shaoxing Yan "Hypergraph neural network for gait recognition based on event camera", Proc. SPIE 12799, Third International Conference on Advanced Algorithms and Signal Image Processing (AASIP 2023), 127994R (10 October 2023); https://doi.org/10.1117/12.3006325
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Gait analysis

Cameras

Neural networks

Feature extraction

Matrices

Convolution

Video

RELATED CONTENT

Using color for face verification
Proceedings of SPIE (August 05 2009)

Back to Top