GaitAE: a cognitive model-based autoencoding technique for gait recognition.

Li, Rui; Li, Huakang; Qiu, Yidan; Ren, Jinchang; Ng, Wing W.Y.; Zhao, Huimin

Abstract

Gait recognition is a long-distance biometric technique with significant potential for applications in crime prevention, forensic identification, and criminal investigations. Existing gait recognition methods typically introduce specific feature refinement modules on designated models, leading to increased parameter volume and computational complexity while lacking flexibility. In response to this challenge, we propose a novel framework called GaitAE. GaitAE efficiently learns gait representations from large datasets and reconstructs gait sequences through an autoencoder mechanism, thereby enhancing recognition accuracy and robustness. In addition, we introduce a horizontal occlusion restriction (HOR) strategy, which introduces horizontal blocks to the original input sequences at random positions during training to minimize the impact of confounding factors on recognition performance. The experimental results demonstrate that our method achieves high accuracy and is effective when applied to existing gait recognition techniques.

Citation

LI, R., LI, H., QIU, Y., REN, J., NG, W.W.Y. and ZHAO, H. 2024. GaitAE: a cognitive model-based autoencoding technique for gait recognition. Mathematics [online], 12(17), article number 2780. Available from: https://doi.org/10.3390/math12172780

• Journal Article Type: Article
• Acceptance Date: Sep 6, 2024
• Online Publication Date: Sep 8, 2024
• Publication Date: Sep 1, 2024
• Deposit Date: Sep 26, 2024
• Publicly Available Date: Sep 26, 2024
• Journal: Mathematics
• Electronic ISSN: 2227-7390
• Publisher: MDPI
• Peer Reviewed: Peer Reviewed
• Volume: 12
• Issue: 17
• Article Number: 2780
• DOI: https://doi.org/10.3390/math12172780
• Keywords: Gait recognition; Biologic recognition; Autoencoder; Deep learning; Computer vision; Covariate reduction
• Public URL: https://rgu-repository.worktribe.com/output/2487235