OS26-2 Optimizing Face Embedding Sizes and Accuracy in Facial Recognition Systems

Title:	OS26-2 Optimizing Face Embedding Sizes and Accuracy in Facial Recognition Systems
Publication:	ICAROB2025
Volume:	30
Pages:	723-728
ISSN:	2188-7829
DOI:	10.5954/ICAROB.2025.OS26-2
Author(s):	Wui Chung Alton Chau, Hao Feng Chan, Dexter Sing Fong Leong, Shakir Hussain Naushad Mohamed, Andi Prademon Yunus, Takao Ito, Zheng Cai, Xinjie Deng, Yit Hong Choo
Publication Date:	February 13, 2025
Keywords:	Face Recognition, Optimization, Hyperparameter Tuning, Embedding Sizes, Deep Learning
Abstract:	Face recognition technology is integral to security, access control and identity verification in finance, healthcare and transportation. It protects personal data, secures online transactions, controls access to areas, and helps prevent identity theft. This paper proposes a novel hybrid optimization algorithm, Moss Particle Swarm Optimization (MPSO) to perform hyperparameter tuning, aiming to identify the neural network, alongside Triple Loss metrics for efficient recognition. The proposed method is tested with the Labeled Faces in the Wild (LFW) dataset, demonstrating its effectiveness in improving facial recognition systems.
PDF File:	https://alife-robotics.co.jp/members2025/icarob/data/html/data/OS/OS26/OS26-2.pdf
Copyright:	© The authors. This article is distributed under the terms of the Creative Commons Attribution License 4.0, which permits non-commercial use, distribution and reproduction in any medium, provided the original work is properly cited. See for details: https://creativecommons.org/licenses/by-nc/4.0/

HOME