Abstract
Side-face biometrics, also known as profile face recognition, is a growing field within facial recognition technology focused on leveraging side-profile images for identity authentication. Unlike traditional facial recognition, which primarily relies on frontal views, profile recognition brings both unique challenges and opportunities. It assesses distinctive features like ear shape and facial contours, which complement conventional landmarks such as the eyes and nose. To enhance recognition, Principal Component Analysis (PCA) is used to reduce the dimensionality of side-profile images, preserving key features while filtering out noise. However, PCA alone struggles to capture complex spatial patterns. To address this, a Convolutional Neural Network (CNN) is integrated into the hybrid model, identifying subtle variations in facial features that PCA might miss, thus boosting recognition accuracy. This hybrid approach achieves an optimal balance between dimensionality reduction and feature extraction, yielding higher performance than either PCA or CNN alone. The combined system ensures efficient processing and reliable authentication, especially in contexts where side faces and ears are crucial biometric markers.
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright (c) 2024 African Journal of Biomedical Research