Masked Faces Classification using Deep Convolutional Neural Network with VGG-16 Architecture

Abstract

Recent years have seen a significant increase in attention given to object detection techniques, particularly in the field of face mask detection, classification, and masked face recognition. Due to the contact-based nature of other biometric methods and the possible outbreak of a pandemic, facial biometrics is now the most secure option for authentication and access management. Experts have advised that adequate preparation must be made to tackle any occurrence of another respiratory-related pandemic in the future. One of the areas worthy of seeking and securing absolute technological breakthroughs is the aspect of face mask detection and classification. The current face mask detection and identification technologies were created using the principles of fair-skinned individuals. This study was carried out with a view to improving the existing systems to perform brilliantly in real- time on dark-skinned faces using a convolutional neural network with VGG-16 architecture. The system was evaluated, and the results show a better performance.