Volume 18, No. 6, 2021
Face Recognition System For People With And Without Mask Using Deep Learning Model
V.Prema , M.Sivajothi , Grasha Jacob
Abstract
Face recognition system is one of the most important biometric authentication systems frequently employed in access control and surveillance. With the superior power of three dimensional (3D) imaging sensors,3D face recognition systems have the potential to attain good recognition accuracy compared to 2D face recognition systems. However, most of the existing 3D face recognition approaches suffer from computational overhead. Recently, the COVID-19 pandemic is spreading throughout the globe which changes the lives of people and creates more problems on people’s health. Wearing a face mask in public places is an effective way to prevent viruses from spreading. This has initiated issues in current face recognition system, motivating the development of an efficient approach for recognizing both masked and non-masked faces. This paper proposes a robust 3D face recognition system based on deep learning model which is capable of recognizing both masked and non-masked faces. Convolutional neural network is designed to classify the facial images into masked and nonmasked faces. Non-masked faces are recognized based on the uncovered facial features while masked faces are recognized based on the features from the preprocessed images. The recognition process is carried out using twin neural network. Performance of the proposed system is validated using four data bases such as FRGC V2.0, Texas 3D, Bosporus and Real World Masked Face Recognition Dataset (RWFRD). Experimental results showed that the proposed system provide excellent performance compared with the existing approaches.
Pages: 4451-4465
Keywords: Biometric authentication system aims to identify a person based on his/her behavioral and psychological trails.