Seamless Security and Convenience: AI-Powered Face Recognition for Smart Office Automation

Abstract

The accuracy of face recognition systems in smart office automation applications is limited by various constraints. These limitations highlight the need of researching mask face recognition. This study presents a groundbreaking deep learning approach called Faster R-CNN, which combines with IoT technology to address security concerns in office environments. The existing employees' photos were collected and stored in a database. These photographs are then subjected to pre-processing in order to train the neural network. The Faster R-CNN algorithm utilises the VGG-16 model as the basis of its architecture to extract structures from pre-processed images. The advancements in Internet of Things (IoT) and deep learning have enabled the use of deep neural networks to tackle the challenges associated with facial recognition. According to the feature classification, when a person who belongs to an organisation styles the gate, it immediately opens. If the person is unfamiliar, the door will remain locked. The cloud saved photographs of both authorised and unauthorised individuals, which were subsequently sent to the office manager for monitoring. The suggested Faster R-CNN model achieves an correctness of 99.3%, which is superior to the correctness of the previous scheme. The proposed Quicker R-CNN demonstrates superior exactness improvements compared to Deep CNN, SVM, LBPH, and OMTCNN. Specifically, it achieves accuracy ranges that are 2.06%, 5.63%, 9.36%, and 3.54% better than the aforementioned methods, respectively.