Deep Learning based Smart Attendance System using Facial Recognition

Abstract

The implementation of an attendance management system is a crucial practice that any institution or organization adopts to record the attendance of their employees or members. The manual process of recording attendance using a paper-based or file-based system is fraught with deficiencies, including the potential for data loss, tampering, or catastrophic events. The current standard specifies the implementation of a smart attendance system that utilizes RFID tags, fingerprints, iris scans, speech recognition, and other technologies. Currently, advancements in technology are promoting the practical application of facial recognition methods for a more effective attendance management system. The facial recognition-based attendance system offers convenience and additional benefits by eliminating the need for human intervention, helping to mitigate the transmission of infections. This study proposes the use of a facial recognition technology in a real-time attendance management system to identify persons. Deep learning-based face recognition systems have demonstrated superior accuracy and performance in comparison to other face identification approaches, such as eigenfaces. Contemporary face recognition systems have distinct stages, including face detection, face alignment, feature extraction, face representation, and face recognition. This paper presents a deep learning methodology for creating a class attendance system that is based on face recognition. The Multitask Convolutional Neural Network (MTCNN) is employed for the purpose of detecting and aligning faces, while a lightweight hybrid high performance model is utilized. Ultimately, the project successfully achieved its goals and objectives, which involved creating a facial recognition system specifically built for tracking attendance and completing a comprehensive evaluation of existing methodologies and outcomes in facial recognition algorithms. In addition, the web application built by Streamlit offers key functionalities such as database creation, facial recognition, and attendance monitoring. Its purpose is to simplify attendance monitoring for end-users. An analysis was conducted to evaluate the system's performance and usability, with the aim of identifying areas for future improvements.