Student attendance system using facial recognition based on deep learning / Syahila Aina Haris and Zulfikri Paidi

The learning process depends on student attendance. There are many ways to track student attendance, and one of them is using their signatures. The procedure has a number of drawbacks, such as taking a long time to complete attendance, attendance papers are lost, the administration must manually enter each student’s attendance information into the computer and there is also a possibility of attendance fraud among students. In order to overcome this problem, this paper suggested a web-based face recognition student attendance system as a solution to this problem. In this suggested system, K-NN is used to categorize student faces, deep metric learning is used to build facial embedding, and Convolutional Neural Network (CNN) is used to detect faces in photos. The development of this system is also assisted by several other software. As a result, the computer can identify faces. This algorithm can identify the faces of students who appear in class, and their attendance will be recorded automatically into the system. As a consequence, tracking attendance information is made easier for student administration.