Pengenalan Wajah Pada Aplikasi Presensi Perkuliahan Menggunakan FaceNet Berbasis Android

Abstract

The attendance system for lecture activities in various educational institutions, including the Department of Electrical Engineering, University of Mataram, still uses paper as a recording medium. The use of paper has a risk of damage and loss, and allows fraud by students. The author developed an Android-based lecture attendance application that uses facial recognition technology with FaceNet, to overcome this problem. When they want to fill in the attendance list, students must take a selfie with the appropriate face position when registering into the attendance system. Through testing facial images, research shows that this system has an accuracy rate of 96.53% in various facial positions, such as upright facing the camera, facing right, and facing left, so that it can be considered very good for recognizing and validating student faces. Thus, the use of this application can help reduce the risk of damage and loss of paper and minimize the possibility of fraud by students.