Automated Attendance System Using Face Recognition with One-Shot Learning and Siamese Neural Networks

Abstract

A face is the identity of a person. Not only human but also machines can now identify a person. Face recognition system is a Biometric Artificial Intelligence application. Now it is widely used everywhere in the world. It can be also used in our classroom. It is compulsory to take attendance in an educational institute like school, college, university. Corny approach for attendance is to call students by their names and record the attendance. We are going to use an automated attendance system in our classroom to avoid it. In this approach, we can use the face detection and recognition method. Face detection differentiates the faces from each other and recognition method recognizes the person for daily attendance. For face detection, we can use Haar cascade classifier and for face recognition, we can use Google’s FaceNet model. It takes less time than a corny approach. It also helps the teacher to maintain a large classroom. The system also helps to avoid a large number of students from skipping the daily classes. In this system, there is no way of cheating like fake attendance. With the help of this system, we can take attendance at any time. Automated attendance system is now also an individual research subject. We can now focus to make the method more efficient in real time scenario.