Attendance Tracking with Face Recognition Through Hidden Markov Models

Facial recognition is one of the most secure ways to identify a person. Manual attendance in organizations, be it in classrooms or libraries or even attendance for teachers is truly a hassle. Due to the inception of Hidden Markov Model (HMM), they have worked well with image data and it has a plethora of facial recognition applications. In this paper, yet another application of face recognition with HMM is explored, where it is integrated with Singular Value Decomposition (SVD) and tracks the attendance of the students present in a database. HMMs deal with data in the form of states and sequences. Face recognition looked through the lens of HMMs can be framed in the following manner: a face is split into regions vertically (forehead, chin, etc.) and a particular sequence is always preserved. A rectangular window of fixed size is passed over every test image, and for every vector obtained, the probability of data is calculated. For training, probability computation is done with the help of the Baum-Welch algorithm. This whole model is connected to a simple program to keep track of the students leaving and entering the classroom, marking their presence only and updating the same information in the college’s database.

2022 International Conference on Electronics and Renewable Systems (ICEARS)