Real-Time Student Surveillance System Using Machine Learning and Computer Vision

Abstract

In a classroom full of students, it is practically not possible for a single teacher to give personal attention to every student. Continuous monitoring and identification of students who show signs of lethargy, sadness or anger in classrooms can help management counsel them and advice some preventive measures – which if kept unchecked can lead the students to take adverse steps. Potential students who need help in some form can be identified. There are some unproductive activities that take place in a classroom which can be easily automated and the time saved can be devoted to productive activities. Manual attendance calling is one of them which can be automated using facial recognition algorithms. In this paper, a robust surveillance system using Machine Learning and Computer Vision algorithms is presented that can take on the above challenges. For facial recognition, Local Binary Pattern Histograms have been used and for emotion recognition, Deep Learning model has been used.