Student Monitoring System Combining Facial Recognition and Identification Methods

Various exam proctoring solutions and tools are used by educational institutes worldwide, offering different methods to reduce the chances of cheating. Taking attendance in class and in exam room is one of the mandatory activities that any teacher must perform. This article presents a solution to do this job automatically, through facial recognition. Attendance is conducted through cameras in the classroom and a student information management website, which utilizes the MySQL data management system to store information related to students, coursework, and exams. The system uses the MTCNN convolutional network to detect facial areas and the FaceNet deep learning technique to recognize student faces in the classroom. The system will identify students’ faces in the class list and save the attendance status of the class in a spreadsheet file in CSV/xlsx format. Teachers will upload this CSV/xlsx file to the system website to automatically calculate attendance scores or confirm exam attendance for students. The system was tested for 5 classes with more than 190 students, with high reliability.