Automated Attendance System in the Classroom Using Artificial Intelligence and Internet of Things Technology

Abstract

Computer vision is recently developing and applying in the utility apps serving people, facial recognition is one of its applications. Although the accuracy of the facial recognition is less than when compared to fingerprint recognition, iris recognition and Radio Frequency Identification (RFID) card recognition. But it is still widely used because the recognition process does not contact the device. With the advantage of the facial recognition method, we propose an automated attendance solution which uses embedded device integrated Artificial Intelligence technology (AI) and Internet of Things technology (IoT) in the smart classrooms. The highlight of the system is the ability to take attendance automatically and continuously throughout the learning period. When the students enter the class, the management department and the parents can know the student’s participation status by viewing the report in the real-time system. The system consists of the main components: embedded device component with attached camera sensor, which is used for the process recognition and interacts with the Cloud server via IoT infrastructure; the Cloud server stores and provides data analysis devices for the administrator and parents. At the beginning of the roll call, the embedded device will receive an instruction to replace the old data with the new recorded data, which contains the characteristics and identifier code of the objects to be attended. The new data goes from the Cloud server in the respective classroom to the embedded device and then it compares with the data collected in the classroom. When the results are available, the embedded device interacts with the Cloud server to update the status of the students. The experimental results of the proposal system achieve accuracy per frame is 89%. The recognition speed of 82ms per face with a distance in the 4 - 15 meter range. The system which is an embedded system-based application solution has low operating costs and rapid deployment.