Deep Learning Powered Class Attendance System Based on Edge Computing

Abstract

While digitalization of the education sector has been going for years, most monotonous administrative tasks are still being performed manually. One of those tasks is class attendance tracking. On all levels of the Slovak educational system, teachers still spend several minutes trying to figure out who is absent from the class. This, mostly manual work, can be easily automatized with the help of the modern technology. We present an edge-based deep-learning powered class attendance system based on the Dlib machine learning framework and NVIDIA Jetson Nano platform to tackle this problem.