Gerenciamento de Frequência Escolar Usando Visão Computacional e Redes Neurais Convolucionais
Eder Almeida Chiquito, Guilherme Eduardo Rodrigues, Julio Becker, Caroline Mazetto Mendes · Anais do XI Computer on the Beach - COTB '20 · 2020
Monitoring students in the classroom plays a key role in the academicfollow-up. In addition to attendance control, it also allowsstudying the impact of absences on learning. In this context, thispaper presents the development of a school attendance managementsystem using computer vision techniques and convolutionalneural networks. A camera captures images from the classroom.The system then detects the faces of the people in the image, andrecognizes students using pre-trained convolutional neural networkmodels. The frequency data generated is stored in the systemdatabase and accessed through a web page. In addition, a mobileapp has been developed to notify student attendance. The initialresults show that the system identifies the registered students andnotifies the mobile device of responsible.