Artificial-Intelligence-Based School Assistant for Detecting the Behavior of University Students
Joseph Lopez-Carreño, Cristhian Calvo-Lavado, Manuel Azpilcueta-Vasquez, Eliseo Zarate-Perez · 2022
In this study, we programmed an artificial-intelligence-based school assistant and integrated it with the structure of surveillance cameras and loudspeakers in a university classroom. This was done to complement the supervision conducted by university teachers to identify student distractions, ensure proper use of protective equipment, and evaluate threatening behaviors, such as confrontations between students. The virtual assistant was developed using Python and generated audio warnings via a loudspeaker through a graphical interface built using the PyCharm environment. The results demonstrated the functionality desired from the virtual assistant and its ability to meet the requirements of a university classroom. Therefore, the effectiveness of the YOLO v5 network and PyCharm, used for training and execution, respectively, at constructing and deploying such a framework was demonstrated.