A Conceptual Architecture for Monitoring Students in Zoom during Online Educational Sessions

Abstract

Recently the education systems around the world have adhered to a distance learning/teaching method. However, this approach don’t provide any guarantee that participated students are presented during the session. The given solution is suggested as an add-on feature in Zoom. This feature enables the teacher to detect the students’ availability. The main objective of the architecture is to check the availability of students in the meeting time to time. In this architecture automatic image capturing and processing techniques are used. According to the proposed method images of students are captured automatically and periodically to a previously defined time i nterval. This is done by manipulating the camera in Android and making the preview of the camera invisible. The captured images are sent to the Zoom cloud to identify the identity of the student. Previously created image vectors (during training period) of students will be there to do the comparisons. Technologies like face embedding, Artificial Neural Network (ANN) are used in identification process of image p rocessing. If they are proven to be the legit participants, the students are accepted to the meeting from the waiting room. Periodically captured images are sent to the face detection by image processing using technologies as Viola Jones algorithm, Haar-like Features, AdaBoost algorithm, Cascading Classifiers, OpenCV. If the student is not available in the seat, a message is sent to the teacher. Even though this conceptual architecture has few limitations, this will be a great help in detecting the students’ identities and availability during learning/teaching environment.