Due to the COVID-19 pandemic, many schools worldwide are using tele-education for class delivery. However, this causes a problem related to students’ active class participation. An AI-based educational application was developed regarding the students’ actions recognition in tele-education. In this integrated system the teacher can visualise student actions in a 3D virtual classroom, where every student is represented by an animated avatar that demonstrates their activity, without affecting the privacy of students. The integrated system incorporates a machine learning model trained to classify seven student actions (absent, attending, hand raising, looking elsewhere, telephone call, using phone and writing) and during an experimental investigation it managed to classify correctly 94.32% image frames of previously unseen students. During the evaluation process, the stakeholders (students, parents and educators) stated that the use of the application is feasible and improves the efficiency of tele-education. A video demonstrating the operation of the system can be found here:
