A Deep Learning-Based Approach for Students' Involvement Assessment in an E-Learning Platform

The multi-dimensional facilities of the e-learning-based platform enforces the students to use it, especially after the pandemic situations. Since many teenagers are using the facility, it is obvious to assess the involvement quotient of the students while accessing the e-learning materials. The proposed work automatic students' involvement assessment system (ASIAS) includes a two-stage vision-based technique to measure the involvement of the students. In the first stage, facial expression-based information is extracted from the live camera to compute the involvement quotient in terms of satisfaction, boredom, confusion, looking away, frustration, etc. In the second stage, screen distance detection is estimated to restrict health hazards. A ranking-based procedure is applied in this work on benchmark and collected datasets making the procedure effective. The performance of the ASIAS model is examined using the datasets FER-2013 and Student Engagement. The outcomes and comparison with cutting-edge methods demonstrate the usefulness of the ASIAS.