A Computer-Vision Based Engagement Evaluation System for More Effective Learning Design

Abstract

Student engagement during online instruction has been a challenge. Particularly in an asynchronous learning environment, assessing and measuring student engagement is a difficult task. This paper proposes a solution to fill in the gap between the need to evaluate student engagement during online instruction and the lack of an effective method to do so. The proposed web-based learning environment fully incorporates the latest development in facial expression recognition to capture the facial expression changes during the course of instruction. The facial expression data can then be visualized for further analysis to help instructors pinpoint the exact locations where the instruction needs improvements. In addition, the proposed system can also be used in corporate training settings, particularly in scenarios where the learners' emotional responses are desired.