Detecting Learning Affect in E-Learning Platform Using Facial Emotion Expression

摘要

Recent trends in education have shifted from traditional classroom learning to an online learning setting; however, research has indicated a high drop out rate among e-learners. Boredom, lack of motivation are among the factors that led to this decline. This study develops a platform that provides feedback to learners in real-time while engaging in an online learning video. The platform detects, predicts and analyses the facial emotions of a learner using Convolutional Neural Network (CNN), and further maps the emotion to a learning affect. The feedback generated provides a reasonable understanding of the comprehension level of the learner.