Student Engagement from Facial Emotions using Deep Learning

Abstract

A fundamental idea in modern education, where it is considered as a goal in itself, is student engagement. In this study, we investigate methods for automatically identifying student engagement from their facial expressions. According to our opinion, the next generation of online learning environments ought to be able to monitor learners' involvement and offer tailored intervention. In this research paper, I used an existing model that is VGG16 to detecting online learner’s engagement through their facial expressions. Students facial expression will be recognized to categorize them into three-level engagement such as engaged, not engaged and neutral. The VGG16 model shows better accuracy that is 97.14%. This experiment conducted of FER2013 Dataset.