E-Monitoring of Student Engagement Level using Facial Gestures

Abstract

Student engagement is a key element to ensure effective learning process. In this work, we presented an automatic system for monitoring engagement level from students’ facial gestures. In this way, the tutor can analyse the engagement level of students and improve the teaching method and strategies to enhance learning process. There has been extensive research on automated classification of engagement level, but most of these methods rely mainly on expensive eye trackers or physiological sensors in controlled settings. The proposed system monitors and classifies engagement level of student based on YOLO algorithm by determining facial gestures, where students move freely and respond naturally to lectures and surroundings. The proposed model gives a mean average precision (mAP) of 0.65 on a complex dataset where students were allowed to move freely during lecture.

Keywords

• face detection
• feature extraction
• YOLO
• mAP
• IoU