Analysis of Facial Expressions for the Estimation of Concentration on Online Lectures

Abstract

The present study aimed to develop a method to estimate the state of attention from facial expressions while students are in online lectures. We con-ducted an experiment to measure the level of attention based on reaction time to detect an auditory target, which was the disappearance of noise sound, while watching lecture videos, assuming that reaction time for the detection of contents-irrelevant noise is longer when learners are focusing attention more to the contents of the videos. We sought facial features that are useful for predicting the reaction time and found that reaction time can be estimated in some amount from facial features. This result indicates that facial expressions are useful for predicting attention state, or concentration level while attending lectures.