Distraction detection of lectures in e-learning using machine learning based on human facial features and postural information

While e-learning lectures allow students to learn at their own pace, it is difficult to manage students’ concentration, which prevents them from receiving valuable information from lectures. Therefore, we propose a method for detecting student distraction during e-learning lectures using machine learning, based on human face and posture information that can be collected using only an ordinary web camera. In this study, we first collected video data of the faces of subjects taking e-learning lectures and used the OpenFace and GAST-Net libraries to obtain face and posture information. Next, from the face and posture data, we extracted features such as the area of the eyes and mouth, the angle of the gaze direction, and the angle of the neck and shoulders. Finally, we used various machine learning models, such as random forest and XGBoost, to detect states of distraction during e-learning lectures. The results show that our binary classification models trained only on the individual’s data achieved more than 90% recall.