Biometric Based E-Learning Application To Detect Learner Attention Prediction

Abstract

Dynamic student’s attention monitoring in the class rooms, online study is essential. Existing model fail to predict the student attention, understanding, engagement automatically using biometric features. Current traditional system is manual with less efficiency and time consuming. Even in both class room based traditional education, e-learning systems, dynamic learner behaviour is not considered. To address this problem statement, this survey paper explores how dynamic learner behaviour of audience, students can be obtained using computer vision and machine learning techniques to predict the engagement level. This research paper also focus on an image processing with deep learning process of invoking, extracting biometric features from the students and detect students attention, distraction towards the class and based on the result outcome the proposed system recommends theory and video based content to obtain the attention back. For deep learning model, we use Convolutional Neural Network (CNN) for training the input models.