Paper Title
Estimation of Mental States Using Facial Features to Improve A Student’s Performance in E-Learning

Students’ interest and involvement in the content being taught direct impact on their academic performance. E-learning and traditional learning have one major difference i.e., education is delivered directlyby the instructors, and their direct interaction with students also allows them to understand the interest level of each student. In e-learning environment, absence of any type of human supervision creates many limitations, and an important limitation is the understanding the learning attitude of student during lecture and act accordingly to improve his learning interest. This calls for an investigation to identify methodologies that can be helpful in understanding a student’s interest and engagement in e-learning where direct human supervision is not available. This study is carried out to to identify the means to gauge student’s interest and facial feature recognition was employed to extract expression pattern to predict the mental state of the student. A survey was carried out among 198 academics to identify pertinent facial features that can represent the facial expressions and mental alertness of a student. These shortlisted facial features were then used to predict facial expressions in real time and using data association based algorithms, interest level of a student was accurately categorized into 5 emotional/mental states.Different combinations of facial expressions were used to estimate mental states and subsequently interest and engagement of a student in an e-learning environment. The results achieved during the course of research showed that facial expressions of a student can quite accurately provide a measure to understand and calculate the student’s involvement in e-learning environment, without any human supervision. The proposed model can be used in enhancing interactive e-learning as it can provide feedback to the e-learning system tovary the content, change delivery methods, and improve machine user interfacein an in order to engage the student. Index Terms- Mental stateextraction, Facial features; Facial recognition; Online education; E-Learning.