Facial Emotion Recognition for Virtual Learning Environments to Reduce Dropout Rate

Virtual learning environments (VLEs) are widening opportunities for people of all age groups to learn courses or acquire skills that appeal to them. Recent trends show that there is an increase in the use of such environments and their resources. However, we observe that there is a significant difference between the number of people registered for such courses and the number of people who complete it. We bring about assistance for the tutors monitoring these courses by helping them recognize the emotion of students over these virtual learning environments. By recognizing specific frames that failed to keep the student engaged, we can successfully increase the efficiency of virtual learning environments. We propose to achieve this by recognizing the student’s emotions using a convolution neural network (CNN) model throughout the content provided, recognizing the emotion observed at each frame, and drawing detailed reports to identify the dominant emotion. By observing the dominant emotion, the tutor can take necessary actions, which would in turn help reduce the dropout rate.