Multi-label Disengagement and Behavior Prediction in Online Learning

Student disengagement prediction in online learning environments is beneficial in various ways, especially to help provide timely cues to make some feedback or stimuli to the students. In this work, we propose a neural network-based model to predict students' disengagement, as well as other behavioral cues, which might be relevant to students' performance, using facial image sequences. For training and evaluating our model, we collected samples from multiple participants and annotated them with temporal segments of disengagement and other relevant behavioral cues with our multiple in-house annotators. We present prediction results of all behavior cues along with baseline comparison.