Measuring the Perceived Difficulty of a Lecture Using Automatic Facial Expression Recognition

This project explores the idea of facial expression for automated feedback in teaching. We show how automatic real-time facial expression recognition can be effectively used to estimate the difficulty level, as perceived by an individual student, of a delivered lecture. We also show that facial expression is predictive of an individual student's preferred rate of curriculum presentation at each moment in time. On a video lecture viewing task, training on less than two minutes of recorded facial expression data and testing on a separate validation set, our system predicted the subjects' self-reported difficulty scores with mean accuracy of 0.42 (Pearson r) and their preferred viewing speeds with mean accuracy of 0.29. Our techniques are fully automatic and have potential applications for both intelligent tutoring systems (ITS) and standard classroom environments.