Detecting Affective States Based on Facial Expressions Among Students Using an Educational Game for Physics

• Main Authors: Tabanao, Emily S, Rodrigo, Ma. Mercedes T.
• Format: text
• Published: Archīum Ateneo 2016
• Subjects: Affect detection; facial expression recognition; Physics Playground; Artificial Intelligence and Robotics; Computer Sciences; Databases and Information Systems; Science and Mathematics Education
• Online Access: https://archium.ateneo.edu/discs-faculty-pubs/319; http://www.et.iitb.ac.in/icce2016/icce2016proceedings.html
• Institution: Ateneo De Manila University

We built logistic regression models to identify facial action units that are associated with affective states of Filipino students using Physics Playground, an educational game for Physics. We found that facial action units AU1 (Inner Brow Raiser), AU5 (Upper Lid Raiser) and AU6 (Cheek Raiser) are significantly associated with engaged concentration, AU20 (Lip Stretcher) with Frustration, AU12 (Lip Corner Puller) with Delight and AU6 with Happiness. We discuss features in the models and future work.