Point of Confusion Estimation Using Facial Features and Gaze Tracking

Abstract

Confusion is an important part of the learning process, but prolonged periods of confusion can have a detrimental effect. Online education environments lack the benefit of face-to-face interaction and instantaneous detection of confusion by teachers. We built an inobtrusive, inexpensive and scalable confusion tracker using only a webcam and a web browser. In this first implementation, the confusion estimator outperforms randomly generated estimates and provides evidence of efficacy in support of an argument to conduct further research.