Real-time Online Learning Attention Tracker
Dexter X. H. Chew, Tee Hui Teo · 2022
As a result of the coronavirus pandemic, online learning has become an essential tool for institutions. Attention-aware systems monitor learners’ attention to ensure an effective e-learning experience, from multi-metric facial analysis tools to electroencephalogram hardware. This paper proposes a low-cost systematic approach that builds on existing web application interfaces to assess students’ attention levels in real-time during online classes. The solution uses an end-to-end approach that estimates average attention scores based on head pose features. Then, this aggregated data can be used to improve teacher efficacy. Test lessons show that the learner attention levels can be effectively captured effectively using a browser-based application hosted on the instructor’s end, with an average of 15 frames per second (FPS).