A crowdsourcing-based approach to assess concentration levels of students in class videos

Concentration is important for students to conduct efficient learning in a class, and an effective assessment of students' concentration level in a class is useful for students to review class materials after lessons, as well as for lecturers to adjust their teaching strategies for self-improvement. Although a number of concentration assessment approaches have been proposed, conventional approaches are generally time/money expensive (e.g., expert opinions), inaccurate (e.g., computer vision-based approaches), and intrusive (e.g., wearable sensor-based approaches). In this study, we propose a novel approach, called Concentration Level Assessment System (CLAS), which combines a markovian Doze-and-Wake Model (DAWM) and the emerging crowdsourcing technique to enable effective concentration assessment of class videos. Using realistic datasets of class videos, we conduct a comprehensive set of synthetic analysis and Internet experiments, the results demonstrate that CLAS is capable of yielding an accuracy up to 98% with 86% cost savings. Moreover, CLAS is simple, effective, and scalable, and it shows promises in facilitating advanced applications for efficiency, productivity, and safety in the future.