Quantifying attention levels in individualized online tutoring: A case of one-an-one sessions

Abstract

After the COVID-19 pandemic, the worldwide reliance and shift towards video communication platforms have highlighted the importance of virtual learning. One major challenge in this aspect is determining the actual engagement of students during virtual sessions. To address this challenge, we present a research study that focuses on developing a system that will help evaluate participant's attentiveness in one-on-one online sessions. We propose a system that will utilize screen sharing detection, face recognition, head pose estimation, and eye gaze estimation to analyze a recorded tutoring session which will help an expert to assess the attention level of both the student and the tutor. It will provide educators valuable insights to optimize their teaching methods and adapt their strategies to boost the participation of students. As the popularity and demand of the global e-learning market continue to grow, systems such as ours can contribute to making online learning more efficient in both educational and corporate training sectors.