Classroom Instrumentation for Gaze Behavioral Tracking Detection
Songpol Bunyang, Thitirat Siriborvornratanakul · 2025
Enhancing learning outcomes and advancing teaching practices in the education sector require access to realworld classroom data. A critical factor is the attentiveness and interest of learners, which offers valuable insights for educators to optimize teaching methods and create enriched educational environments. However, traditional gaze-tracking systems have not gained widespread adoption due to their reliance on sophisticated sensors, which often lack practicality and efficiency in real-classroom settings. To address this limitation, this paper introduces a gaze-sensing system specifically designed for classroom deployments, coupled with a cost-effective processing system to enable broader implementation. The proposed approach assesses class gaze patterns by estimating individual six degrees of freedom (6-DOF) head movements using two RGB cameras. This system also provides educators with actionable feedback by visualizing students' gaze behaviors, enabling them to adapt instructional strategies to meet learners' needs effectively. By automating the detection and analysis of gaze behaviors, this research equips educators with objective, pedagogically relevant data, empowering them to make informed decisions. Furthermore, the proposed platform offers extensibility for future developments, paving the way for further advancements in data-driven educational practices.