Real-Time User Attention and Typing Behavior Analysis via MediaPipe Face Mesh and Spatial Threshold Calibration
International Research Journal of Innovation in Science and Technology · 07.11.2026 · 1 (4) · 46-56
DOI: 10.67308/irjist.101 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Research Journal of Innovation in Science and Technology, online erschienen am 07.11.2026. Verfasst von Arnav Wakekar und 5 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Automated user attention mapping frameworks have attained an important role in measuring human-computer interaction (HCI) metrics and online educational methods. Conventional vision-based solutions usually result in substantial computational overhead from convolutional models or do not feature precise geometric thresholds, which cannot be adjusted for different seating positions. This paper suggests an efficient and non-intrusive software system able to track the user’s screen-versus-keyboard attention. More precisely, it is capable of distinguishing when a person is looking at the computer screen and when the person is gazing at a keyboard. The solution uses a normal webcam video stream to receive data through the MediaPipe Face Mesh technology that enables automatic 3D coordinate system construction for getting a vertical head-pose offset. The two-level automatic system calibration process is carried out by means of a special model allowing data processing through asynchronous Flask web server work. In addition, to guarantee that no frame loss occurs at the hardware level, thread safety is ensured via a special mutex lock.
Zitieren
APA
Wakekar, A., Wanjari, K., Yelurkar, A., Wagh, P., Mahatme, S., & Zalte, R. (2026). Real-Time User Attention and Typing Behavior Analysis via MediaPipe Face Mesh and Spatial Threshold Calibration. International Research Journal of Innovation in Science and Technology, 1(4), 46-56. https://doi.org/10.67308/irjist.101
Vancouver
Wakekar A, Wanjari K, Yelurkar A, Wagh P, Mahatme S, Zalte R. Real-Time User Attention and Typing Behavior Analysis via MediaPipe Face Mesh and Spatial Threshold Calibration. International Research Journal of Innovation in Science and Technology. 2026;1(4):46-56. doi:10.67308/irjist.101
BibTeX
@article{Wakekar2026,
author = {Wakekar, Arnav and Wanjari, Kartik and Yelurkar, Anis and Wagh, Padmali and Mahatme, Supriya and Zalte, Rakshit},
title = {Real-Time User Attention and Typing Behavior Analysis via MediaPipe Face Mesh and Spatial Threshold Calibration},
journal = {International Research Journal of Innovation in Science and Technology},
year = {2026},
volume = {1},
number = {4},
pages = {46--56},
doi = {10.67308/irjist.101}
}
EUVIMED