Eye-Tracking Keyboard Using a Webcam: An Exploratory Analysis of Deployment and Calibration
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This paper presents an open-source, webcam-based gaze-typing system implemented in Next.js. It leverages MediaPipe FaceMesh for real-time facial landmark tracking and a lightweight Eye Aspect Ratio (EAR) heuristic for blink detection, enabling hands-free text entry on a virtual keyboard. We describe the pipeline, tunable parameters, deployment strategy, and a calibration-oriented attention heat-map that helps set key “knobs” (EAR threshold, gaze scale, smoothing) to improve accuracy. We also position our contribution in the context of prior browser-native gaze tracking efforts. Overall, the system demonstrates how commodity hardware can deliver low-cost, accessible gaze-based interaction while remaining extensible for advanced machine learning and HCI research.
Fields of Study & Subjects
Gaze Tracking and Assistive Technology
Hand Gesture Recognition Systems
User Authentication and Security Systems
Publication details
- Journal / source
- Figshare
- Publication date
- 29 March 2026
- Language
- EN
- Document type
- Thesis