Development and Evaluation of a Webcam-Based Eye Tracking System (Bachelor and Master)

Eye tracking has become an essential tool in XR research, with applications ranging from user behavior analysis to gaze-based interaction. This thesis topic aims to develop an open-source eye tracking framework that operates using consumer cameras, such as webcams, or smartphone cameras. The student will survey existing approaches (e.g., WebGazer.js, Beam Eye Tracker), design and implement a prototype, and evaluate its performance against commercial webcam eye trackers in terms of spatial and temporal accuracy, calibration effort, robustness, ease of use, and user experience. Beyond gaze estimation, the framework will support features such as gaze visualization, blink detection, gaze-event detection, and potentially pupillometry. The thesis will cover prototype development, pilot testing, evaluation, data analysis, and the development of a demonstrator, such as a gaze-controlled application or 360° video navigator. Students interested in this topic should have an interest in programming, computer vision, behavioral experiments, and physiological data analysis. Experience with Python or JavaScript is an advantage.

Suitable for Bachelor and Master theses. Contact person: Abhinav Bhattacharya M.Sc.

thesis_abhinav
AI Powered Iris Tracking with WebGazer.Js, image reference: https://www.youtube.com/watch?v=Wh77ZGdIaZQ&t=14s