Eye movement event detection is crucial for understanding human visual behaviour, and can be used in multiple use cases, e.g., visual attention monitoring [1]. This thesis will focus on implementing and comparing between multiple eye movement event detection algorithms, specifically targeting head-mounted eye trackers. The algorithms include well-established baselines such as I-VT (Velocity-Threshold Identification) [4] and I-DT (Dispersion-Threshold Identification) [4], and more advanced approaches such as those proposed by Drews & Dierkes (2024) [2], Nejad et al. (2024) [3], and Steil et al. (2018) [6].

The thesis can also focus on exploring novel machine learning approaches, e.g. [5], to differentiate between various events. By offering multiple detection algorithms and robust pre-processing techniques, the goal is to provide a comprehensive solution for real-time eye movement event detection.

 

Skills Required: This project requires machine learning and software development skills. The topic can be suitable and adjusted for both a bachelor’s and a master’s thesis.

If interested, please send an email to