Recent developments in mobile eye-tracking technology have enabled various fields of research to study visual perception and human behavior in natural environments. Due to several sources of noise in such settings, e.g., motion blur, changing illuminations conditions, reflections, etc., the recorded eye movement data lacks robustness. Therefore, data collected in studies outside the laboratory has to be manually post-processed, which is a laborious and time-consuming procedure. Furthermore, post-processing constitutes the major bottleneck in the implementation of gazed-based and real-time applications. In this talk, I will first introduce a novel analysis pipeline based on machine learning techniques to approach robust and real-time eye movements’ analysis. I will then showcase the applicability of these methods to several use cases in highly dynamic scenarios, showing thus, how such methods can pave the way for new gaze-based technology in everyday life.
Prof. Dr. Enkelejda Kasneci, Universität Tübingen
Ort: INB Seminarraum
