Diploma Thesis

Visual Analysis of Gaze Data in Virtual Environments

Abstract

Eye tracking technology allows the recording of eye movements for post-hoc analysis of visual attention. One of its major strengths and a reason for its popularity is presenting easily comprehensible, superimposed visualizations for two-dimensional stimuli - for example heat maps and gaze plots. However, such gaze visualization techniques do not adequately support interactive analysis of three-dimensional (3D) virtual environments (VE).

This work intends to facilitate eye tracking studies in 3D VEs by enhancing visual gaze analysis techniques and methodologies. For this purpose, a survey was conducted for investigating improvements of existing gaze analysis techniques. This led to the development of a taxonomy of gaze visualizations to determine strengths and weaknesses of current techniques. Based on feature requests from the survey and findings from the classification of gaze visualizations, a gaze analysis tool was developed that integrates visualizations for static 3D VEs. The implementation is based on the Tobii 1750 Eye Tracker, the XNA Framework, and Windows Forms.

The impact of this thesis is twofold: Firstly it helps eye tracking mature into a technology that can be used to easily investigate visual attention in simulation and, for example, serious gaming environments. For this purpose, a software solution is provided to generate and explore gaze visualizations in VEs. Secondly it serves as a foundation for future research into 3D visualizations of human gaze by providing the methodological and taxonomical framing for these visualization techniques.

Results

Prototyp
Screenshot

Examples for gaze visualizations in SVEETER
(Sophie's Visualization Environment for Eye Tracking & Experimental Research)