Document Type

Conference Paper


Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence


Electrical and electronic engineering, Health-related biotechnology

Publication Details

In the Proceedings of IEEE EMBSS UKRI PGBIOMED '05, University of Reading, UK, July 18-20, 2005.


In this paper we describe a novel mode of human computer interaction based on gaze tracking using the electro-oculogram (EOG). Despite the relative simplicity of recording this signal, it is often discounted as a reliable method of gaze tracking because of problems arising from changing sensitivity and DC drift. We describe an original technique, Target Position Variation (TPV), which addresses this issue by presenting moving icons which, when followed with the eye, create a corresponding pattern in the EOG signal which can be used to infer the correct gaze position and to compensate for variations in sensitivity.