Single-Pixel Eye Tracking via Patterned Contact Lenses: Design and Evaluation in HCI Domain

Abstract

This paper presents a preliminary study of an eye tracking technique suitable for use in devices with low-power consumption demands, e.g. Google Glass. The method uses a patterned contact lens and a single-pixel imaging sensor. Its applicability is explored via a semi-simulated user study, where real eye movements from 50 subjects are used to animate a 3-D graphics replica of an eye wearing a patterned contact lens. An accurate single-pixel camera simulator is used to perform gaze estimation via capturing of the imprinted pattern. The results show the promising potential of the technique in the field of eye tracking and eye gesture recognition.