Survey on Eye Movement Based Authentication Systems

Abstract

No matter how sophisticated an authentication system has been devised, human is often considered as the weakest link in the security chain. Security problems can stem from bad interactions between humans and systems. Eye movement is a natural interaction modality. The application of eye tracking technology in authentication offers a promising and feasible solution to the trading-off between the usability and the security of an authentication system. This paper conducts a comprehensive survey on existing Eye Movement Based Authentication (EMBA) methodologies and systems, and briefly outlines the technical and methodological aspects of EMBA systems. We decompose the EMBA technique into three fundamental aspects: (1) eye movement input modality, (2) eye movement interaction mechanism, and (3) eye movement data recognition. The features and functions of the EMBA modules are further analyzed. An emphasis is put on the interrelationship among the modules and their general impacts on the formation and function of the EMBA framework. The paper attempts to provide a systemic treatment on the state of the art technology and also to outline some potential future development directions in eye movement based interaction or security systems.