Experimental Studies of Biometric User Identification Based on the Data of Eye Tracker Tobii X2-30
Abstract
Purpose: Biometric identification based on eye movement analysis has high protection against forgery. The accuracy of this identification method depends on the analyzed parameters of the eye movement, on the applied stimuli and also on the frequency of the device which records the gaze trajectory. The purpose of our experimental studies is to reveal which eye movement parameters provide the best identification results and which stimuli should be used to reveal personal characteristics of one's eye movement. Results: The results of experimental studies of person identification based on eye movement data analysis obtained by eye tracker Tobii X2-30 under two different stimuli are presented. Eye-tracking data obtained by gaze trajectory registration systems, including low-frequency ones, can potentially be used to reveal person's identity. Data analysis for the two experiments showed that the identification error could vary from 0 to 40% over the observed parameter and the gaze event type. The results of the second experiment showed that analyzing saccades' data is more efficient due to the gaze velocity threshold set in order to determine the gaze event type correctly. The second experiment revealed that gaze acceleration parameter of the saccade data could be used to identify a person. Practical relevance: The results obtained during the experimental study proved the opportunity of using low-frequency eye-tracking systems for person identification.Published
2015-10-20
How to Cite
Lyamin, A., & Cherepovskaya, E. (2015). Experimental Studies of Biometric User Identification Based on the Data of Eye Tracker Tobii X2-30. Information and Control Systems, (5), 26-33. https://doi.org/10.15217/issn1684-8853.2015.5.26
Issue
Section
Information processing and control