Nonlinear Dynamic Model of the Oculo-Motor System Human Based on the Volterra Series
摘要
Eye tracking is a powerful tool that decodes eye movements and translates them into insights. Information such as pupil position, the gaze vector, and gaze point form physical characteristics that can be used in biometric systems. The data of the human oculo-motor system (OMS) responses to the test visual stimuli was obtained by use of the Tobii Pro TX300 screen-based eye tracker. The purpose of this experiment was to test the effectiveness of a new method of biometric identification based upon the definition of the integral Volterra model of the OMS in accordance with “input-output” research. Pursuant to the data received, determined first, second and third orders transient functions for two people. The informativeness of the proposed heuristic features was investigated to build a personality classifier. Resistant to computational errors pairs of features were discovered—the probability of correct recognition (PCR) value is in the range 0.92–0.97.