Corey Holland

Biometric identification via eye movement scanpaths in reading

This paper presents an objective evaluation of various eye movement-based biometric features and their ability to accurately and precisely distinguish unique individuals. Eye movements are uniquely counterfeit resistant due to the complex neurological interactions and the extraocular muscle properties involved in their generation. Considered biometric candidates cover a number of basic eye movements and their aggregated scanpath characteristics, including: fixation count, average fixation duration, average saccade amplitudes, average saccade velocities, average saccade peak velocities, the velocity waveform, scanpath length, scanpath area, regions of interest, scanpath inflections, the amplitude-duration relationship, the main sequence relationship, and the pairwise distance between fixations. As well, an information fusion method for combining these metrics into a single identification algorithm is presented. With limited testing this method was able to identify subjects with an equal error rate of 27%. These results indicate that scanpath-based biometric identification holds promise as a behavioral biometric technique.

Complex eye movement pattern biometrics: Analyzing fixations and saccades

This paper presents an objective evaluation of previously unexplored biometric techniques utilizing patterns identifiable in human eye movements to distinguish individuals. The distribution of primitive eye movement features are compared between eye movement recordings using algorithms based on the following statistical tests: the Ansari-Bradley test, the Mann-Whitney U-test, the two-sample Kolmogorov-Smirnov test, the two-sample t-test, and the two-sample Cramer-von Mises test. Score-level information fusion is applied and evaluated by: weighted mean, support vector machine, random forest, and likelihood ratio. The accuracy of each comparison/jusion algorithm is evaluated, with results suggesting that, on high resolution eye tracking equipment, it is possible to obtain equal error rates of 16.5% and rank-1 identification rates of 82.6% using the two-sample Cramér-von Mises test and score-level information fusion by random forest, the highest accuracy results on the considered dataset.

Eye tracking on unmodified common tablets: challenges and solutions

This work describes the design and implementation of an eye tracking system on an unmodified common tablet PC. A neural network eye tracker is employed as a solution to eye tracking in the visible spectrum of light. We discuss the challenges related to image recognition and processing, and provide an objective evaluation of the accuracy and sampling rate of eye-gaze-based interaction with such an eye tracker. The results indicate that it is possible to obtain an average accuracy of 4.42° and a sampling rate of 0.70 Hz with the described system.

Multimodal ocular biometrics approach: A feasibility study

Growing efforts have been concentrated on the development of alternative biometric recognition strategies, the intended goal to increase the accuracy and counterfeit-resistance of existing systems without increased cost. In this paper, we propose and evaluate a novel biometric approach using three fundamentally different traits captured by the same camera sensor. Considered traits include: 1) the internal, non-visible, anatomical properties of the human eye, represented by Oculomotor Plant Characteristics (OPC); 2) the visual attention strategies employed by the brain, represented by Complex Eye Movement patterns (CEM); and, 3) the unique physical structure of the iris. Our experiments, performed using a low-cost web camera, indicate that the combined ocular traits improve the accuracy of the resulting system. As a result, the combined ocular traits have the potential to enhance the accuracy and counterfeit-resistance of existing and future biometric systems.

Complex Eye Movement Pattern Biometrics: The Effects of Environment and Stimulus

This paper presents an objective evaluation of the effects of eye tracking specification and stimulus presentation on the biometric viability of complex eye movement patterns. Six spatial accuracy tiers (0.5^°, 1.0^°, 1.5^°, 2.0^°, 2.5^°, 3.0^°), six temporal resolution tiers (1000, 500, 250, 120, 75, 30 Hz), and five stimulus types (simple, complex, cognitive, textual, random) are evaluated to identify acceptable conditions under which to collect eye movement data. The results suggest the use of eye tracking equipment capable of at least 0.5^° spatial accuracy and 250 Hz temporal resolution for biometric purposes, whereas stimulus had little effect on the biometric viability of eye movements.

Attack of Mechanical Replicas: Liveness Detection With Eye Movements

This paper investigates liveness detection techniques in the area of eye movement biometrics. We investigate a specific scenario, in which an impostor constructs an artificial replica of the human eye. Two attack scenarios are considered: 1) the impostor does not have access to the biometric templates representing authentic users, and instead utilizes average anatomical values from the relevant literature and 2) the impostor gains access to the complete biometric database, and is able to employ exact anatomical values for each individual. In this paper, liveness detection is performed at the feature and match score levels for several existing forms of eye movement biometric, based on different aspects of the human visual system. The ability of each technique to differentiate between live and artificial recordings is measured by its corresponding false spoof acceptance rate, false live rejection rate, and classification rate. The results suggest that eye movement biometrics are highly resistant to circumvention by artificial recordings when liveness detection is performed at the feature level. Unfortunately, not all techniques provide feature vectors that are suitable for liveness detection at the feature level. At the match score level, the accuracy of liveness detection depends highly on the biometric techniques employed.

Biometric verification via complex eye movements: The effects of environment and stimulus

This paper presents an objective evaluation of the effects of stimulus type and eye tracking specifications on the accuracy of biometric verification based on complex eye movement patterns (CEM). Five stimulus types (simple, complex, cognitive, random, textual), six spatial accuracy tiers (0.5°, 1.0°, 1.5°, 2.0°, 2.5°, 3.0°), and six temporal resolution tiers (1000 Hz, 500 Hz, 250 Hz, 120 Hz, 75 Hz, 30 Hz) are evaluated to identify their effects. The results suggest the use of eye tracking equipment capable of 0.5° spatial accuracy and 250 Hz temporal resolution for biometric purposes, though biometric accuracy remains achievable for systems capable of at least 1.0° spatial accuracy and 30 Hz temporal resolution. While not conclusive, the complex and textual pattern stimuli provided the greatest accuracy, with little difference between the remaining stimuli.

Usability evaluation of eye tracking on an unmodified common tablet

This paper describes the design, implementation, and usability evaluation of a neural network based eye tracking system on an unmodified common tablet and discusses the challenges and implications of neural networks as an eye tracking component on a mobile platform. We objectively and subjectively evaluate the usability and performance tradeoffs of calibration, one of the fundamental components of eye tracking. The described system obtained an average spatial accuracy of 3.95° and an average temporal resolution of 0.65 Hz during trials. Results indicate that an increased neural network training set may be utilized to increase spatial accuracy, at the cost of greater physical effort and fatigue.

Biometric authentication via complex oculomotor behavior

This paper presents an objective evaluation of previously unexplored biometric techniques utilizing patterns identifiable in complex oculomotor behavior to distinguish individuals. Considered features include: saccadic dysmetria, compound saccades, dynamic overshoot, and express saccades. Score-level information fusion is applied and evaluated by: likelihood ratio, support vector machine, and random forest. The results suggest that it is possible to obtain equal error rates of 25% and rank-1 identification rates of 47% using score-level fusion by likelihood ratio.

CUE: counterfeit-resistant usable eye movement-based authentication via oculomotor plant characteristics and complex eye movement patterns

The widespread use of computers throughout modern society introduces the necessity for usable and counterfeit-resistant authentication methods to ensure secure access to personal resources such as bank accounts, e-mail, and social media. Current authentication methods require tedious memorization of lengthy pass phrases, are often prone to shouldersurfing, and may be easily replicated (either by counterfeiting parts of the human body or by guessing an authentication token based on readily available information). This paper describes preliminary work toward a counterfeit-resistant usable eye movement-based (CUE) authentication method. CUE does not require any passwords (improving the memorability aspect of the authentication system), and aims to provide high resistance to spoofing and shoulder-surfing by employing the combined biometric capabilities of two behavioral biometric traits: 1) oculomotor plant characteristics (OPC) which represent the internal, non-visible, anatomical structure of the eye; 2) complex eye movement patterns (CEM) which represent the strategies employed by the brain to guide visual attention. Both OPC and CEM are extracted from the eye movement signal provided by an eye tracking system. Preliminary results indicate that the fusion of OPC and CEM traits is capable of providing a 30% reduction in authentication error when compared to the authentication accuracy of individual traits.