A Semiparametric Model for Bayesian Reader Identification

We study the problem of identifying individuals based on their characteristic gaze patterns during reading of arbitrary text. The motivation for this problem is an unobtrusive biometric setting in which a user is observed during access to a document, but no specific challenge protocol requiring the user's time and attention is carried out. Existing models of individual differences in gaze control during reading are either based on simple aggregate features of eye movements, or rely on parametric density models to describe, for instance, saccade amplitudes or word fixation durations. We develop flexible semiparametric models of eye movements during reading in which densities are inferred under a Gaussian process prior centered at a parametric distribution family that is expected to approximate the true distribution well. An empirical study on reading data from 251 individuals shows significant improvements over the state of the art.

[1]  Cecilia R. Aragon,et al.  Biometric identification via an oculomotor plant mathematical model , 2010, ETRA.

[2]  Ioannis Rigas,et al.  Human eye movements as a trait for biometrical identification , 2012, 2012 IEEE Fifth International Conference on Biometrics: Theory, Applications and Systems (BTAS).

[3]  Tomi Kinnunen,et al.  Eye-Movements as a Biometric , 2005, SCIA.

[4]  Yoshinobu Kano,et al.  Predicting Word Fixations in Text with a CRF Model for Capturing General Reading Strategies among Readers , 2012 .

[5]  E. B. Huey.  The Psychology And Pedagogy Of Reading , 1908 .

[6]  Pawel Kasprowski,et al.  Eye Movements in Biometrics , 2004, ECCV Workshop BioAW.

[7]  Oleg V. Komogortsev,et al.  Biometric identification via eye movement scanpaths in reading , 2011, 2011 International Joint Conference on Biometrics (IJCB).

[8]  Anders Søgaard,et al.  With Blinkers on: Robust Prediction of Eye Movements across Readers , 2013, EMNLP.

[9]  Ryan P. Adams,et al.  The Gaussian Process Density Sampler , 2008, NIPS.

[10]  Ying Yang,et al.  Penalized semiparametric density estimation , 2009, Stat. Comput..

[11]  P. Lenk,et al.  Bayesian Semiparametric Density Estimation and Model Verification Using a Logistic–Gaussian Process , 2003 .

[12]  J. Heister,et al.  Analysing large datasets of eye movements during reading: Julian Heister, Kay-Michael Würzner, and Reinhold Kliegl , 2012 .

[13]  Tom Leonard.  Density Estimation, Stochastic Processes and Prior Information , 1978 .

[14]  Ioannis Rigas,et al.  Biometric Recognition via Eye Movements: Saccadic Vigor and Acceleration Cues , 2016, TAP.

[15]  K. Rayner.  Eye movements in reading and information processing: 20 years of research. , 1998, Psychological bulletin.

[16]  Ralf Engbert,et al.  Tracking the mind during reading: the influence of past, present, and future words on fixation durations. , 2006, Journal of experimental psychology. General.

[17]  Oleg V. Komogortsev,et al.  Person verification via eye movement-driven text reading model , 2015, 2015 IEEE 7th International Conference on Biometrics Theory, Applications and Systems (BTAS).

[18]  Miles A. Tinker.  Studies in the Psychology of Reading. , 1953 .

[19]  Reinhold Kliegl,et al.  SWIFT: a dynamical model of saccade generation during reading. , 2005, Psychological review.

[20]  Ioannis Rigas,et al.  Biometric identification based on the eye movements and graph matching techniques , 2012, Pattern Recognit. Lett..

[21]  Jayanta K. Ghosh,et al.  Bayesian density regression with logistic Gaussian process and subspace projection , 2010 .

[22]  Guy T. Buswell,et al.  Studies in the psychology of reading , 1952 .

[23]  Erik D. Reichle,et al.  Toward a model of eye movement control in reading. , 1998, Psychological review.

[24]  J. Adelman.  Meaning and context, individuals and development , 2012 .

[25]  N. Hjort,et al.  Nonparametric Density Estimation with a Parametric Start , 1995 .

[26]  Reinhold Kliegl,et al.  A Model of Individual Differences in Gaze Control During Reading , 2014, EMNLP.