Abstract
The appearance of a face is severely altered by illumination conditions that makes automatic face recognition a challenging task. In this paper we propose a Gaussian Mixture Models (GMM)-based human face identification technique built in the Fourier or frequency domain that is robust to illumination changes and does not require “illumination normalization” (removal of illumination effects) prior to application unlike many existing methods. The importance of the Fourier domain phase in human face identification is a well-established fact in signal processing. A maximum a posteriori (or, MAP) estimate based on the posterior likelihood is used to perform identification, achieving misclassification error rates as low as 2% on a database that contains images of 65 individuals under 21 different illumination conditions. Furthermore, a misclassification rate of 3.5% is observed on the Yale database with 10 people and 64 different illumination conditions. Both these sets of results are significantly better than those obtained from traditional PCA and LDA classifiers. Statistical analysis pertaining to model selection is also presented.
Highlights
Biometric Authentication denotes the technique of identifying people based on their unique physical or behavioral traits
In this paper we propose a Gaussian Mixture Models (GMM)-based human face identification technique built in the Fourier or frequency domain that is robust to illumination changes and does not require “illumination normalization” prior to application unlike many existing methods
We use a subset of the publicly available “CMUPIE Database” ([32]) which contains images of 65 people captured under 21 different illumination conditions ranging from shadows to balanced and overall dark
Summary
Biometric Authentication denotes the technique of identifying people based on their unique physical (e.g. face, fingerprints, iris) or behavioral (e.g. gait, voiceprint) traits. The modern world has seen a rapid evolution of the technology of biometric authentication, prompted by an increasing urgency for security following the attacks of 9/11. They are used everywhere today from law enforcement to immigration to e-commerce transactions. Model-based systems use a statistical model to represent the pattern of some facial features (often the ones mentioned above), and some characteristics of the fitted model (parameter estimates, likelihood, etc.) are used as the matching criteria. I represents the model parameters for the ith mixture component and π π1, , πg T is the g vector of the mixing proportions with πi 1.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call