scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

Online learning in biometrics: A case study in face classifier update

TL;DR: The proposed online classifier is used in a face recognition application for classifying genuine and impostor match scores impacted by different covariates and not only improves the verification accuracy but also significantly reduces the computational cost.
Abstract: In large scale applications, hundreds of new subjects may be regularly enrolled in a biometric system. To account for the variations in data distribution caused by these new enrollments, biometric systems require regular re-training which usually results in a very large computational overhead. This paper formally introduces the concept of online learning in biometrics. We demonstrate its application in classifier update algorithms to re-train classifier decision boundaries. Specifically, the algorithm employs online learning technique in a 2ν-Granular Soft Support Vector Machine for rapidly training and updating face recognition systems. The proposed online classifier is used in a face recognition application for classifying genuine and impostor match scores impacted by different covariates. Experiments on a heterogeneous face database of 1,194 subjects show that the proposed online classifier not only improves the verification accuracy but also significantly reduces the computational cost.
Citations
More filters
01 Jan 2010
TL;DR: The next generation biometric technology must overcome many hurdles and challenges to improve the recognition accuracy, including ability to handle poor quality and incomplete data, achieve scalability to accommodate hundreds of millions of users, ensure interoperability, and protect user privacy while reducing system cost and enhancing system integrity.
Abstract: Prevailing methods of human identification based on credentials (identification documents and PIN) are not able to meet the growing demands for stringent security in applications such as national ID cards, border crossings, government benefits, and access control. As a result, biometric recognition, or simply biometrics, which is based on physiological and behavioural characteristics of a person, is being increasingly adopted and mapped to rapidly growing person identification applications. Unlike credentials (documents and PIN), biometric traits (e.g., fingerprint, face, and iris) cannot be lost, stolen, or easily forged; they are also considered to be persistent and unique. Use of biometrics is not new; fingerprints have been successfully used for over one hundred years in law enforcement and forensics to identify and apprehend criminals. But, as biometrics permeates our society, this recognition technology faces new challenges. The design and suitability of biometric technology for person identification depends on the application requirements. These requirements are typically specified in terms of identification accuracy, throughput, user acceptance, system security, robustness, and return on investment. The next generation biometric technology must overcome many hurdles and challenges to improve the recognition accuracy. These include ability to handle poor quality and incomplete data, achieve scalability to accommodate hundreds of millions of users, ensure interoperability, and protect user privacy while reducing system cost and enhancing system integrity. This chapter presents an overview of biometrics, some of the emerging biometric technologies and their limitations, and examines future challenges.

155 citations

Book ChapterDOI
01 Jan 2012
TL;DR: The next generation biometric technology must overcome many hurdles and challenges to improve the recognition accuracy, including ability to handle poor quality and incomplete data, achieve scalability to accommodate hundreds of millions of users, ensure interoperability, and protect user privacy while reducing system cost and enhancing system integrity.
Abstract: Prevailing methods of human identification based on credentials (identification documents and PIN) are not able to meet the growing demands for stringent security in applications such as national ID cards, border crossings, government benefits, and access control. As a result, biometric recognition, or simply biometrics, which is based on physiological and behavioural characteristics of a person, is being increasingly adopted and mapped to rapidly growing person identification applications. Unlike credentials (documents and PIN), biometric traits (e.g., fingerprint, face, and iris) cannot be lost, stolen, or easily forged; they are also considered to be persistent and unique. Use of biometrics is not new; fingerprints have been successfully used for over 100 years in law enforcement and forensics to identify and apprehend criminals. But, as biometrics permeates our society, this recognition technology faces new challenges. The design and suitability of biometric technology for person identification depends on the application requirements. These requirements are typically specified in terms of identification accuracy, throughput, user acceptance, system security, robustness, and return on investment. The next generation biometric technology must overcome many hurdles and challenges to improve the recognition accuracy. These include ability to handle poor quality and incomplete data, achieve scalability to accommodate hundreds of millions of users, ensure interoperability, and protect user privacy while reducing system cost and enhancing system integrity. This chapter presents an overview of biometrics, some of the emerging biometric technologies and their limitations, and examines future challenges.

123 citations


Cites background from "Online learning in biometrics: A ca..."

  • ...Therefore, another aspect of an adaptive biometric system is online learning , which can periodically update the matcher (Singh et al. 2009 ) ....

    [...]

Journal ArticleDOI
TL;DR: An encoding scheme is devised that compresses high-dimensional dense features into a compact representation by maximizing the intrauser correlation and an adaptive feature matching algorithm is developed for effective classification.
Abstract: Dense feature extraction is becoming increasingly popular in face recognition tasks. Systems based on this approach have demonstrated impressive performance in a range of challenging scenarios. However, improvements in discriminative power come at a computational cost and with a risk of over-fitting. In this paper, we propose a new approach to dense feature extraction for face recognition, which consists of two steps. First, an encoding scheme is devised that compresses high-dimensional dense features into a compact representation by maximizing the intrauser correlation. Second, we develop an adaptive feature matching algorithm for effective classification. This matching method, in contrast to the previous methods, constructs and chooses a small subset of training samples for adaptive matching, resulting in further performance gains. Experiments using several challenging face databases, including labeled Faces in the Wild data set, Morph Album 2, CUHK optical-infrared, and FERET, demonstrate that the proposed approach consistently outperforms the current state of the art.

49 citations


Cites background or methods from "Online learning in biometrics: A ca..."

  • ...The feature matching models used in practice fall into two categories: discriminative models [7]–[13], [47], [48], [51], [52], [54] and generative models [14]–[16]....

    [...]

  • ...On the other hand, discriminative approaches focus more directly on the classification task and thus can yield superior performance over the generative methods; representative methods include [7]–[13], [47], [48], [51], [52], [54]....

    [...]

Journal ArticleDOI
TL;DR: In this work, the performance of the following imputation methods are compared in the context of multibiometric fusion and it is observed that the Gaussian mixture model (GMM)-based KNN imputation scheme results in the best recognition accuracy.

48 citations

Journal ArticleDOI
TL;DR: Eight different scores are proposed and evaluated in order to quantify the differences between gestures, obtaining an optimal EER result of 3.42% when analyzing a random set of 40 users of a database made up of 80 users with real attempts of falsification.

11 citations


Cites background from "Online learning in biometrics: A ca..."

  • ...Actually, this is a common practice in other biometric techniques where the biometric characteristic may change over time, as in hand recognition (Amayeh et al., 2009), fingerprint (Freni et al., 2008) or face (Singh et al, 2009; Marcialis et al, 2008; Rattani et al., 2008)....

    [...]

References
More filters
Book
Vladimir Vapnik1
01 Jan 1995
TL;DR: Setting of the learning problem consistency of learning processes bounds on the rate of convergence ofLearning processes controlling the generalization ability of learning process constructing learning algorithms what is important in learning theory?
Abstract: Setting of the learning problem consistency of learning processes bounds on the rate of convergence of learning processes controlling the generalization ability of learning processes constructing learning algorithms what is important in learning theory?.

40,147 citations


"Online learning in biometrics: A ca..." refers methods in this paper

  • ..., N , K(xi,xj) is the kernel function [15], αi, αj are the Lagrange multipliers such that 0 ≤ αi ≤ Ci, ∑ i αiyi = 0, and ∑ i αi ≥ ν....

    [...]

  • ...C is a regularization factor between the total distance of each error from the margin and the width of the margin, and ψi is the slack variable used to allow classification errors [15]....

    [...]

Book
01 Jan 2009
TL;DR: The aim of this book is to provide a Discussion of Constrained Optimization and its Applications to Linear Programming and Other Optimization Problems.
Abstract: Preface Table of Notation Part 1: Unconstrained Optimization Introduction Structure of Methods Newton-like Methods Conjugate Direction Methods Restricted Step Methods Sums of Squares and Nonlinear Equations Part 2: Constrained Optimization Introduction Linear Programming The Theory of Constrained Optimization Quadratic Programming General Linearly Constrained Optimization Nonlinear Programming Other Optimization Problems Non-Smooth Optimization References Subject Index.

7,278 citations


"Online learning in biometrics: A ca..." refers methods in this paper

  • ...In this process, the KarushKuhn Tucker conditions [4] are maintained so that the 2νOGSSVM provides an optimal decision hyperplane....

    [...]

01 Oct 2008
TL;DR: The database contains labeled face photographs spanning the range of conditions typically encountered in everyday life, and exhibits “natural” variability in factors such as pose, lighting, race, accessories, occlusions, and background.
Abstract: Most face databases have been created under controlled conditions to facilitate the study of specific parameters on the face recognition problem. These parameters include such variables as position, pose, lighting, background, camera quality, and gender. While there are many applications for face recognition technology in which one can control the parameters of image acquisition, there are also many applications in which the practitioner has little or no control over such parameters. This database, Labeled Faces in the Wild, is provided as an aid in studying the latter, unconstrained, recognition problem. The database contains labeled face photographs spanning the range of conditions typically encountered in everyday life. The database exhibits “natural” variability in factors such as pose, lighting, race, accessories, occlusions, and background. In addition to describing the details of the database, we provide specific experimental paradigms for which the database is suitable. This is done in an effort to make research performed with the database as consistent and comparable as possible. We provide baseline results, including results of a state of the art face recognition system combined with a face alignment system. To facilitate experimentation on the database, we provide several parallel databases, including an aligned version.

5,742 citations


"Online learning in biometrics: A ca..." refers background in this paper

  • ...The Labeled Faces in the Wild database [6] contains real world images of celebrities and popular individuals (this database contains images of more than 1,600 subjects from which we selected 294 subjects that have at least 6 images)....

    [...]

Journal ArticleDOI
TL;DR: Two of the most critical requirements in support of producing reliable face-recognition systems are a large database of facial images and a testing procedure to evaluate systems.
Abstract: Two of the most critical requirements in support of producing reliable face-recognition systems are a large database of facial images and a testing procedure to evaluate systems. The Face Recognition Technology (FERET) program has addressed both issues through the FERET database of facial images and the establishment of the FERET tests. To date, 14,126 images from 1,199 individuals are included in the FERET database, which is divided into development and sequestered portions of the database. In September 1996, the FERET program administered the third in a series of FERET face-recognition tests. The primary objectives of the third test were to 1) assess the state of the art, 2) identify future areas of research, and 3) measure algorithm performance.

4,816 citations


"Online learning in biometrics: A ca..." refers methods in this paper

  • ...The AR face database [8] contains face images with varying illumination and accessories, and the FERET database [10] has face images with different variations over a time interval of 3-4 years....

    [...]