Biometric Authentication: A Machine Learning Approach [Hardcover]

S.Y. Kung (Author), M.W. Mak (Author), S.H. Lin (Author)

Book Description

Publication Date: September 24, 2004 | ISBN-10: 0131478249 | ISBN-13: 978-0131478244

• A breakthrough approach to improving biometrics performance

• Constructing robust information processing systems for face and voice recognition

• Supporting high-performance data fusion in multimodal systems

• Algorithms, implementation techniques, and application examples

Machine learning: driving significant improvements in biometric performance

As they improve, biometric authentication systems are becoming increasingly indispensable for protecting life and property. This book introduces powerful machine learning techniques that significantly improve biometric performance in a broad spectrum of application domains.

Three leading researchers bridge the gap between research, design, and deployment, introducing key algorithms as well as practical implementation techniques. They demonstrate how to construct robust information processing systems for biometric authentication in both face and voice recognition systems, and to support data fusion in multimodal systems.

Coverage includes:

• How machine learning approaches differ from conventional template matching

• Theoretical pillars of machine learning for complex pattern recognition and classification

• Expectation-maximization (EM) algorithms and support vector machines (SVM)

• Multi-layer learning models and back-propagation (BP) algorithms

• Probabilistic decision-based neural networks (PDNNs) for face biometrics

• Flexible structural frameworks for incorporating machine learning subsystems in biometric applications

• Hierarchical mixture of experts and inter-class learning strategies based on class-based modular networks

• Multi-cue data fusion techniques that integrate face and voice recognition

• Application case studies

Editorial Reviews

From the Back Cover

• A breakthrough approach to improving biometrics performance

• Constructing robust information processing systems for face and voice recognition

• Supporting high-performance data fusion in multimodal systems

• Algorithms, implementation techniques, and application examples

Machine learning: driving significant improvements in biometric performance

As they improve, biometric authentication systems are becoming increasingly indispensable for protecting life and property. This book introduces powerful machine learning techniques that significantly improve biometric performance in a broad spectrum of application domains.

Three leading researchers bridge the gap between research, design, and deployment, introducing key algorithms as well as practical implementation techniques. They demonstrate how to construct robust information processing systems for biometric authentication in both face and voice recognition systems, and to support data fusion in multimodal systems.

Coverage includes:

• How machine learning approaches differ from conventional template matching

• Theoretical pillars of machine learning for complex pattern recognition and classification

• Expectation-maximization (EM) algorithms and support vector machines (SVM)

• Multi-layer learning models and back-propagation (BP) algorithms

• Probabilistic decision-based neural networks (PDNNs) for face biometrics

• Flexible structural frameworks for incorporating machine learning subsystems in biometric applications

• Hierarchical mixture of experts and inter-class learning strategies based on class-based modular networks

• Multi-cue data fusion techniques that integrate face and voice recognition

• Application case studies

About the Author

Sun-Yuan Kung is a professor of electrical engineering at Princeton University. His research and teaching interests include VLSI signal processing; neural networks; digital signal, image, and video processing; and multimedia information systems. His books include VLSI Array Processors and Digital Neural Networks (Prentice Hall PTR).

Man-Wai Mak is an assistant professor at The Hong Kong Polytechnic University and chairman of the IEEE Hong Kong Section Computer Chapter. His research interests include speaker recognition, machine learning, and neural networks.

Shang-Hung Lin is a senior architect at Nvidia, a leader in video and imaging products.

See all Editorial Reviews

Product Details

• Hardcover: 496 pages
• Publisher: Prentice Hall (September 24, 2004)
• Language: English
• ISBN-10: 0131478249
• ISBN-13: 978-0131478244
• Product Dimensions: 9.6 x 7.3 x 1.4 inches
• Shipping Weight: 2.5 pounds
• Average Customer Review: 2.0 out of 5 stars  See all reviews (1 customer review)
• Amazon Best Sellers Rank: #2,017,301 in Books (See Top 100 in Books)

More About the Author

Discover books, learn about writers, read author blogs, and more.

Customer Reviews

Most Helpful Customer Reviews

2.0 out of 5 stars Too much information, not enough detail, February 8, 2007

By 

calvinnme - See all my reviews
(HALL OF FAME REVIEWER)    (VINE VOICE)    (TOP 50 REVIEWER)   

This review is from: Biometric Authentication: A Machine Learning Approach (Hardcover)

Any time you can pick up a used copy of a recently published technical book on an interesting topic at one-fourth of the retail price, you know there must be a problem. You would be right. This book tries to do three things at the same time and fails with at least two of its goals. It tries to talk about the business issues of biometrics, technical issues of specific biometric technologies (face recognition, speech recogniton, etc.), and finally machine learning techniques used to accomplish the biometric measurements. Only at this last goal do I think the book comes close to success, and even then only on a sketchy high level. The first seven chapters do an OK job of explaining machine learning techniques and give you some very instructive figures that are often lacking in academic textbooks, especially on neural networks. Also, these chapters do a pretty good job of explaining the equations involved. What's lacking, though, even in these early chapters, are some simple numerical examples or algorithmic steps that would give you some guidance on how to approach a task. When the book tries to make the leap to connecting the machine learning techniques to biometric authentication in a meaningful way such that a computer scientist could code up an algorithm, the book really falls on its face. There are some nice block diagrams of biometric systems, but no real details on algorithmic steps that would allow you to realize any of those blocks. Instead, there is quite a bit of verbage on the competition involved on building particular kinds of systems, and some rhetoric on possible pitfalls in specific biometric designs. However, with you standing there not knowing where to start with your design, this advice is really not very helpful.

I would say pass on this book and if you need to learn machine learning techniques, start with the older book by Mitchell entitled "Machine Learning". It talks about all of the machine learning techniques mentioned in this book, plus there are plenty of examples. Used copies are still relatively inexpensive, and its content is accessible and complete. As for biometric techniques, I've found the best books concentrate on one technique, such as fingerprint verification, and don't stray into other forms of authentication. The following is the table of contents:

Chapter 1. Overview

Chapter 2. Biometric Authentication Systems

Chapter 3. Expectation-Maximization Theory

Chapter 4. Support Vector Machines

Chapter 5. Multi-Layer Neural Networks

Chapter 6. Modular and Hierarchical Networks

Chapter 7. Decision-Based Neural Networks

Chapter 8. Biometric Authentication by Face Recognition

Chapter 9. Biometric Authentication by Voice Recognition

Chapter 10. Multicue Data Fusion

Appendix A: Convergence Properties of EM

Appendix B: Average Det Curves

Appendix C: Matlab Projects