A hyperbolic function model for multiple biometrics decision fusion

Kar Ann Toh, Xudong Jiang, Wei Yun Yau

Research output: Contribution to journalArticle

Abstract

In this paper, we treat the problem of combining fingerprint and speech biometric decisions as a classifier fusion problem. The Feed-forward Neural Network provides a natural choice for such data fusion as it has been shown to be a universal approximator. However, the training process remains much to be a trial-and-error effort since no learning algorithm can guarantee convergence to optimal solution within finite iterations. In this work, we propose a network model to generate different combinations of the hyperbolic functions to achieve some approximation and classification properties. This is to circumvent the iterative training problem as seen in neural networks learning. The proposed hyperbolic functions network model is applied to combine the fingerprint and speaker verification decisions which show either better or comparable results with respect to several commonly used methods.

Original languageEnglish
Pages (from-to)655-662
Number of pages8
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3072
Publication statusPublished - 2004 Dec 1

Fingerprint

Decision Fusion
Hyperbolic functions
Hyperbolic function
Biometrics
Network Model
Feedforward neural networks
Data fusion
Fingerprint Verification
Classifier Fusion
Learning algorithms
Speaker Verification
Classifiers
Trial and error
Data Fusion
Feedforward Neural Networks
Fingerprint
Neural networks
Learning Algorithm
Optimal Solution
Model

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

@article{27ae4c819abb4255bcadf44a3b0ae30a,
title = "A hyperbolic function model for multiple biometrics decision fusion",
abstract = "In this paper, we treat the problem of combining fingerprint and speech biometric decisions as a classifier fusion problem. The Feed-forward Neural Network provides a natural choice for such data fusion as it has been shown to be a universal approximator. However, the training process remains much to be a trial-and-error effort since no learning algorithm can guarantee convergence to optimal solution within finite iterations. In this work, we propose a network model to generate different combinations of the hyperbolic functions to achieve some approximation and classification properties. This is to circumvent the iterative training problem as seen in neural networks learning. The proposed hyperbolic functions network model is applied to combine the fingerprint and speaker verification decisions which show either better or comparable results with respect to several commonly used methods.",
author = "Toh, {Kar Ann} and Xudong Jiang and Yau, {Wei Yun}",
year = "2004",
month = "12",
day = "1",
language = "English",
volume = "3072",
pages = "655--662",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - A hyperbolic function model for multiple biometrics decision fusion

AU - Toh, Kar Ann

AU - Jiang, Xudong

AU - Yau, Wei Yun

PY - 2004/12/1

Y1 - 2004/12/1

N2 - In this paper, we treat the problem of combining fingerprint and speech biometric decisions as a classifier fusion problem. The Feed-forward Neural Network provides a natural choice for such data fusion as it has been shown to be a universal approximator. However, the training process remains much to be a trial-and-error effort since no learning algorithm can guarantee convergence to optimal solution within finite iterations. In this work, we propose a network model to generate different combinations of the hyperbolic functions to achieve some approximation and classification properties. This is to circumvent the iterative training problem as seen in neural networks learning. The proposed hyperbolic functions network model is applied to combine the fingerprint and speaker verification decisions which show either better or comparable results with respect to several commonly used methods.

AB - In this paper, we treat the problem of combining fingerprint and speech biometric decisions as a classifier fusion problem. The Feed-forward Neural Network provides a natural choice for such data fusion as it has been shown to be a universal approximator. However, the training process remains much to be a trial-and-error effort since no learning algorithm can guarantee convergence to optimal solution within finite iterations. In this work, we propose a network model to generate different combinations of the hyperbolic functions to achieve some approximation and classification properties. This is to circumvent the iterative training problem as seen in neural networks learning. The proposed hyperbolic functions network model is applied to combine the fingerprint and speaker verification decisions which show either better or comparable results with respect to several commonly used methods.

UR - http://www.scopus.com/inward/record.url?scp=35048898880&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35048898880&partnerID=8YFLogxK

M3 - Article

VL - 3072

SP - 655

EP - 662

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -