Service-oriented architecture based on biometric using random features and incremental neural networks

Kwontaeg Choi, Kar Ann Toh, Youngjung Uh, Hyeran Byun

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

We propose a service-oriented architecture based on biometric system where training and classification tasks are used by millions of users via internet connection. Such a large-scale biometric system needs to consider template protection, accuracy and efficiency issues. This is a challenging problem since there are tradeoffs among these three issues. In order to simultaneously handle these issues, we extract both global and local features via controlling the sparsity of random bases without training. Subsequently, the extracted features are fused with a sequential classifier. In the proposed system, the random basis features are not stored for security reason. The non-training based on feature extraction followed by a sequential learning contributes to computational efficiency. The overall accuracy is consequently improved via an ensemble of classifiers. We evaluate the performance of the proposed system using equal error rate under a stolen-token scenario. Our experimental results show that the proposed method is robust over severe local deformation with efficient computation for simultaneous transactions. Although we focus on face biometrics in this paper, the proposed method is generic and can be applied to other biometric traits.

Original languageEnglish
Pages (from-to)1539-1553
Number of pages15
JournalSoft Computing
Volume16
Issue number9
DOIs
Publication statusPublished - 2012 Sep 1

Fingerprint

Service-oriented Architecture
Biometrics
Service oriented architecture (SOA)
Neural Networks
Neural networks
Classifiers
Classifier
Local Features
Computational efficiency
Sparsity
Computational Efficiency
Feature Extraction
Transactions
Error Rate
Feature extraction
Template
Ensemble
Trade-offs
Face
Internet

All Science Journal Classification (ASJC) codes

  • Software
  • Theoretical Computer Science
  • Geometry and Topology

Cite this

@article{da4499b6866543f78c20ba5eba4f763d,
title = "Service-oriented architecture based on biometric using random features and incremental neural networks",
abstract = "We propose a service-oriented architecture based on biometric system where training and classification tasks are used by millions of users via internet connection. Such a large-scale biometric system needs to consider template protection, accuracy and efficiency issues. This is a challenging problem since there are tradeoffs among these three issues. In order to simultaneously handle these issues, we extract both global and local features via controlling the sparsity of random bases without training. Subsequently, the extracted features are fused with a sequential classifier. In the proposed system, the random basis features are not stored for security reason. The non-training based on feature extraction followed by a sequential learning contributes to computational efficiency. The overall accuracy is consequently improved via an ensemble of classifiers. We evaluate the performance of the proposed system using equal error rate under a stolen-token scenario. Our experimental results show that the proposed method is robust over severe local deformation with efficient computation for simultaneous transactions. Although we focus on face biometrics in this paper, the proposed method is generic and can be applied to other biometric traits.",
author = "Kwontaeg Choi and Toh, {Kar Ann} and Youngjung Uh and Hyeran Byun",
year = "2012",
month = "9",
day = "1",
doi = "10.1007/s00500-012-0827-3",
language = "English",
volume = "16",
pages = "1539--1553",
journal = "Soft Computing",
issn = "1432-7643",
publisher = "Springer Verlag",
number = "9",

}

Service-oriented architecture based on biometric using random features and incremental neural networks. / Choi, Kwontaeg; Toh, Kar Ann; Uh, Youngjung; Byun, Hyeran.

In: Soft Computing, Vol. 16, No. 9, 01.09.2012, p. 1539-1553.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Service-oriented architecture based on biometric using random features and incremental neural networks

AU - Choi, Kwontaeg

AU - Toh, Kar Ann

AU - Uh, Youngjung

AU - Byun, Hyeran

PY - 2012/9/1

Y1 - 2012/9/1

N2 - We propose a service-oriented architecture based on biometric system where training and classification tasks are used by millions of users via internet connection. Such a large-scale biometric system needs to consider template protection, accuracy and efficiency issues. This is a challenging problem since there are tradeoffs among these three issues. In order to simultaneously handle these issues, we extract both global and local features via controlling the sparsity of random bases without training. Subsequently, the extracted features are fused with a sequential classifier. In the proposed system, the random basis features are not stored for security reason. The non-training based on feature extraction followed by a sequential learning contributes to computational efficiency. The overall accuracy is consequently improved via an ensemble of classifiers. We evaluate the performance of the proposed system using equal error rate under a stolen-token scenario. Our experimental results show that the proposed method is robust over severe local deformation with efficient computation for simultaneous transactions. Although we focus on face biometrics in this paper, the proposed method is generic and can be applied to other biometric traits.

AB - We propose a service-oriented architecture based on biometric system where training and classification tasks are used by millions of users via internet connection. Such a large-scale biometric system needs to consider template protection, accuracy and efficiency issues. This is a challenging problem since there are tradeoffs among these three issues. In order to simultaneously handle these issues, we extract both global and local features via controlling the sparsity of random bases without training. Subsequently, the extracted features are fused with a sequential classifier. In the proposed system, the random basis features are not stored for security reason. The non-training based on feature extraction followed by a sequential learning contributes to computational efficiency. The overall accuracy is consequently improved via an ensemble of classifiers. We evaluate the performance of the proposed system using equal error rate under a stolen-token scenario. Our experimental results show that the proposed method is robust over severe local deformation with efficient computation for simultaneous transactions. Although we focus on face biometrics in this paper, the proposed method is generic and can be applied to other biometric traits.

UR - http://www.scopus.com/inward/record.url?scp=84865727735&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865727735&partnerID=8YFLogxK

U2 - 10.1007/s00500-012-0827-3

DO - 10.1007/s00500-012-0827-3

M3 - Article

VL - 16

SP - 1539

EP - 1553

JO - Soft Computing

JF - Soft Computing

SN - 1432-7643

IS - 9

ER -