Fusion of visual and infra-red face scores by weighted power series

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

This paper proposes a weighted power series model for face verification scores fusion. Essentially, a linear parametric power series model is adopted to directly minimize an approximated total error rate for fusion of multi-modal face verification scores. Unlike the conventional least-squares error minimization approach which involves fitting of a learning model to data density and then perform a threshold process for error counting, this work directly formulates the required target error count rate in terms of design model parameters with a closed-form solution. The solution is found to belong to a specific setting of the weighted least squares. Our experiments on fusing scores from visual and infra-red face images as well as on public data sets show promising results.

Original languageEnglish
Pages (from-to)603-615
Number of pages13
JournalPattern Recognition Letters
Volume29
Issue number5
DOIs
Publication statusPublished - 2008 Apr 1

Fingerprint

Fusion reactions
Infrared radiation
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

@article{b4a39b5216eb4ea9b0c3ff76abc33c03,
title = "Fusion of visual and infra-red face scores by weighted power series",
abstract = "This paper proposes a weighted power series model for face verification scores fusion. Essentially, a linear parametric power series model is adopted to directly minimize an approximated total error rate for fusion of multi-modal face verification scores. Unlike the conventional least-squares error minimization approach which involves fitting of a learning model to data density and then perform a threshold process for error counting, this work directly formulates the required target error count rate in terms of design model parameters with a closed-form solution. The solution is found to belong to a specific setting of the weighted least squares. Our experiments on fusing scores from visual and infra-red face images as well as on public data sets show promising results.",
author = "Toh, {Kar Ann} and Youngsung Kim and Sangyoun Lee and Jaihie Kim",
year = "2008",
month = "4",
day = "1",
doi = "10.1016/j.patrec.2007.11.010",
language = "English",
volume = "29",
pages = "603--615",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",
number = "5",

}

Fusion of visual and infra-red face scores by weighted power series. / Toh, Kar Ann; Kim, Youngsung; Lee, Sangyoun; Kim, Jaihie.

In: Pattern Recognition Letters, Vol. 29, No. 5, 01.04.2008, p. 603-615.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Fusion of visual and infra-red face scores by weighted power series

AU - Toh, Kar Ann

AU - Kim, Youngsung

AU - Lee, Sangyoun

AU - Kim, Jaihie

PY - 2008/4/1

Y1 - 2008/4/1

N2 - This paper proposes a weighted power series model for face verification scores fusion. Essentially, a linear parametric power series model is adopted to directly minimize an approximated total error rate for fusion of multi-modal face verification scores. Unlike the conventional least-squares error minimization approach which involves fitting of a learning model to data density and then perform a threshold process for error counting, this work directly formulates the required target error count rate in terms of design model parameters with a closed-form solution. The solution is found to belong to a specific setting of the weighted least squares. Our experiments on fusing scores from visual and infra-red face images as well as on public data sets show promising results.

AB - This paper proposes a weighted power series model for face verification scores fusion. Essentially, a linear parametric power series model is adopted to directly minimize an approximated total error rate for fusion of multi-modal face verification scores. Unlike the conventional least-squares error minimization approach which involves fitting of a learning model to data density and then perform a threshold process for error counting, this work directly formulates the required target error count rate in terms of design model parameters with a closed-form solution. The solution is found to belong to a specific setting of the weighted least squares. Our experiments on fusing scores from visual and infra-red face images as well as on public data sets show promising results.

UR - http://www.scopus.com/inward/record.url?scp=38749124716&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38749124716&partnerID=8YFLogxK

U2 - 10.1016/j.patrec.2007.11.010

DO - 10.1016/j.patrec.2007.11.010

M3 - Article

AN - SCOPUS:38749124716

VL - 29

SP - 603

EP - 615

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

IS - 5

ER -