Face localization via hierarchical CONDENSATION with Fisher Boosting feature selection

Jilin Tu, Zhenqiu Zhang, Zhihong Zeng, Thomas Huang

Research output: Contribution to journalConference article

Abstract

We formulate face localization as a Maximum A Posteriori Probability(MAP) problem of finding the best estimation of human face configuration in a given image. The a prior distribution for intrinsic face configuration is defined by Active Shape Model(ASM). The likelihood model for local facial features is parameterized as Mixture of Gaussions in feature space. A hierarchical CONDENSATION framework is then proposed to estimate the face configuration parameter. In order to improve the discriminative power of likelihood distribution in feature space, a new feature subspace, Fisher Boosting feature space, is proposed and compared against PCA subspace and biased PCA subspace. Experiments show that, Fisher Boosting algorithm can generate strong classifier with less number of weaker classifiers comparing to conventional Adaboosting algorithm as illustrated in a toy problem, that the face localization with Fisher Boosting feature subspace outperforms that with PCA feature subspaces in localization accuracy and convergence rate, and that the design of hierarchical CONDENSATION framework alleviates the local minima problem which is frequently encountered by previous ASM optimization algorithms.

Original languageEnglish (US)
Pages (from-to)II719-II724
JournalProceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Volume2
StatePublished - Oct 19 2004
EventProceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2004 - Washington, DC, United States
Duration: Jun 27 2004Jul 2 2004

Fingerprint

Feature extraction
Classifiers
Experiments

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition

Cite this

Face localization via hierarchical CONDENSATION with Fisher Boosting feature selection. / Tu, Jilin; Zhang, Zhenqiu; Zeng, Zhihong; Huang, Thomas.

In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Vol. 2, 19.10.2004, p. II719-II724.

Research output: Contribution to journalConference article

@article{9afda0ab167449afbf64f1be929b4a2e,
title = "Face localization via hierarchical CONDENSATION with Fisher Boosting feature selection",
abstract = "We formulate face localization as a Maximum A Posteriori Probability(MAP) problem of finding the best estimation of human face configuration in a given image. The a prior distribution for intrinsic face configuration is defined by Active Shape Model(ASM). The likelihood model for local facial features is parameterized as Mixture of Gaussions in feature space. A hierarchical CONDENSATION framework is then proposed to estimate the face configuration parameter. In order to improve the discriminative power of likelihood distribution in feature space, a new feature subspace, Fisher Boosting feature space, is proposed and compared against PCA subspace and biased PCA subspace. Experiments show that, Fisher Boosting algorithm can generate strong classifier with less number of weaker classifiers comparing to conventional Adaboosting algorithm as illustrated in a toy problem, that the face localization with Fisher Boosting feature subspace outperforms that with PCA feature subspaces in localization accuracy and convergence rate, and that the design of hierarchical CONDENSATION framework alleviates the local minima problem which is frequently encountered by previous ASM optimization algorithms.",
author = "Jilin Tu and Zhenqiu Zhang and Zhihong Zeng and Thomas Huang",
year = "2004",
month = "10",
day = "19",
language = "English (US)",
volume = "2",
pages = "II719--II724",
journal = "Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition",
issn = "1063-6919",
publisher = "IEEE Computer Society",

}

TY - JOUR

T1 - Face localization via hierarchical CONDENSATION with Fisher Boosting feature selection

AU - Tu, Jilin

AU - Zhang, Zhenqiu

AU - Zeng, Zhihong

AU - Huang, Thomas

PY - 2004/10/19

Y1 - 2004/10/19

N2 - We formulate face localization as a Maximum A Posteriori Probability(MAP) problem of finding the best estimation of human face configuration in a given image. The a prior distribution for intrinsic face configuration is defined by Active Shape Model(ASM). The likelihood model for local facial features is parameterized as Mixture of Gaussions in feature space. A hierarchical CONDENSATION framework is then proposed to estimate the face configuration parameter. In order to improve the discriminative power of likelihood distribution in feature space, a new feature subspace, Fisher Boosting feature space, is proposed and compared against PCA subspace and biased PCA subspace. Experiments show that, Fisher Boosting algorithm can generate strong classifier with less number of weaker classifiers comparing to conventional Adaboosting algorithm as illustrated in a toy problem, that the face localization with Fisher Boosting feature subspace outperforms that with PCA feature subspaces in localization accuracy and convergence rate, and that the design of hierarchical CONDENSATION framework alleviates the local minima problem which is frequently encountered by previous ASM optimization algorithms.

AB - We formulate face localization as a Maximum A Posteriori Probability(MAP) problem of finding the best estimation of human face configuration in a given image. The a prior distribution for intrinsic face configuration is defined by Active Shape Model(ASM). The likelihood model for local facial features is parameterized as Mixture of Gaussions in feature space. A hierarchical CONDENSATION framework is then proposed to estimate the face configuration parameter. In order to improve the discriminative power of likelihood distribution in feature space, a new feature subspace, Fisher Boosting feature space, is proposed and compared against PCA subspace and biased PCA subspace. Experiments show that, Fisher Boosting algorithm can generate strong classifier with less number of weaker classifiers comparing to conventional Adaboosting algorithm as illustrated in a toy problem, that the face localization with Fisher Boosting feature subspace outperforms that with PCA feature subspaces in localization accuracy and convergence rate, and that the design of hierarchical CONDENSATION framework alleviates the local minima problem which is frequently encountered by previous ASM optimization algorithms.

UR - http://www.scopus.com/inward/record.url?scp=5044236978&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=5044236978&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:5044236978

VL - 2

SP - II719-II724

JO - Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition

JF - Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition

SN - 1063-6919

ER -