Unsupervised discriminant projection analysis for feature extraction

Jian Yang; Dapeng Zhang; Zhong Jin; Jing Yu Yang

doi:10.1109/ICPR.2006.1143

Unsupervised discriminant projection analysis for feature extraction

Jian Yang, Dapeng Zhang, Zhong Jin, Jing Yu Yang

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

22 Citations (Scopus)

Abstract

This paper develops an unsupervised discriminant projection (UDP) technique for feature extraction. UDP takes the local and non-local information into account, seeking to find a projection that maximizes the non-local scatter and minimizes the local scatter simultaneously. This characteristic makes UDP more intuitive and more powerful than the up-to-date method - Locality preserving projection (LPP, which considers the local information only) for classification tasks. The proposed method is applied to face biometrics and examined using the ORL and FERET face image databases. Our experimental results show that UDP consistently outperforms LPP, PCA, and LDA.

Original language	English
Title of host publication	Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006
Pages	904-907
Number of pages	4
Volume	1
DOIs	https://doi.org/10.1109/ICPR.2006.1143
Publication status	Published - 1 Dec 2006
Event	18th International Conference on Pattern Recognition, ICPR 2006 - Hong Kong, Hong Kong Duration: 20 Aug 2006 → 24 Aug 2006

Conference

Conference	18th International Conference on Pattern Recognition, ICPR 2006
Country/Territory	Hong Kong
City	Hong Kong
Period	20/08/06 → 24/08/06

ASJC Scopus subject areas

Computer Vision and Pattern Recognition

More information

10.1109/ICPR.2006.1143

http://hdl.handle.net/10397/1187

Cite this

@inproceedings{12b95f6467104b9e8013ee3e595a7a16,

title = "Unsupervised discriminant projection analysis for feature extraction",

abstract = "This paper develops an unsupervised discriminant projection (UDP) technique for feature extraction. UDP takes the local and non-local information into account, seeking to find a projection that maximizes the non-local scatter and minimizes the local scatter simultaneously. This characteristic makes UDP more intuitive and more powerful than the up-to-date method - Locality preserving projection (LPP, which considers the local information only) for classification tasks. The proposed method is applied to face biometrics and examined using the ORL and FERET face image databases. Our experimental results show that UDP consistently outperforms LPP, PCA, and LDA.",

author = "Jian Yang and Dapeng Zhang and Zhong Jin and Yang, {Jing Yu}",

year = "2006",

month = dec,

day = "1",

doi = "10.1109/ICPR.2006.1143",

language = "English",

isbn = "0769525210",

volume = "1",

pages = "904--907",

booktitle = "Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006",

note = "18th International Conference on Pattern Recognition, ICPR 2006 ; Conference date: 20-08-2006 Through 24-08-2006",

}

Unsupervised discriminant projection analysis for feature extraction. / Yang, Jian; Zhang, Dapeng; Jin, Zhong et al.
Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006. Vol. 1 2006. p. 904-907 1699036.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - Unsupervised discriminant projection analysis for feature extraction

AU - Yang, Jian

AU - Zhang, Dapeng

AU - Jin, Zhong

AU - Yang, Jing Yu

PY - 2006/12/1

Y1 - 2006/12/1

N2 - This paper develops an unsupervised discriminant projection (UDP) technique for feature extraction. UDP takes the local and non-local information into account, seeking to find a projection that maximizes the non-local scatter and minimizes the local scatter simultaneously. This characteristic makes UDP more intuitive and more powerful than the up-to-date method - Locality preserving projection (LPP, which considers the local information only) for classification tasks. The proposed method is applied to face biometrics and examined using the ORL and FERET face image databases. Our experimental results show that UDP consistently outperforms LPP, PCA, and LDA.

AB - This paper develops an unsupervised discriminant projection (UDP) technique for feature extraction. UDP takes the local and non-local information into account, seeking to find a projection that maximizes the non-local scatter and minimizes the local scatter simultaneously. This characteristic makes UDP more intuitive and more powerful than the up-to-date method - Locality preserving projection (LPP, which considers the local information only) for classification tasks. The proposed method is applied to face biometrics and examined using the ORL and FERET face image databases. Our experimental results show that UDP consistently outperforms LPP, PCA, and LDA.

UR - http://www.scopus.com/inward/record.url?scp=34047237342&partnerID=8YFLogxK

U2 - 10.1109/ICPR.2006.1143

DO - 10.1109/ICPR.2006.1143

M3 - Conference article published in proceeding or book

SN - 0769525210

SN - 9780769525211

VL - 1

SP - 904

EP - 907

BT - Proceedings - 18th International Conference on Pattern Recognition, ICPR 2006

T2 - 18th International Conference on Pattern Recognition, ICPR 2006

Y2 - 20 August 2006 through 24 August 2006

ER -

Unsupervised discriminant projection analysis for feature extraction

Abstract

Conference

ASJC Scopus subject areas

More information

Other files and links

Fingerprint

Cite this