Weighted neighbourhood preserving embedding in face recognition

Andrew Beng Jin Teoh; Pang Ying Han; Lim Heng Siong

doi:10.1109/ICIEA.2010.5516836

Weighted neighbourhood preserving embedding in face recognition

Andrew Beng Jin Teoh, Pang Ying Han, Lim Heng Siong

Department of Electrical and Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Graph Embedding (GE) along with its linearization outperforms the traditional linear dimension reduction techniques in face recognition, but there is still room for improvement on GE. This paper proposes an eigenvector weighting technique for a realization of linear GE, namely Neighbourhood Preserving Embedding (NPE) in face verification. The proposed method is called Eigenvector Weighting Function - NPE (EWF-NPE). The eigenspace is decomposed into three subspaces: (1) a subspace that is attributed to facial intra-class variations, (2) a subspace comprises of intrinsic facial characteristics, and (3) a subspace that is attributed to sensor and other external noises. Eigenfeatures are weighted differently in these subspaces. The proposed EWF-NPE ensures that only stable face subspace which yields informative data is emphasized, while the other two noise subspaces are deemphasized. Experimental investigations on FRGC and FERET databases demonstrate promising results of the proposed method.

Original language	English
Title of host publication	Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010
Pages	295-300
Number of pages	6
DOIs	https://doi.org/10.1109/ICIEA.2010.5516836
Publication status	Published - 2010
Event	5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010 - Taichung, Taiwan, Province of China Duration: 2010 Jun 15 → 2010 Jun 17

Publication series

Name	Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010

Other

Other	5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010
Country	Taiwan, Province of China
City	Taichung
Period	10/6/15 → 10/6/17

All Science Journal Classification (ASJC) codes

Electrical and Electronic Engineering

Access to Document

10.1109/ICIEA.2010.5516836

Fingerprint Dive into the research topics of 'Weighted neighbourhood preserving embedding in face recognition'. Together they form a unique fingerprint.

Cite this

Teoh, A. B. J., Han, P. Y., & Siong, L. H. (2010). Weighted neighbourhood preserving embedding in face recognition. In Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010 (pp. 295-300). [5516836] (Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010). https://doi.org/10.1109/ICIEA.2010.5516836

@inproceedings{2b8455c5b6754b9a86a01443f6a1824b,

title = "Weighted neighbourhood preserving embedding in face recognition",

abstract = "Graph Embedding (GE) along with its linearization outperforms the traditional linear dimension reduction techniques in face recognition, but there is still room for improvement on GE. This paper proposes an eigenvector weighting technique for a realization of linear GE, namely Neighbourhood Preserving Embedding (NPE) in face verification. The proposed method is called Eigenvector Weighting Function - NPE (EWF-NPE). The eigenspace is decomposed into three subspaces: (1) a subspace that is attributed to facial intra-class variations, (2) a subspace comprises of intrinsic facial characteristics, and (3) a subspace that is attributed to sensor and other external noises. Eigenfeatures are weighted differently in these subspaces. The proposed EWF-NPE ensures that only stable face subspace which yields informative data is emphasized, while the other two noise subspaces are deemphasized. Experimental investigations on FRGC and FERET databases demonstrate promising results of the proposed method.",

author = "Teoh, {Andrew Beng Jin} and Han, {Pang Ying} and Siong, {Lim Heng}",

year = "2010",

doi = "10.1109/ICIEA.2010.5516836",

language = "English",

isbn = "9781424450466",

series = "Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010",

pages = "295--300",

booktitle = "Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010",

}

Teoh, ABJ, Han, PY & Siong, LH 2010, Weighted neighbourhood preserving embedding in face recognition. in Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010., 5516836, Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010, pp. 295-300, 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010, Taichung, Taiwan, Province of China, 10/6/15. https://doi.org/10.1109/ICIEA.2010.5516836

Weighted neighbourhood preserving embedding in face recognition. / Teoh, Andrew Beng Jin; Han, Pang Ying; Siong, Lim Heng.

Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010. 2010. p. 295-300 5516836 (Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Weighted neighbourhood preserving embedding in face recognition

AU - Teoh, Andrew Beng Jin

AU - Han, Pang Ying

AU - Siong, Lim Heng

PY - 2010

Y1 - 2010

N2 - Graph Embedding (GE) along with its linearization outperforms the traditional linear dimension reduction techniques in face recognition, but there is still room for improvement on GE. This paper proposes an eigenvector weighting technique for a realization of linear GE, namely Neighbourhood Preserving Embedding (NPE) in face verification. The proposed method is called Eigenvector Weighting Function - NPE (EWF-NPE). The eigenspace is decomposed into three subspaces: (1) a subspace that is attributed to facial intra-class variations, (2) a subspace comprises of intrinsic facial characteristics, and (3) a subspace that is attributed to sensor and other external noises. Eigenfeatures are weighted differently in these subspaces. The proposed EWF-NPE ensures that only stable face subspace which yields informative data is emphasized, while the other two noise subspaces are deemphasized. Experimental investigations on FRGC and FERET databases demonstrate promising results of the proposed method.

AB - Graph Embedding (GE) along with its linearization outperforms the traditional linear dimension reduction techniques in face recognition, but there is still room for improvement on GE. This paper proposes an eigenvector weighting technique for a realization of linear GE, namely Neighbourhood Preserving Embedding (NPE) in face verification. The proposed method is called Eigenvector Weighting Function - NPE (EWF-NPE). The eigenspace is decomposed into three subspaces: (1) a subspace that is attributed to facial intra-class variations, (2) a subspace comprises of intrinsic facial characteristics, and (3) a subspace that is attributed to sensor and other external noises. Eigenfeatures are weighted differently in these subspaces. The proposed EWF-NPE ensures that only stable face subspace which yields informative data is emphasized, while the other two noise subspaces are deemphasized. Experimental investigations on FRGC and FERET databases demonstrate promising results of the proposed method.

UR - http://www.scopus.com/inward/record.url?scp=77956050018&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77956050018&partnerID=8YFLogxK

U2 - 10.1109/ICIEA.2010.5516836

DO - 10.1109/ICIEA.2010.5516836

M3 - Conference contribution

AN - SCOPUS:77956050018

SN - 9781424450466

T3 - Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010

SP - 295

EP - 300

BT - Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010

T2 - 5th IEEE Conference on Industrial Electronics and Applications, ICIEA 2010

Y2 - 15 June 2010 through 17 June 2010

ER -