Incremental Nonnegative Matrix Factorization for Face Recognition
Nonnegative matrix factorization (NMF) is a promising approach for local feature extraction in face recognition tasks. However, there are two major drawbacks in almost all existing NMF-based methods. One shortcoming is that the computational cost is expensive for large matrix decomposition. The othe...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Hindawi Limited
2008-01-01
|
Series: | Mathematical Problems in Engineering |
Online Access: | http://dx.doi.org/10.1155/2008/410674 |
id |
doaj-0b1565d3a89542bf89c76ef5427de9f1 |
---|---|
record_format |
Article |
spelling |
doaj-0b1565d3a89542bf89c76ef5427de9f12020-11-25T00:58:22ZengHindawi LimitedMathematical Problems in Engineering1024-123X1563-51472008-01-01200810.1155/2008/410674410674Incremental Nonnegative Matrix Factorization for Face RecognitionWen-Sheng Chen0Binbin Pan1Bin Fang2Ming Li3Jianliang Tang4College of Mathematics and Computational Science, Shenzhen University, Shenzhen 518060, ChinaCollege of Mathematics and Computational Science, Shenzhen University, Shenzhen 518060, ChinaCollege of Computer Science, Chongqing University, Chongqing 400044, ChinaSchool of Information Science & Technology, East China Normal University, Shanghai 200241, ChinaCollege of Mathematics and Computational Science, Shenzhen University, Shenzhen 518060, ChinaNonnegative matrix factorization (NMF) is a promising approach for local feature extraction in face recognition tasks. However, there are two major drawbacks in almost all existing NMF-based methods. One shortcoming is that the computational cost is expensive for large matrix decomposition. The other is that it must conduct repetitive learning, when the training samples or classes are updated. To overcome these two limitations, this paper proposes a novel incremental nonnegative matrix factorization (INMF) for face representation and recognition. The proposed INMF approach is based on a novel constraint criterion and our previous block strategy. It thus has some good properties, such as low computational complexity, sparse coefficient matrix. Also, the coefficient column vectors between different classes are orthogonal. In particular, it can be applied to incremental learning. Two face databases, namely FERET and CMU PIE face databases, are selected for evaluation. Compared with PCA and some state-of-the-art NMF-based methods, our INMF approach gives the best performance.http://dx.doi.org/10.1155/2008/410674 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Wen-Sheng Chen Binbin Pan Bin Fang Ming Li Jianliang Tang |
spellingShingle |
Wen-Sheng Chen Binbin Pan Bin Fang Ming Li Jianliang Tang Incremental Nonnegative Matrix Factorization for Face Recognition Mathematical Problems in Engineering |
author_facet |
Wen-Sheng Chen Binbin Pan Bin Fang Ming Li Jianliang Tang |
author_sort |
Wen-Sheng Chen |
title |
Incremental Nonnegative Matrix Factorization for Face Recognition |
title_short |
Incremental Nonnegative Matrix Factorization for Face Recognition |
title_full |
Incremental Nonnegative Matrix Factorization for Face Recognition |
title_fullStr |
Incremental Nonnegative Matrix Factorization for Face Recognition |
title_full_unstemmed |
Incremental Nonnegative Matrix Factorization for Face Recognition |
title_sort |
incremental nonnegative matrix factorization for face recognition |
publisher |
Hindawi Limited |
series |
Mathematical Problems in Engineering |
issn |
1024-123X 1563-5147 |
publishDate |
2008-01-01 |
description |
Nonnegative matrix factorization (NMF) is a promising approach for local feature extraction in face recognition tasks. However, there are two major drawbacks in almost all existing NMF-based methods. One shortcoming is that the computational cost is expensive for large matrix decomposition. The other is that it must conduct repetitive learning, when the training samples or classes are updated. To overcome these two limitations, this paper proposes a novel incremental nonnegative matrix factorization (INMF) for face representation and recognition. The proposed INMF approach is based on a novel constraint criterion and our previous block strategy. It thus has some good properties, such as low computational complexity, sparse coefficient matrix. Also, the coefficient column vectors between different classes are orthogonal. In particular, it can be applied to incremental learning. Two face databases, namely FERET and CMU PIE face databases, are selected for evaluation. Compared with PCA and some state-of-the-art NMF-based methods, our INMF approach gives the best performance. |
url |
http://dx.doi.org/10.1155/2008/410674 |
work_keys_str_mv |
AT wenshengchen incrementalnonnegativematrixfactorizationforfacerecognition AT binbinpan incrementalnonnegativematrixfactorizationforfacerecognition AT binfang incrementalnonnegativematrixfactorizationforfacerecognition AT mingli incrementalnonnegativematrixfactorizationforfacerecognition AT jianliangtang incrementalnonnegativematrixfactorizationforfacerecognition |
_version_ |
1725220375650893824 |