Learning spatially localized, parts-based representation

S.Z. Li; Xin Hou; Hong Jiang Zhang; Qian Cheng

doi:10.1109/cvpr.2001.990477

Abstract

In this paper, we propose a novel method, called local non-negative matrix factorization (LNMF), for learning spatially localized, parts-based subspace representation of visual patterns. An objective function is defined to impose a localization constraint, in addition to the non-negativity constraint in the standard NMF. This gives a set of bases which not only allows a non-subtractive (part-based) representation of images but also manifests localized features. An algorithm is presented for the learning of such basic components. Experimental results are presented to compare LNMF with the NMF and PCA methods for face representation and recognition, which demonstrates advantages of LNMF.

Keywords

Non-negative matrix factorizationRepresentation (politics)Computer scienceConstraint (computer-aided design)Artificial intelligenceSubspace topologyPattern recognition (psychology)Matrix decompositionSet (abstract data type)Facial recognition systemFactorizationFace (sociological concept)AlgorithmMathematics

Affiliated Institutions

Related Publications

Effective representation using ICA for face recognition robust to local distortion and partial occlusion

Jongsun Kim , Jongmoo Choi , Juneho Yi +1 more

The performance of face recognition methods using subspace projection is directly related to the characteristics of their basis images, especially in the cases of local distorti...

2005 IEEE Transactions on Pattern Analysis... 215 citations

Face recognition using Laplacianfaces

Xiaofei He , Shuicheng Yan , Yuxiao Hu +2 more

We propose an appearance-based face recognition method called the Laplacianface approach. By using Locality Preserving Projections (LPP), the face images are mapped into a face ...

2005 IEEE Transactions on Pattern Analysis... 3273 citations

Feature Extraction and Uncorrelated Discriminant Analysis for High-Dimensional Data

Wenhui Yang , Dao‐Qing Dai , Hong Yan

High-dimensional data and the small sample size problem occur in many modern pattern classification applications such as face recognition and gene expression data analysis. To d...

2008 IEEE Transactions on Knowledge and Da... 56 citations

Human Face Detection in Visual Scenes

Henry A. Rowley , Shumeet Baluja , Takeo Kanade

We present a neural network-based face detection system. A retinally connected neural network examines small windows of an image, and decides whether each window contains a face...

1995 354 citations

Dropout Training of Matrix Factorization and Autoencoder for Link Prediction in Sparse Graphs

Shuangfei Zhai , Zhongfei Zhang

Matrix factorization (MF) and Autoencoder (AE) are among the most successful approaches of unsupervised learning. While MF based models have been extensively exploited in the gr...

2015 33 citations

Publication Info

Year: 2005
Type: article
Volume: 1
Pages: I-207
Citations: 780
Access: Closed

External Links

View on DOI.org

Social Impact

Altmetric

Learning spatially localized, parts-based representation

PlumX Metrics

Social media, news, blog, policy document mentions

Citation Metrics

780

OpenAlex

Cite This

APA Style

                            
                                    S.Z. Li, 
                                
                                    Xin Hou, 
                                
                                    Hong Jiang Zhang
                                
                                et al.
                            
                            (2005). 
                            Learning spatially localized, parts-based representation. 
                            
                            , 1
                            
                            , I-207.
                            https://doi.org/10.1109/cvpr.2001.990477

Identifiers

DOI: 10.1109/cvpr.2001.990477