Semi-supervised neighborhood discrimination analysis method for face recognition

A technology of discriminant analysis and face recognition, applied in the field of image processing, which can solve the problems of insufficient labeled data, money-consuming and time-consuming, etc.

Inactive Publication Date: 2010-11-03
BEIJING JIAOTONG UNIV
View PDF0 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For example, images can be easily obtained from digital cameras, digital videos, or from the Internet, however, the labeled data is insufficient due to the costly and time-consuming nature of labeling work

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Semi-supervised neighborhood discrimination analysis method for face recognition
  • Semi-supervised neighborhood discrimination analysis method for face recognition
  • Semi-supervised neighborhood discrimination analysis method for face recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] The invention provides a semi-supervised neighborhood discriminant analysis (SSNDA) method for face recognition. Combine below Figure 1-7 and different algorithms, the present invention will be further described based on spectrogram theory.

[0022] Spectral graph theory for dimensionality reduction

[0023] In classification problems, the training sample set can be expressed as a matrix X=[x 1 , x 2 ,...,x N ], x i ∈R M , where N is the number of samples, M is the feature dimension, and x i is the jth face sample, x i ∈R M Indicates the face sample x i belongs to the M-dimensional real number space. In a supervised learning problem, a sample x i The class label of c i ∈{1,2,...,n c}, where n c is the number of classes, N j is the number of samples belonging to the jth class.

[0024] Construct an undirected weighted graph G={X, A} using the graph-based dimensionality reduction method [2,3], where the vertex set is X, and the adjacency matrix or weight ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a semi-supervised neighborhood discrimination analysis method for face recognition, belonging to the technical range of image treatment. The method can retain a local structure of data and have the discrimination capacity. The semi-supervised neighborhood discrimination analysis (SSNDA) method is designed against facial image data and comprises the steps of utilizing the spectral graph theory as a tool, fully utilizing information provided by a marked data type label and similarity information between the marked data and the unmarked data for constructing an adjacent matrix ASSNDA, and fully utilizing the information provided by the marked data type label and the similarity information between the marked data and the unmarked data. Therefore, the method constructs the adjacent matrix A-SSNDA: the low-dimensional feature of the SSNDA represents that the SSNDA contains discrimination information of the marked data and the information of the local structure of the marked and the unmarked data. Actual face recognition experiments verify the high efficiency and the stability of the SSNDA, and the performances of the method are better than those of the LDA method.

Description

technical field [0001] The invention belongs to the technical scope of image processing, in particular to a semi-supervised neighborhood discriminant analysis method for face recognition. Background technique [0002] In the field of pattern recognition and machine learning, many applications such as image classification and information retrieval have the problem of high-dimensional data and too little labeled data. Dimensionality reduction, that is, the construction of low-dimensional representations to represent the original high-dimensional data, has become a fundamental task in classification and visualization problems. In the case of unclear class label information, unsupervised dimensionality reduction methods for constructing low-dimensional representations include PCA (Principal Component Analysis) and KPCA (Kernel Principal Component Analysis). In recent years, various approaches to unsupervised dimensionality reduction have received a lot of attention, such as LE ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62
Inventor 赵嘉莉黄雅平田媚王文秀罗四维
Owner BEIJING JIAOTONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap