Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for face recognition by adopting two-dimensional non-negative sparse partial least squares

A partial least squares, non-negative sparse technology, applied in the field of two-dimensional non-negative sparse partial least squares for face recognition, it can solve the problems that the algorithm does not consider the non-negative constraints and the results are not interpretable enough.

Inactive Publication Date: 2014-08-06
CHONGQING UNIV
View PDF3 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But these algorithms do not consider the non-negativity constraint
Therefore, the interpretability of the results is insufficient

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
  • Method for face recognition by adopting two-dimensional non-negative sparse partial least squares
  • Method for face recognition by adopting two-dimensional non-negative sparse partial least squares
  • Method for face recognition by adopting two-dimensional non-negative sparse partial least squares

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0095] Example 1: for figure 1 (a) is performed on 11 original images of a person without occlusion, and the highest recognition rate results are as follows figure 2shown. It can be seen from this that the method of 2DNSPLS [the method of the present invention, that is, when the face training sample set X is projected on the base matrix, the non-negativity constraint and the sparsity constraint are considered] and the method of 2DNPLS [two-dimensional non-negative partial least squares Multiplication (Two-dimensional Nonnegative PLS, 2DNPLS, that is, when the face training sample set X is projected on the base matrix, only the non-negativity constraint is considered, and the converged base matrix is ​​finally obtained through the iterative format (5) iteration] when there are fewer training samples The results are better than other two-dimensional algorithms. The 2DLDA and 2DLPP algorithms perform poorly when the training samples are small. In general, 2DNSPLS and 2DNPLS are...

Embodiment 2

[0096] Embodiment 2: This embodiment is in figure 1 (b), figure 1 (c), figure 1 (d), figure 1 (e) and figure 1 (f) In the face database corresponding to the five cases, 2DPCA, 2DLDA, 2DLPP, and 2DPLS respectively represent two-dimensional principal component analysis, two-dimensional discriminant analysis, two-dimensional local preservation mapping, two-dimensional partial least squares 2DNPLS, 2DNSPLS algorithm for testing. Experimental results such as Figure 3-7 shown.

[0097] Such as image 3 shown, at 9% occlusion of the figure 1 In (b), 2DLDA performs the worst when there are fewer training samples, followed by 2DLPP. 2DNSPLS algorithm shows better recognition rate and robustness under each training sample. The recognition rate of 2DNPLS algorithm decreases with the occurrence of occlusion, and other algorithms are greatly affected. Both 2DLDA and 2DLPP are greatly affected when the sample size is small.

[0098] Compare image 3 , 4 , 5, that is, in the fa...

Embodiment 3

[0100] Embodiment 3: Adopt the experiment on PIE face bank [face bank is the face bank that includes posture (pose), illumination (illumination) and expression (exprsssion) change that Carnegie Mellon University establishs] it includes from 68 people The 40,000 photos of each person have photos under 13 pose conditions, 43 lighting conditions and 4 expressions. In this embodiment, 45 photos of each person are selected as the total sample of the experiment. Figure 8 There are 45 pictures of 2 of them.

[0101] Select the first 5, 10, 15, 20, 25, 30, 35, and 40 pictures of each person for training, and the remaining pictures for testing. Experimental results such as Figure 9 shown.

[0102] Depend on Figure 9 It can be seen that the recognition rate of 2DNSPLS is significantly better than other algorithms when the number of training samples is small, and the advantage decreases with the increase of the number of training samples, but it is still better than other algorith...

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 relates to a method for face recognition by adopting two-dimensional non-negative sparse partial least squares. The method comprises the following steps: firstly constructing a category matrix of a face training sample set, then constructing an objective function for realizing minimum loss of the amount of information in the projection of the face training sample set and the category matrix, afterwards adding non-negativity constraint and sparsity constraint into the objective function to obtain a convergent non-negative basis matrix, projecting the face training sample set on the basis matrix to obtain a test sample coefficient matrix, also carrying out the operation on a test sample to obtain a coefficient matrix of the test sample, determining that certain element matrices in the coefficient matrix of the test sample and the test sample coefficient matrix fall into the same category by using a nearest neighbor strategy, and then considering that a person on the test sample corresponding to the coefficient matrix of the test sample is the same person on the training sample corresponding to the element matrix. The method is high in recognition rate and robustness, and as only the basis matrix needs iterative solution, the operation is simplified, the time complexity is reduced, and the recognition speed is high.

Description

technical field [0001] The invention belongs to the technical field of image processing and pattern recognition, and in particular relates to a face recognition method using two-dimensional non-negative sparse partial least squares. Background technique [0002] With the advancement of science and technology, face recognition, as an important biometric identification technology, is receiving more and more attention and research. The partial least squares (Partial Least Squares, PLS) proposed by Wold H can solve the problem of multiple correlations between variables, and is suitable for application when the sample size is smaller than the number of variables. In recent years, PLS and its improved algorithm have been widely used in face recognition. J. Beak. etc. used PLS for face recognition for the first time, and got a good recognition rate. Sun Quansen et al. extended PLS to a two-dimensional algorithm and proposed two-dimensional partial least squares (Two-dimensional P...

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
IPC IPC(8): G06K9/64
Inventor 葛永新徐玲杨梦宁洪明坚张小洪杨丹步文斌
Owner CHONGQING UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com