CN111611963B - Face recognition method based on neighbor preservation canonical correlation analysis - Google Patents
Face recognition method based on neighbor preservation canonical correlation analysis Download PDFInfo
- Publication number
- CN111611963B CN111611963B CN202010473892.2A CN202010473892A CN111611963B CN 111611963 B CN111611963 B CN 111611963B CN 202010473892 A CN202010473892 A CN 202010473892A CN 111611963 B CN111611963 B CN 111611963B
- Authority
- CN
- China
- Prior art keywords
- face
- neighbor
- correlation analysis
- matrix
- preservation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000010219 correlation analysis Methods 0.000 title claims abstract description 35
- 238000004321 preservation Methods 0.000 title claims abstract description 16
- 239000011159 matrix material Substances 0.000 claims abstract description 34
- 230000004927 fusion Effects 0.000 claims abstract description 23
- 238000005457 optimization Methods 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000012360 testing method Methods 0.000 claims abstract description 10
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 4
- 239000013598 vector Substances 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000012562 intraclass correlation Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 description 12
- 238000005516 engineering process Methods 0.000 description 9
- 238000004458 analytical method Methods 0.000 description 8
- 238000002474 experimental method Methods 0.000 description 5
- 238000011160 research Methods 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000001815 facial effect Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003909 pattern recognition Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
- 210000000697 sensory organ Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
Abstract
The invention discloses a face recognition method based on neighbor preservation typical correlation analysis, which comprises the following steps: 1: input face training dataset X epsilon R m×N ，Y∈R n×N Neighbor weight reconstruction matrix U for computing images by neighbor preserving learning x And U y The method comprises the steps of carrying out a first treatment on the surface of the 2: finding two sets of projection vectors w using a canonical correlation analysis x And w y Introducing neighbor preservation into a frame of typical correlation analysis by using an optimization method, and calculating a projection matrix W by using generalized eigenvalue decomposition x And W is y The method comprises the steps of carrying out a first treatment on the surface of the 3: low-dimensional projection of test face images by adopting two feature fusion strategiesAndfusing; 4: and using the nearest neighbor classifier to use the fused features for face recognition. According to the invention, the neighbor is introduced into a typical correlation analysis frame by using an optimization method through the neighbor preservation learning face neighbor weight reconstruction matrix, and then the label information of the face is utilized, so that the extracted face features not only maximize the correlation among different faces, but also maintain the neighbor structure of the face to the greatest extent, and improve the face recognition capability and stability.
Description
Technical Field
The invention relates to the field of classification recognition in machine learning, in particular to a face recognition method based on neighbor preservation typical correlation analysis.
Background
With the rapid development of modern information technology, the technology of identity authentication is transferred to a biological characteristic layer. Modern biological identification technology is mainly to carry out personal identification by closely combining a computer with a high-tech means and utilizing the inherent physiological characteristics and behavior characteristics of human bodies. Face recognition refers to the distribution of facial five sense organs and outlines of people, and the distribution characteristics are different from person to person and are natural. Face recognition is a technique based on facial feature information of a person. The method comprises the steps of collecting images or video streams containing faces by using a camera or a camera, automatically detecting and tracking the faces in the images, and then determining one or more persons in a scene by using an existing face database. The current face recognition research scope mainly comprises several aspects: face detection and positioning, face feature representation, face recognition, expression and gesture analysis, and physiological analysis and classification. The research methods of face recognition at present comprise a geometric feature-based method, a local feature analysis method, a feature face method and a latest neural network method. Future research on face recognition from a three-dimensional perspective is also one of the trends.
The research of the face recognition system starts in the 60 th century of the 20 th century, and the 80 th year later is improved along with the development of computer technology and optical imaging technology, while the real application stage entering the primary stage is in the later 90 th year, and is mainly realized by the technology of the United states, germany and Japan. The key to success of the face recognition system is whether to have a sophisticated core algorithm, and the recognition result has practical recognition rate and recognition speed. The face recognition system integrates a plurality of professional technologies such as artificial intelligence, pattern recognition, machine learning, model theory, expert system, video image processing and the like, and meanwhile, the theory and the realization of intermediate value processing are combined, so that the face recognition system is the latest application of biological feature recognition, and the realization of core technology shows the conversion from weak artificial intelligence to strong artificial intelligence.
Unlike the conventional way of fusing multiple sets of features in parallel to form a high-dimensional feature, typical correlation analysis (Canonical Correlation Analysis, CCA) achieves the goal of feature fusion by maximizing the correlation of two sets of features after projection, and has been successfully applied to the field of pattern recognition, such as face recognition and expression recognition, image processing, digital word recognition, and the like. However, as with PCA, the typical correlation analysis can only extract the linear features of the pattern, and as an unsupervised learning method, the classification effect is often poor, and many researchers have proposed various improved methods based on the typical correlation analysis, such as a kernel-based typical correlation analysis (KCCA) combining the typical correlation analysis with the kernel method to extract the nonlinear features of the pattern; sparse Canonical Correlation Analysis (SCCA) can preserve sparse reconstruction relationships of samples. In addition, researchers introduce supervision and Semi-supervision technologies into typical correlation analysis, and a distinguishing typical correlation analysis (DCCA) and Semi-supervised typical correlation analysis (Semi-CCA) are provided, and on the basis, students introduce paired constraint information among samples, and a supervision typical correlation analysis algorithm using relative strength is provided.
In recent years, manifold learning has been applied in the field of face recognition by its characteristics such as nonlinearity, ability to reveal the internal structure of a data set and to maintain local geometric characteristics of data. The main manifold learning algorithms include equi-scale mapping, local linear embedding, laplace feature mapping, and local tangential spatial arrangement. Because the implicit mapping established between the high-dimensional observation space and the internal low-dimensional space by the manifold learning algorithm is defined on the training data set, new observation data points cannot be directly mapped, and therefore the manifold learning algorithm cannot be directly used in face recognition, two new dimension reduction methods are provided for the problem, namely local maintenance projection and neighbor maintenance embedding. The traditional CCA face recognition method is an unsupervised method, label information of a good face is not utilized, and sometimes the face image data volume is too large, so that the recognition speed and the recognition effect are affected.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a face recognition method based on neighbor preservation typical correlation analysis, and the extracted face features not only maximize the correlation among different face data, but also optimally preserve the adjacent structure of the face, thereby improving the face recognition capability and stability.
The purpose of the invention is realized in the following way: a face recognition method based on neighbor preservation canonical correlation analysis comprises the following steps:
step 1: input face training dataset X epsilon R m×N ，Y∈R n×N Neighbor weight reconstruction matrix U for computing images by neighbor preserving learning x And U y ；
Step 2: finding two sets of projection vectors w using a canonical correlation analysis x And w y Introducing neighbor preservation into a frame of typical correlation analysis by using an optimization method, and calculating a projection matrix W by using generalized eigenvalue decomposition x And W is y ；
Step 3: in the feature fusion stage, two feature fusion strategies are adopted to test the low-dimensional projection of the face imageAnd->Fusing;
step 4: and using the nearest neighbor classifier to use the fused features for face recognition.
As a further limitation of the present invention, the step 1 specifically includes inputting face training data x= [ X ] 1 ,x 2 ,...,x N ]∈R m×N And Y= [ Y ] 1 ,y 2 ,...,y N ]∈R n×N Calculating k-nearest neighbor reconstruction weight matrix U of training sample x And U y ，U x And U y Can be obtained by minimizing the following objective function:
and
wherein the method comprises the steps ofAnd->Respectively represent human face samples x i And y i K-neighbor samples of (2) to obtain U by calculation x ＝(u x,ij ) And U y ＝(u y,ij )。
As a further definition of the present invention, the step 2 specifically includes: establishing an optimization function of neighbor preservation representative correlation analysis:
wherein S is xy ＝XY T Representing the cross covariance matrix of X and Y, describing the correlation between the two sets of variables; s is S w Representing the intra-class correlation matrix, S b Represents an inter-class correlation matrix, wherein,
m represents the same type of face sample set, C represents different type of face sample set,and->Representing a local linear reconstruction error matrix defined as follows:
wherein U is x And U y Reconstructing a weight matrix for the neighbor, wherein I is an identity matrix;
using the evaluation function method, the optimization function (1) can be transformed into the following optimization function (2):
by using Lagrange multiplier method, the optimization function (2) can be converted into the following two generalized eigenvalue problems:
in actual calculation, only d eigenvectors corresponding to the first d maximum non-zero eigenvalues of one generalized eigenvalue problem need to be solved, and then the other group of projection directions are solved by using the following relation:
obtaining projections of two sets of face features X and YMatrix arrayAnd
as a further limitation of the present invention, the step 3 specifically includes, for any one of the test samples [ x ] T ,y T ] T Can obtain its low-dimensional projection asAnd->And adopts two feature combination strategies to fuse, wherein the two feature combination strategies comprise a feature fusion strategy 1: />Feature fusion strategy 2: />
Compared with the prior art, the invention has the beneficial effects that: aiming at the condition that label information and local information of a face are not introduced in the traditional typical correlation analysis face recognition, the method not only provides that the extracted two groups of face features have the largest correlation coefficient, but also requires that local linear reconstruction errors in the two face feature sets after projection are as small as possible. In order to better utilize the face label information, improve the discrimination capability of extracting the face characteristics, maximize the correlation between the same face characteristics and minimize the correlation between different face characteristics, when calculating the neighbor weight reconstruction matrix of the image, label information is introduced, and adjacent points are selected from the similar face images. In the feature fusion stage, two feature fusion strategies are adopted. According to the invention, the neighbor is introduced into a typical correlation analysis frame by using an optimization method through the neighbor preservation learning face neighbor weight reconstruction matrix, and then the label information of the face is utilized, so that the extracted face features not only maximize the correlation among different faces, but also maintain the neighbor structure of the face to the greatest extent, and improve the face recognition capability and stability.
Drawings
FIG. 1 is a flow chart of an implementation of the present invention.
Fig. 2 is an ORL face recognition rate graph under feature fusion strategy 1.
Fig. 3 is an ORL face recognition rate graph under feature fusion strategy 2.
Fig. 4 is a Yale face recognition rate diagram under the feature fusion strategy 1.
Fig. 5 is a Yale face recognition rate diagram under the feature fusion strategy 2.
Detailed Description
The face recognition method based on the neighbor preserving typical correlation analysis as shown in fig. 1 comprises the following steps:
step 1: input face training dataset X epsilon R m×N ，Y∈R n×N Neighbor weight reconstruction matrix U for computing images by neighbor preserving learning x And U y ；
Inputting face training data X= [ X ] 1 ,x 2 ,...,x N ]∈R m×N And Y= [ Y ] 1 ,y 2 ,...,y N ]∈R n×N Calculating k-nearest neighbor reconstruction weight matrix U of training sample x And U y ，U x And U y Can be obtained by minimizing the following objective function:
and
wherein the method comprises the steps ofAnd->Respectively represent human face samples x i And y i K-neighbor samples of (2) to obtain U by calculation x ＝(u x,ij ) And U y ＝(u y,ij )。
Step 2: finding two sets of projection vectors w using a canonical correlation analysis x And w y Introducing neighbor preservation into a frame of typical correlation analysis by using an optimization method, and calculating a projection matrix W by using generalized eigenvalue decomposition x And W is y ；
Establishing an optimization function of neighbor preservation representative correlation analysis:
wherein S is xy ＝XY T Representing the cross covariance matrix of X and Y, describing the correlation between the two sets of variables; s is S w Representing the intra-class correlation matrix, S b Represents an inter-class correlation matrix, wherein,
m represents the same type of face sample set, C represents different type of face sample set,and->Representing a local linear reconstruction error matrix defined as follows:
wherein U is x And U y Reconstructing a weight matrix for the neighbor, wherein I is an identity matrix;
using the evaluation function method, the optimization function (1) can be transformed into the following optimization function (2):
by using Lagrange multiplier method, the optimization function (2) can be converted into the following two generalized eigenvalue problems:
in actual calculation, only d eigenvectors corresponding to the first d maximum non-zero eigenvalues of one generalized eigenvalue problem need to be solved, and then the other group of projection directions are solved by using the following relation:
obtaining projection matrix of two groups of face features X and YAnd
step 3: in the feature fusion stage, two feature fusion strategies are adopted to test the low-dimensional projection of the face imageAnd->Fusing;
for any one test sample [ x T ,y T ] T Can obtain its low-dimensional projection asAnd->And adopts two feature combination strategies to fuse, wherein the two feature combination strategies comprise a feature fusion strategy 1: />Feature fusion strategy 2: />
Step 4: and using the nearest neighbor classifier to use the fused features for face recognition.
In order to test the effectiveness of the present invention, the following experiments were performed, first the experimental parameter settings were performed:
ORL face database: the database contains 400 face images of 40 persons, 10 images per person, each image having a resolution of 92 x 112. These face images include expression changes, posture changes, and the like. During an experiment, the first 5 images of each person are selected as training samples, the last 5 images are selected as test samples, and wavelet 4-degree analysis is carried out on the original face images to obtain low-frequency components as a second feature set; in addition, the main component analysis (PCA) is adopted to reduce the dimension of the face image, so that the calculation complexity is reduced.
Yale face database: the database contains 165 gray-scale face images of 15 persons, and the resolution of each image is 120 multiplied by 91. Each person had 11 images, including three different illumination directions left, right and front, and eye, expression changes (normal, happy, sad, drowsy, surprised and blinked); in the experiment, each person selects the first 5 images as a training set, the other 6 images as a testing set, and performs wavelet 4-component analysis on the original face image to obtain a low-frequency component as a second feature set, and PCA is adopted to perform dimension reduction, so that the calculation complexity is reduced.
The experiment adopts a nearest neighbor classifier for classification, the performance measurement parameter is identification accuracy (A), and the formula is as follows:
experiment 1ORL face database recognition rate analysis
As shown in fig. 2 and 3, the classification recognition rate of the ORL face database of university of cambridge in united kingdom under different feature fusion strategies of the algorithm is shown, the abscissa represents the feature dimension after projection, and the ordinate represents the classification recognition accuracy. As can be seen from experimental results, the classification recognition rate of a plurality of algorithms can be increased along with the increase of the feature dimension, when the feature dimension is increased to a certain dimension, the recognition rate can be maintained at a stable level, the recognition rate of the algorithm provided by the invention is higher than that of other related algorithms of CCA under most conditions of an ORL face database, and when the feature dimension reaches 20 dimensions in the ORL database, the classification recognition rate is over 90 percent, and finally, the recognition rate is stabilized to be about 95 percent.
Experimental 2Yale face database recognition rate analysis
As shown in fig. 4 and fig. 5, the classification recognition rate of the Yale face database of the algorithm under different feature fusion strategies is shown, the abscissa represents the feature dimension after projection, and the ordinate represents the recognition rate. As shown by experimental results, as fewer Yale face training samples are used for 165 faces in total, under the feature fusion strategies 1 and 2, the face identification information extracted by the algorithm is weaker, the fluctuation of the identification rate is larger, the identification result is lower than that of an ORL face database, but the identification rate is still higher than that of other algorithms.
The invention is not limited to the above embodiments, and based on the technical solution disclosed in the invention, a person skilled in the art may make some substitutions and modifications to some technical features thereof without creative effort according to the technical content disclosed, and all the substitutions and modifications are within the protection scope of the invention.
Claims (3)
1. The face recognition method based on the neighbor preserving canonical correlation analysis is characterized by comprising the following steps of:
step 1: input face training dataset X epsilon R m×N ，Y∈R n×N Wherein m and N respectively represent the dimensions of face training samples in data sets X and Y, N represents the number of face training samples, R represents a real set, and a neighbor weight reconstruction matrix U of an image is calculated through neighbor preservation learning x And U y ；
Step 2: finding two sets of projection vectors w using a canonical correlation analysis x And w y Introducing neighbor preservation into a frame of typical correlation analysis by using an optimization method, and calculating a projection matrix W by using generalized eigenvalue decomposition x And W is y ；
The step 2 specifically includes: establishing an optimization function of neighbor preservation representative correlation analysis:
wherein S is xy ＝XY T Representing the cross covariance matrix of X and Y, describing the correlation between the two sets of variables; s is S w Representing the intra-class correlation matrix, S b Represents an inter-class correlation matrix, wherein,
M(x i ) Representation and x i Similar face sample set, M (y i ) Representation and y i Similar face sample set, C (x i ) Representation and x i Different classes of face sample sets, C (y i ) Representation and y i Different types of face sample sets, x i And y i Representing the ith face sample in training sample sets X and Y, i=1, 2, …, N, respectively, when calculating S w When x is j And y j Respectively represent M (x) i ) And M (y) i ) The j-th face sample in (1) when calculating S b When x is j And y j Respectively represent C (x) i ) And C (y) i ) In the (c) j-th face sample,and->Representing a local linear reconstruction error matrix defined as follows:
wherein U is x And U y Reconstructing a weight matrix for the neighbor, wherein I is an identity matrix;
using the evaluation function method, the optimization function (1) can be transformed into the following optimization function (2):
by using Lagrange multiplier method, the optimization function (2) can be converted into the following two generalized eigenvalue problems:
in actual calculation, only d eigenvectors corresponding to the first d maximum non-zero eigenvalues of one generalized eigenvalue problem need to be solved, and then the other group of projection directions are solved by using the following relation:
obtaining projection matrix of two groups of face features X and YAnd->
Step 3: in the feature fusion stage, two feature fusion strategies are adopted to test the low-dimensional projection of the face imageAndfusing;
step 4: and using the nearest neighbor classifier to use the fused features for face recognition.
2. The face recognition method based on the neighbor preserving canonical correlation analysis of claim 1, wherein the step 1 specifically includes inputting face training data x= [ X ] 1 ,x 2 ,...,x N ]∈R m×N And Y= [ Y ] 1 ,y 2 ,...,y N ]∈R n×N Calculating k-nearest neighbor reconstruction weight matrix U of training sample x And U y ，U x And U y Can be obtained by minimizing the following objective function:
and
wherein the method comprises the steps ofAnd->Respectively represent human face samples x i And y i K-neighbor samples of (2) to obtain U by calculation x ＝(u x,ij ) And U y ＝(u y,ij )。
3. The face recognition method based on the neighbor preserving canonical correlation analysis of claim 1, wherein the step 3 specifically includes: for any one test sample [ x T ,y T ] T Can obtain its low-dimensional projection asAnd->And adopts two feature combination strategies to fuse, wherein the two feature combination strategies comprise a feature fusion strategy 1: />Feature fusion strategy 2: />
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010473892.2A CN111611963B (en) | 2020-05-29 | 2020-05-29 | Face recognition method based on neighbor preservation canonical correlation analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010473892.2A CN111611963B (en) | 2020-05-29 | 2020-05-29 | Face recognition method based on neighbor preservation canonical correlation analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111611963A CN111611963A (en) | 2020-09-01 |
| CN111611963B true CN111611963B (en) | 2023-12-26 |
Family
ID=72204717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010473892.2A Active CN111611963B (en) | 2020-05-29 | 2020-05-29 | Face recognition method based on neighbor preservation canonical correlation analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111611963B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112966735B (en) * | 2020-11-20 | 2023-09-12 | 扬州大学 | Method for fusing supervision multi-set related features based on spectrum reconstruction |
| CN112989218B (en) * | 2021-03-12 | 2022-06-28 | 西华大学 | Identity linking method based on multilevel attribute embedding and constraint canonical correlation analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105893947A (en) * | 2016-03-29 | 2016-08-24 | 江南大学 | Bi-visual-angle face identification method based on multi-local correlation characteristic learning |
| CN106203256A (en) * | 2016-06-24 | 2016-12-07 | 山东大学 | A kind of low resolution face identification method based on sparse holding canonical correlation analysis |
| CN107133648A (en) * | 2017-05-05 | 2017-09-05 | 南京航空航天大学 | The sparse one-dimensional range profile recognition methods for keeping projecting is merged based on self-adapting multi-dimension |
| CN108520279A (en) * | 2018-04-12 | 2018-09-11 | 上海海洋大学 | A kind of semi-supervised dimension reduction method of high spectrum image of the sparse insertion in part |
| CN108647690A (en) * | 2017-10-17 | 2018-10-12 | 南京工程学院 | The sparse holding projecting method of differentiation for unconstrained recognition of face |
-
2020
- 2020-05-29 CN CN202010473892.2A patent/CN111611963B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105893947A (en) * | 2016-03-29 | 2016-08-24 | 江南大学 | Bi-visual-angle face identification method based on multi-local correlation characteristic learning |
| CN106203256A (en) * | 2016-06-24 | 2016-12-07 | 山东大学 | A kind of low resolution face identification method based on sparse holding canonical correlation analysis |
| CN107133648A (en) * | 2017-05-05 | 2017-09-05 | 南京航空航天大学 | The sparse one-dimensional range profile recognition methods for keeping projecting is merged based on self-adapting multi-dimension |
| CN108647690A (en) * | 2017-10-17 | 2018-10-12 | 南京工程学院 | The sparse holding projecting method of differentiation for unconstrained recognition of face |
| CN108520279A (en) * | 2018-04-12 | 2018-09-11 | 上海海洋大学 | A kind of semi-supervised dimension reduction method of high spectrum image of the sparse insertion in part |
Non-Patent Citations (4)
| Title |
|---|
| 张其文 等."基于监督协同近邻保持投影的人脸识别算法".《计算机科学》.2015,第1-3. * |
| 杨林 等.基于特征融合的KLPCCA及其在人脸识别中的应用.科技信息.2009,(33),全文. * |
| 许洁 等."稀疏保持典型相关分析特征选择与模式识别".《小型微型计算机系统》.2017,第1-5节. * |
| 黄晓华 等.基于彩色人脸图像的信息融合与识别方法.中国图象图形学报.2010,(03),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111611963A (en) | 2020-09-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zuo et al. | Bidirectional PCA with assembled matrix distance metric for image recognition | |
| CN109711283B (en) | Occlusion expression recognition method combining double dictionaries and error matrix | |
| CN110659665B (en) | Model construction method of different-dimension characteristics and image recognition method and device | |
| Zhai et al. | BeautyNet: Joint multiscale CNN and transfer learning method for unconstrained facial beauty prediction | |
| Fellenz et al. | Comparing template-based, feature-based and supervised classification of facial expressions from static images | |
| CN107220627B (en) | Multi-pose face recognition method based on collaborative fuzzy mean discrimination analysis | |
| CN108932501B (en) | Face recognition method based on multi-core association integration dimension reduction | |
| CN111611963B (en) | Face recognition method based on neighbor preservation canonical correlation analysis | |
| Hosseini et al. | GF-CapsNet: Using gabor jet and capsule networks for facial age, gender, and expression recognition | |
| Wang et al. | Research on face recognition technology based on PCA and SVM | |
| CN110956113B (en) | Robust face recognition method based on secondary cooperation representation identification projection | |
| CN109543637A (en) | A kind of face identification method, device, equipment and readable storage medium storing program for executing | |
| CN110399814B (en) | Face recognition method based on local linear representation field adaptive measurement | |
| CN114511901A (en) | Age classification assisted cross-age face recognition algorithm | |
| Bhattacharya et al. | Edge-enhanced bi-dimensional empirical mode decomposition-based emotion recognition using fusion of feature set | |
| Zhang et al. | Compound structure classifier system for ear recognition | |
| CN111723759B (en) | Unconstrained face recognition method based on weighted tensor sparse graph mapping | |
| CN111079657B (en) | Face recognition method based on semi-supervised linear regression | |
| Lacroix et al. | Modeling recognition memory using the similarity structure of natural input | |
| Gül | Holistic face recognition by dimension reduction | |
| CN111783615B (en) | Unconstrained face recognition method based on weighting block tensor sparse map mapping | |
| Hiremath et al. | Face recognition using Eigenface approach | |
| Zhou | Analysis of Psychological and Emotional Tendency Based on Brain Functional Imaging and Deep Learning | |
| CN112241680A (en) | Multi-mode identity authentication method based on vein similar image knowledge migration network | |
| Hasan et al. | A Study of Gender Classification Techniques Based on Iris Images: A Deep Survey and Analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |