CN110298224A - A kind of face age estimation method analyzed based on direction gradient and hidden variable - Google Patents
A kind of face age estimation method analyzed based on direction gradient and hidden variable Download PDFInfo
- Publication number
- CN110298224A CN110298224A CN201910242005.8A CN201910242005A CN110298224A CN 110298224 A CN110298224 A CN 110298224A CN 201910242005 A CN201910242005 A CN 201910242005A CN 110298224 A CN110298224 A CN 110298224A
- Authority
- CN
- China
- Prior art keywords
- feature
- age
- formula
- face
- gradient
- 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.)
- Pending
Links
Classifications
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration by the use of histogram techniques
-
- 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
- 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
-
- 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
- G06V40/178—Human faces, e.g. facial parts, sketches or expressions estimating age from face image; using age information for improving recognition
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Collating Specific Patterns (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of face age estimation method analyzed based on more dimensional directions histogram of gradients (Histogram of oriented gradient) feature and hidden variable, overall step are as follows: S1: data set is simultaneously divided into training set, verifying collection two parts by building data set；S2: the histograms of oriented gradients feature of more sizes is extracted to picture；S3: training hidden variable analysis model simultaneously filters out age unrelated feature using the model from the feature that S2 is extracted；S4: device is returned using S3 treated feature training, completes the estimation to personage's age in facial image.More size characteristic extracting methods based on face priori knowledge that present invention firstly provides a kind of, then the age for including in feature unrelated information is filtered out using hidden variable analysis method, compared to the accuracy that more traditional method based on local shape factor improves age estimation.
Description
Technical field
The present invention relates to technical field of computer vision more particularly to a kind of people analyzed based on direction gradient and hidden variable
Face age estimation method.
Background technique
The estimation of face age, i.e., estimate the age of people by individual human face photo, is important in face character analysis
Component part.Since it is in intelligent monitoring, business analysis, the extensive use in the fields such as human-computer interaction is always in the latest 20 years
The popular problem of one research.But the estimation of face age is also a challenging problem, because of the aging of different people
Process has very big otherness in the performance of face, that is to say, that the face appearance from a people is to its actual age
A kind of fuzzy mapping relations, and everyone mapping relations are different from.This also explains why some seem
It is practical more older than it and some are less than normal.
Current face age estimation method can be roughly divided into two types, the method based on classification and the side based on recurrence
Method.Method based on classification is that all ages and classes are regarded as to different classifications, it is different classes of between there is no correlation.Such processing
It has been ignored as stronger succession and correlation between all ages and classes, age estimation method performance is caused to decline.And based on recurrence
Method be that the age is regarded as to a continuous number, fitting age label is gone by the training picture feature extracted, in this way
Processing method be more in line with intuition.But there is scholar to point out in some research work, the former is compared, the side based on recurrence
Method is more easily trapped into over-fitting.By being further analyzed to the feature of extraction, rejecting unrelated noise and reducing data dimension
Degree can effectively alleviate overfitting problem.
Hidden variable is the variable that can not directly observe in finger to finger test, can usually be seen using statistical model to hidden variable
It examines, counts its probability nature to deduce hidden variable.Hidden variable is analyzed in psychology, economics, across age recognition of face etc.
Field has to be widely applied very much.It include various information, such as age, identity, mood etc., these letters in face
Breath can regard one group of hidden variable as, can analyze these information targetedly using hidden variable analysis method to improve people
The performance of face age estimation method.
Summary of the invention
It is an object of the invention to overcome the shortcomings of the prior art, provide a kind of based on multiple dimensioned direction gradient histogram
The face age estimation method for scheming (Histogram of oriented gradient) feature and hidden variable analysis, comprising following
Step:
Step 1: data preprocessing phase: using existing Face datection and face key point location algorithm to face figure
Piece carries out Face datection and crucial point location, cuts out human face region and picture is zoomed to the preservation of 128 × 160 sizes.
Step 2: data set divides: random division goes out 80% data of age data concentration as training set, is left 20%
Collect as verifying, guarantees that the data of the same person only occur in a set.
Step 3: training set is grouped: by the training set marked off in step 2 respectively according to age bracket and identity information point
Group simultaneously saves.
Step 4: extracting multiple dimensioned histograms of oriented gradients feature: the training set picture divided to step 2 carries out first
Gray processing handles and does gamma normalization, calculates each pixel horizontal direction of picture and vertical direction using gradient operator later
Gradient value, calculation formula such as formula one and formula two:
Gx(x, y)=H (x+1, y)-H (x-1, y) formula one
Gy(x, y)=H (x, y+1)-H (x, y-1) formula two
Wherein, Gx(x, y), Gy(x, y), H (x, y) respectively indicate the ladder of the horizontal direction in input picture at pixel (x, y)
Degree, vertical direction gradient and pixel value.Then calculate pixel (x, y) at gradient magnitude G (x, y) and gradient direction α (x,
Y), calculation formula such as formula three and formula four:
Then large-sized histograms of oriented gradients feature is synthesized in full figure, and according to the knot of point location crucial in step 1
Fruit extracts the histograms of oriented gradients feature of small size in circumference of eyes, finally by the conduct together of the merging features of two sizes
The feature of whole face.
Step 5: Feature Dimension Reduction: using Principal Component Analysis Algorithm to the histograms of oriented gradients feature extracted in step 4
Dimensionality reduction is carried out, 98% energy is retained.
Step 6: training hidden variable analysis model: the picture feature extracted in step 6 is modeled as age correlated characteristic
With the linear combination of identity correlated characteristic and other uncorrelated noises, modeling format such as formula five:
T=β+Ux+Vy+ ε formula five
Wherein, t is the feature extracted, and β is the mean value of feature in training set, and that Ux is indicated is age-dependent feature, Vy
What is indicated is the relevant feature of identity, and ε indicates other uncorrelated noises, it is assumed that its Normal Distribution: ε~N (0, δ2I).First
β is calculated, then according to the sample group divided in step 3 according to age bracket and identity information, utilizes greatest hope
(Expectation maximization) algorithm estimates parameter U, V, δ.It is calculated from former feature then according to formula six
Age-dependent feature simultaneously saves:
F=UUTΣ-1(t- β) formula six
Wherein, Σ=δ2I+UUT+VVT。
Step 7: training returns device: using the age-dependent feature extracted in step 6 as input, training one linear
It returns device model and saves, the present invention completes to return the training of device using existing linear (LibLinear) function library.
Step 8: method is tested and assessed: after the above step is finished, the picture that verifying is concentrated first extracts more rulers according to step 5
Very little histograms of oriented gradients feature.Trained hidden variable analysis model in step 6 is reused, year is extracted according to formula six
Age relevant feature.Age correlated characteristic is finally sent into the age that trained linear regressor is predicted in step 7,
And use mean absolute error as evaluation index assessment algorithm performance, the calculation formula of mean absolute error is shown in formula seven:
In conclusion the priori knowledge of the present invention first according to face information, after extracting global characteristics to face picture, needle
To the extraction that the minutia abundant for including around eyes is refined, the histograms of oriented gradients of more sizes has been synthesized
Feature.Since the feature of extraction contains unrelated noise of many ages, the present invention does feature using hidden variable parser
Further analysis, feature is all that height is age-dependent so that treated, to improve the accuracy of age estimation.
Detailed description of the invention
Fig. 1 is extraction histograms of oriented gradients feature flow chart in the present invention.
Fig. 2 is the overview flow chart of invention.
Specific embodiment
The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
A kind of face age estimation method analyzed based on direction gradient and hidden variable, overall step are as follows.
Step 1: data preprocessing phase: using existing Face datection and face key point location algorithm to face figure
Piece carries out Face datection and key point (68 key points) positioning, cuts out human face region (if it is not detected that face then will
The picture is deleted), picture is aligned according to the position of pupil and upper lip and zooms to the preservation of 128 × 160 sizes.This step packet
It includes but is not limited to be aligned based on 68 points into pedestrian's face.
Step 2: data set divide: in order to avoid model on data set over-fitting, improve the generalization ability of model.It needs
Data set is divided, the data that random division goes out 80% collect as training set, remaining 20% as verifying, and guarantee
The data of the same person only occur in a set.
Step 3: training set is grouped: in order to train hidden variable analysis model, needing the data of training set respectively according to year
Age section and identity information grouping, it is notable that the age distribution in view of data set sample is needed according to age bracket grouping
Situation, the age bracket of division will guarantee that the training sample number in each section is close.
Step 4: extracting multiple dimensioned histograms of oriented gradients feature: the training set picture divided to step 2 carries out first
Gray processing handles and does gamma normalization, calculates each pixel horizontal direction of picture and vertical direction using gradient operator later
Gradient value, calculation formula such as formula one and formula two:
Gx(x, y)=H (x+1, y)-H (x-1, y) formula one
Gy(x, y)=H (x, y+1)-H (x, y-1) formula two
Wherein, Gx(x, y), Gy(x, y), H (x, y) respectively indicate the horizontal direction in input picture at pixel (x, y)
Gradient, vertical direction gradient and pixel value.Then calculate pixel (x, y) at gradient magnitude G (x, y) and gradient direction α (x,
Y), calculation formula such as formula three and formula four:
Then large-sized histograms of oriented gradients is synthesized using 16 × 16 (or 8 × 8) as cell factory size in full figure
Feature, and according to the result of point location crucial in step 1 in the side of the small size of circumference of eyes 8 × 8 (or 4 × 4) sizes of extraction
To histogram of gradients feature, the merging features of two sizes are finally used as to the feature of whole face, the process of extraction together
As shown in Figure 1.
Step 5: Feature Dimension Reduction: since the histograms of oriented gradients extracted in step 4 is characterized in more sizes, leading to spy
It is very high to levy dimension, it is therefore necessary to carry out dimension-reduction treatment to feature to reduce the operand of subsequent algorithm, the present invention using it is main at
The energy for dividing parser to carry out dimensionality reduction and reservation 98%.This step is including but not limited to the dimensionality reduction side for using principal component analysis
Method.
Step 6: training hidden variable analysis model: the picture feature extracted in step 6 is modeled as age correlated characteristic
With the linear combination of identity correlated characteristic and other uncorrelated noises, modeling format such as formula five:
T=β+Ux+Vy+ ε formula five
Wherein, t is the feature extracted, and β is the mean value of feature in training set, and what Ux was indicated is age-dependent feature, and x is
The hidden variable factor of age correlated characteristic, it is assumed that its Normal Distribution: x~N (0, I), U are parameters to be estimated.Vy is indicated
Be the relevant feature of identity, y is the hidden variable factor of identity correlated characteristic, it is assumed that its Normal Distribution: x~N (0, I), V
It is parameter to be estimated.ε indicates other uncorrelated noises, it is assumed that its Normal Distribution: ε~N (0, δ 2I).β is calculated first,
Then according to the sample group divided in step 3 according to age bracket and identity information, greatest hope (Expectation is utilized
Maximization) algorithm estimates parameter U, V, δ.Age-dependent feature is calculated from former feature then according to formula six
And it saves:
F=UUTΣ-1(t- β) formula six
Wherein, Σ=δ2I+UUT+VVT。
Step 7: training returns device: the age-dependent feature extracted in step 6 is linear as input training one
Device and preservation model are returned, the present invention completes to return the training of device using existing linear (LibLinear) function library.
Step 8: method is tested and assessed: after the above step is finished, the picture that verifying is concentrated first extracts more rulers according to step 5
Very little histograms of oriented gradients feature.Trained hidden variable analysis model in step 6 is reused, year is extracted according to formula six
Age relevant feature.Age correlated characteristic is finally sent into the age that trained linear regressor is predicted in step 7,
And use mean absolute error as evaluation index assessment algorithm performance, the calculation formula of mean absolute error is shown in formula seven:
Overall procedure of the invention is as shown in Fig. 2, innovative point and key point of the invention is as follows.
(1) priori knowledge according to face information, after extracting large scale histograms of oriented gradients feature to face picture,
The equal region comprising detailed information very abundant, reduces the ruler of cell factory in histograms of oriented gradients feature around eyes
It is very little, to extract finer feature, two kinds of various sizes of merging features are played to the spy as whole face picture later
Sign.Such processing method both can largely retain the detailed information in face, while it is excessive to avoid intrinsic dimensionality
Situation.
It (2) is a kind of unsupervised method due to using feature operator to extract picture feature, so in the feature extracted not
The information and noise item unrelated there are many ages avoidablely.The present invention is using hidden variable analysis method to more rulers of extraction
Very little histograms of oriented gradients feature is further analyzed, and eliminates the interference that age irrelevant information estimates the age, to mention
The high accuracy of age estimation.
Above embodiment is not limitation of the present invention, and the present invention is also not limited to the example above, this technology neck
The variations, modifications, additions or substitutions that the technical staff in domain is made within the scope of technical solution of the present invention, also belong to this hair
Bright protection scope.
Claims (3)
1. a kind of face age estimation method analyzed based on direction gradient and hidden variable, it is characterised in that: the method it is whole
Body step are as follows:
Step 1: data preprocessing phase: using existing Face datection and face key point location algorithm to face picture into
Row Face datection and crucial point location cut out human face region and picture are zoomed to the preservation of 128 × 160 sizes；
Step 2: data set divides: random division goes out 80% data of age data concentration as training set, is left 20% conduct
Verifying collection guarantees that the data of the same person only occur in a set；
Step 3: training set is grouped: the training set marked off in step 2 is grouped simultaneously according to age bracket and identity information respectively
It saves；
Step 4: extracting multiple dimensioned histograms of oriented gradients (Histogram of oriented gradient) feature: to step
The rapid two training set pictures divided carry out gray processing processing first and do gamma normalization, calculate picture using gradient operator later
The gradient value of each pixel horizontal direction and vertical direction, calculation formula such as formula one and formula two:
Gx(x, y)=H (x+1, y)-H (x-1, y) formula one
Gy(x, y)=H (x, y+1)-H (x, y-1) formula two
Wherein, Gx(x, y), Gy(x, y), H (x, y) respectively indicate horizontal direction gradient in input picture at pixel (x, y),
Vertical direction gradient and pixel value；
Then the gradient magnitude G (x, y) and gradient direction α (x, y) at pixel (x, y), calculation formula such as three He of formula are calculated
Formula four:
Then large-sized histograms of oriented gradients feature is synthesized in full figure, and is existed according to the result of point location crucial in step 1
Circumference of eyes extracts the histograms of oriented gradients feature of small size, finally regard the merging features of two sizes as whole together
The feature of face；
Step 5: Feature Dimension Reduction: being carried out using Principal Component Analysis Algorithm to the histograms of oriented gradients feature extracted in step 4
Dimensionality reduction retains 98% energy；
Step 6: training hidden variable analysis model: the picture feature extracted in step 6 is modeled as age correlated characteristic and body
The linear combination of part correlated characteristic and other uncorrelated noises, modeling format such as formula five:
T=β+Ux+Vy+ ε formula five
Wherein, t is the feature extracted, and β is the mean value of feature in training set, and what Ux was indicated is age-dependent feature, and Vy is indicated
Be the relevant feature of identity, ε indicates other uncorrelated noises, it is assumed that its Normal Distribution: ε~N (0, δ2I)；
β is calculated first, then according to the sample group divided in step 3 according to age bracket and identity information, utilizes greatest hope
(Expectation maximization) algorithm estimates parameter U, V, δ；
Age-dependent feature is calculated from former feature then according to formula six and is saved:
F=UUTΣ-1(t- β) formula six
Wherein, Σ=δ2I+UUT+VVT；
Step 7: training returns device: using the age-dependent feature extracted in step 6 as input, one linear regression of training
Device model simultaneously saves, and the present invention completes to return the training of device using existing linear (LibLinear) function library；
Step 8: method is tested and assessed: after the above step is finished, the picture that verifying is concentrated first extracts more size sides according to step 5
To histogram of gradients feature；
Trained hidden variable analysis model in step 6 is reused, age-dependent feature is extracted according to formula six；
Age correlated characteristic is finally sent into the age that trained linear regressor is predicted in step 7, and using average
Absolute error is shown in formula seven as evaluation index assessment algorithm performance, the calculation formula of mean absolute error:
2. the method as described in claim 1, it is characterised in that: first extract large scale direction to whole picture in the step 4
Histogram of gradients feature (cell factory is having a size of 16 × 16 or 8 × 8), then to ocular vicinity extracted region small size direction gradient
Histogram feature (cell factory is having a size of 8 × 8 or 4 × 4).
3. method according to claim 1 or 2, it is characterised in that: the multiple dimensioned direction gradient that will be extracted in the step 5
Histogram feature first carries out dimensionality reduction, the feature modeling after dimensionality reduction at age correlated characteristic and identity correlated characteristic and other nothings
The linear combination of noise is closed, and uses the parameter of EM algorithm estimation hidden variable analysis model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910242005.8A CN110298224A (en) | 2019-03-28 | 2019-03-28 | A kind of face age estimation method analyzed based on direction gradient and hidden variable |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910242005.8A CN110298224A (en) | 2019-03-28 | 2019-03-28 | A kind of face age estimation method analyzed based on direction gradient and hidden variable |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110298224A true CN110298224A (en) | 2019-10-01 |
Family
ID=68026555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910242005.8A Pending CN110298224A (en) | 2019-03-28 | 2019-03-28 | A kind of face age estimation method analyzed based on direction gradient and hidden variable |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110298224A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112328907A (en) * | 2020-11-05 | 2021-02-05 | 重庆第二师范学院 | Learning content recommendation method |
| CN117095434A (en) * | 2023-07-24 | 2023-11-21 | 山东睿芯半导体科技有限公司 | Face recognition method, chip and terminal for different ages |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100217743A1 (en) * | 2007-09-28 | 2010-08-26 | Nec Soft, Ltd. | Attribute estimation system, age estimation system, gender estimation system, age and gender estimation system and attribute estimation method |
| CN104463190A (en) * | 2014-10-30 | 2015-03-25 | 华为技术有限公司 | Age estimation method and equipment |
| CN105550641A (en) * | 2015-12-04 | 2016-05-04 | 康佳集团股份有限公司 | Age estimation method and system based on multi-scale linear differential textural features |
| CN106778584A (en) * | 2016-12-08 | 2017-05-31 | 南京邮电大学 | A kind of face age estimation method based on further feature Yu shallow-layer Fusion Features |
-
2019
- 2019-03-28 CN CN201910242005.8A patent/CN110298224A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100217743A1 (en) * | 2007-09-28 | 2010-08-26 | Nec Soft, Ltd. | Attribute estimation system, age estimation system, gender estimation system, age and gender estimation system and attribute estimation method |
| CN104463190A (en) * | 2014-10-30 | 2015-03-25 | 华为技术有限公司 | Age estimation method and equipment |
| CN105550641A (en) * | 2015-12-04 | 2016-05-04 | 康佳集团股份有限公司 | Age estimation method and system based on multi-scale linear differential textural features |
| CN106778584A (en) * | 2016-12-08 | 2017-05-31 | 南京邮电大学 | A kind of face age estimation method based on further feature Yu shallow-layer Fusion Features |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112328907A (en) * | 2020-11-05 | 2021-02-05 | 重庆第二师范学院 | Learning content recommendation method |
| CN117095434A (en) * | 2023-07-24 | 2023-11-21 | 山东睿芯半导体科技有限公司 | Face recognition method, chip and terminal for different ages |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108229330A (en) | Face fusion recognition methods and device, electronic equipment and storage medium | |
| CN103810490B (en) | A kind of method and apparatus for the attribute for determining facial image | |
| Chen et al. | Why recognition in a statistics-based face recognition system should be based on the pure face portion: a probabilistic decision-based proof | |
| Dehshibi et al. | A new algorithm for age recognition from facial images | |
| US20180181834A1 (en) | Method and apparatus for security inspection | |
| CN103218609B (en) | A kind of Pose-varied face recognition method based on hidden least square regression and device thereof | |
| CN110084259A (en) | A kind of facial paralysis hierarchical synthesis assessment system of combination face texture and Optical-flow Feature | |
| CN110464366A (en) | A kind of Emotion identification method, system and storage medium | |
| CN109034092A (en) | Accident detection method for monitoring system | |
| CN108898125A (en) | One kind being based on embedded human face identification and management system | |
| CN106599785B (en) | Method and equipment for establishing human body 3D characteristic identity information base | |
| CN106778489A (en) | The method for building up and equipment of face 3D characteristic identity information banks | |
| CN106529377A (en) | Age estimating method, age estimating device and age estimating system based on image | |
| Monwar et al. | Pain recognition using artificial neural network | |
| CN110298224A (en) | A kind of face age estimation method analyzed based on direction gradient and hidden variable | |
| CN109117774A (en) | A kind of multi-angle video method for detecting abnormality based on sparse coding | |
| CN106778491B (en) | The acquisition methods and equipment of face 3D characteristic information | |
| Liu et al. | Human gait recognition for multiple views | |
| CN108288040A (en) | Multi-parameter face identification system based on face contour | |
| CN111694980A (en) | Robust family child learning state visual supervision method and device | |
| Lanitis | Facial Biometric Templates and Aging: Problems and Challenges for Artificial Intelligence. | |
| CN110135362A (en) | A kind of fast face recognition method based under infrared camera | |
| Thanh Do et al. | Facial feature extraction using geometric feature and independent component analysis | |
| Pirozmand et al. | Age estimation, a gabor pca-lda approach | |
| TW201839635A (en) | Emotion detection system and method |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191001 |
|
| WD01 | Invention patent application deemed withdrawn after publication |