CN109800643A - A kind of personal identification method of living body faces multi-angle - Google Patents
A kind of personal identification method of living body faces multi-angle Download PDFInfo
- Publication number
- CN109800643A CN109800643A CN201811537149.8A CN201811537149A CN109800643A CN 109800643 A CN109800643 A CN 109800643A CN 201811537149 A CN201811537149 A CN 201811537149A CN 109800643 A CN109800643 A CN 109800643A
- Authority
- CN
- China
- Prior art keywords
- face
- recognition result
- angle
- positive
- formula
- 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.)
- Granted
Links
Abstract
The invention discloses a kind of personal identification methods of living body faces multi-angle, comprising: if the coordinate of 5 key points, which calculates, meets the first formula, judges the face that is positive；If key point coordinate when positive face and side face meets the second formula or third formula, judges that user's head turns left or turns right, acquire user's left side of the face or right side face image；The facial image of three the positive face of user, left side of the face, right side face angles is respectively fed to convolutional neural networks model, exports the feature vector of 2622 dimensions respectively；Feature vector in this feature vector and face database does normalization dot product one by one, obtains similarity score list；One of highest scoring in similarity list is taken as face recognition result, final face recognition result is voted by the recognition result of three angle faces and generated, and poll is at most final recognition result.This invention removes the influences of head pose variability；At the same time, the left and right head shaking movement during user shows multi-orientation Face can be used as the foundation of face In vivo detection.
Description
Technical field
The present invention relates to face In vivo detection and field of face identification more particularly to a kind of identity of living body faces multi-angle
Recognition methods.
Background technique
Recognition of face is a hot issue in computer vision and machine learning field.It is in video monitoring, access control
Key effect is played in the application scenarios such as system, man-machine interface.Face recognition technology be used to identify user identity, compared to fingerprint,
Other human body living things feature recognitions such as gene, it is not noticeable and under non-mated condition that recognition of face provides a kind of user
Recognition methods.In general, face identification system include four modules: Face datection, face critical point detection, face characterization and
Identification.
In the 1970s, face recognition technology has had begun research like a raging fire.Classical face recognition algorithms
There are PCA (principal component analysis) method, LDA (linear discriminant analysis) method, Elastic Matching technology and bayes method etc..With
Many computer vision problems are the same, and recognition of face problem can be divided into conventional method and deep learning method.Conventional method packet
Include geometrical measurers, sub-space analysis method, statistical nature method, template matching method etc..Deep learning method emerges DeepID[1]、
DeepFace[2]、VGGFace[3]、FaceNet[4]Etc. serial of methods.
Although recognition of face is academicly having been achieved for significant progress, and takes in various complicated face data sets
Obtain very high accuracy rate.But in practical applications, the variability of complicated illumination condition and face head pose makes people
Face identification is still challenging.On the other hand, many safety issues can be generated by carrying out identification using only face planar picture,
Such as cheated using the face picture of legitimate user, video, model or mask means.It is therefore desirable in recognition of face
In vivo detection is carried out in the process.
A kind of Multi-angle human face recognition method and system (publication number CN102609695A, publication date 2012.07.25) use
Multiple cameras acquire facial image, but only select face image to collected Multi-angle human face image and identify.Base
Fusion is used in the living body determination method and equipment (publication number CN105389554A, publication date 2016.03.09) of recognition of face
Method is extracted the texture variations feature near specular reflective characteristics and face key point and is merged, but extracts everyone
Texture variations feature calculation amount near face key point is very big.
Summary of the invention
The present invention provides a kind of personal identification method of living body faces multi-angle, the present invention utilizes multi-orientation Face feature
Information carries out decision level fusion, eliminates the influence of head pose variability；At the same time, user shows the process of multi-orientation Face
In left and right head shaking movement can be used as the foundation of face In vivo detection, it is described below:
A kind of personal identification method of living body faces multi-angle, the described method comprises the following steps:
Learn the local binary feature of each key point by cascading regression tree, and combine, is examined using linear regression
Key point is surveyed, 5 key points are finally obtained；
If the coordinate of 5 key points, which calculates, meets the first formula, the face that is positive is judged；
If key point coordinate when positive face and side face meets the second formula or third formula, judge user's head turn left or
It turns right, acquires user's left side of the face or right side face image；
The facial image of three the positive face of user, left side of the face, right side face angles is respectively fed to convolutional neural networks model, point
It Shu Chu not 2622 feature vectors tieed up；Feature vector in this feature vector and face database does normalization dot product one by one, obtains phase
Like property score list；
One of highest scoring in similarity list is taken as face recognition result, final face recognition result is by three
The recognition result of angle face, which is voted, to be generated, and poll is at most final recognition result.
Wherein, 5 key points specifically: the left and right tail of the eye, nose, the left and right corners of the mouth.
Further, first formula specifically:
Wherein, (ax, ay), (bx, by), (cx, cy), (dx, dy), the transverse and longitudinal that (ex, ey) is 5 key points of face are sat
Mark.
Wherein, second formula specifically:
Wherein, the ordinate of face key point A1 and B1 when ay1, by1 are side face, face when ax1, bx1, cx1 are side face
The abscissa of key point A1, B1, C1, ax0, bx0, cx0 be positive face when face key point A0, B0, C0 abscissa.
Further, the third formula specifically:
。
When specific implementation, the method also includes:
If the recognition result of three angle faces is different, using the face recognition result of positive face angle as final people
Face recognition result.
The beneficial effect of the technical scheme provided by the present invention is that:
1, this method comprehensively considers the face characteristic of three angles, ballot obtains recognition result, effective solution head
Posture variability bring misidentifies problem, enhances the stability and accuracy rate of recognition of face；
2, the present invention can carry out In vivo detection to active user, increase while acquiring user's Multi-angle human face image
The strong safety of recognition of face, effectively antagonizes the attack of human face photo；
3, this method can be adapted for cell phone application face registration login system, intelligent vehicle device driver identification identifying system, building
The concrete application scenes such as space gate inhibition's testimony of a witness identifying system and intelligent camera head's identification system.
Detailed description of the invention
Fig. 1 is a kind of flow chart of the personal identification method of living body faces multi-angle；
Fig. 2 is the schematic diagram of 68 key points of face；
Fig. 3 is the schematic diagram of the coordinate representation of 5 key points of face；
Fig. 4 is positive face Face datection and critical point detection lab diagram；
Fig. 5 is left side of the face Face datection and critical point detection lab diagram；
Fig. 6 is right side face Face datection and critical point detection lab diagram；
Fig. 7 is multi-orientation Face similarity score list lab diagram.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, embodiment of the present invention is made below further
Ground detailed description.
Embodiment 1
A kind of personal identification method of living body faces multi-angle, referring to Fig. 1, method includes the following steps:
101: learning the local binary feature of each key point by cascading regression tree, and combine, returned using linear
Return detection key point, finally obtains 5 key points；
102: if the coordinate of 5 key points, which calculates, meets the first formula, judge the face that is positive, executes step 103, otherwise,
It prompts user to show positive face, until detecting positive face, executes step 103；
103: if key point coordinate when positive face and side face meets the second formula or third formula, judging a user's head left side
Turn or turn right, acquires user's left side of the face or right side face image；
104: the facial image of three the positive face of user, left side of the face, right side face angles is respectively fed to convolutional neural networks mould
Type exports the feature vector of 2622 dimensions respectively；Feature vector in this feature vector and face database does normalization dot product one by one, obtains
To similarity score list；
105: take one of highest scoring in similarity list as face recognition result, final face recognition result by
The recognition result of three angle faces, which is voted, to be generated, and poll is at most final recognition result.
Wherein, 5 key points in step 101 specifically: the left and right tail of the eye, nose, the left and right corners of the mouth.
When specific implementation, this method further include:
If the recognition result of three angle faces is different, using the face recognition result of positive face angle as final people
Face recognition result.
In conclusion this method has comprehensively considered the face characteristic of three angles, ballot obtains recognition result, effective to solve
Head pose variability bring of having determined misidentifies problem, enhances the stability and accuracy rate of recognition of face.
Embodiment 2
The scheme in embodiment 1 is carried out further below with reference to specific example, Fig. 2, Fig. 3 and calculation formula
It introduces, described below:
Step 1: image preprocessing
Homomorphic filtering processing is carried out to images to be recognized, adjusts image grayscale range, it is unbalanced to eliminate illumination on image
Problem enhances the image detail of dark space, while not losing the image detail in clear zone again.
Wherein, the concrete operations of the step are known to those skilled in the art, and the embodiment of the present invention does not repeat them here this.
Step 2: Face datection
It detects in pretreated image with the presence or absence of face.Using Haar feature and Adaboost cascade classifier to figure
As carrying out Face datection.Haar feature calculation is quick, and Adaboost integrates multiple Weak Classifiers and is combined into strong classifier, can be fast
Fast effective progress Face datection.
Step 3: face critical point detection
To the image after Face datection, tree method is returned using cascade, learns the local binary feature of each key point, so
Local binary feature is combined afterwards, detects key point using linear regression.
Wherein, the calibration of 68 key points of face is as shown in Figure 2.For ease of calculation, 5 people are taken from 68 key points
Face key point: the intermediate point for the two o'clock line that tail of the eye number is 39,36 in left eye is taken, point A is denoted as；The tail of the eye in right eye is taken to compile
Number for 42,45 two o'clock line intermediate point, be denoted as point B；The point that nose number is 33 is taken, point C is denoted as；Take the corners of the mouth number be
48,54 two o'clock is denoted as point D and point E.The coordinate representation of 5 key points of face is as shown in Figure 3.
Step 4: In vivo detection
In vivo detection is carried out according to the variation of face key point.Steps are as follows:
1) judge whether face is positive face, if the coordinate calculating of 5 key points of face meets formula (1), judgement is positive
Otherwise face prompts user to show positive face, until detecting positive face, after detecting positive face, acquisition user's face image is used for
The processing of 5th step.
Wherein, (ax, ay), (bx, by), (cx, cy), (dx, dy), the transverse and longitudinal that (ex, ey) is 5 key points of face are sat
Mark.
2) coordinate of 5 key points of face is A0 (ax0, ay0), B0 (bx0, by0), C0 when remembering positive face in step 1)
(cx0,cy0),D0(dx0,dy0),E0(ex0,ey0).The coordinate of 5 key points of face when side face be denoted as A1 (ax1, ay1),
B1(bx1,by1)、C1(cx1,cy1)、D1(dx1,dy1)、E1(ex1,ey1)。
Programming system prompts user's left-right rotation head, if face key point coordinate when calculating positive face and side face meets public affairs
Formula (2), then judge that user's head is turned left, and acquires the processing that user right face image is used for the 5th step at this time；Calculate positive face and side
Face key point coordinate when face judges that user's head is turned right if meeting formula (3), acquires face image on the left of user at this time and uses
In the processing of the 5th step.
Wherein, the ordinate of face key point A1 and B1 when ay1, by1 are side face, face when ax1, bx1, cx1 are side face
The abscissa of key point A1, B1, C1, ax0, bx0, cx0 be positive face when face key point A0, B0, C0 abscissa.
Step 5: positive face face identification
The facial image of the user front angle of 4th step acquisition is sent into convolutional neural networks model VGGFace, output
The feature vector of 2622 dimensions.
Feature vector in this feature vector and face database does normalization dot product one by one, obtains similarity score list, phase
It is [0,1] like property score value range, score is higher, indicates more similar.
One of highest scoring in similarity list is taken, remembers that its similarity score is Smax.If Smax is greater than given threshold
(empirical value 0.8) then regard this as recognition of face final result, no longer the 6th step of execution；If Smax is less than given threshold
(empirical value 0.8) then carries out the 6th step multi-orientation Face identification.
Since face image provides main identity information, when positive face similarity score is reliable enough, so that it may determine
Identity, it is not necessary to carry out the identification of side face.When positive face similarity score is less reliable, then comprehensively consider the face of multi-angle
Image.The algorithm speed of service can be improved in this way without losing accuracy.
Step 6: multi-orientation Face identification
The facial image of three the positive face of user of 4th step acquisition, left side of the face, right side face angles is respectively fed to convolution mind
Through network model VGGFace, the feature vector of 2622 dimension of output.
Feature vector in this feature vector and face database does normalization dot product one by one, obtains similarity score list, phase
It is [0,1] like property score value range, score is higher, indicates more similar.
One of highest scoring in similarity list is taken as face recognition result, the face of three angles respectively obtains three
A recognition result, final face recognition result are voted by the recognition result of these three angle faces and are generated, and poll is at most
Final recognition result.If there is extreme case: the recognition result of three angle faces is different, that is, the feelings of flat ticket
Condition, then using the face recognition result of positive face angle as final face recognition result.
In conclusion this method has comprehensively considered the face characteristic of three angles, ballot obtains recognition result, effective to solve
Head pose variability bring of having determined misidentifies problem, enhances the stability and accuracy rate of recognition of face.
Embodiment 3
Below with reference to Fig. 4-Fig. 7, the scheme in Examples 1 and 2 is further introduced, described below:
Step 1: image preprocessing
Homomorphic filtering processing is carried out to images to be recognized, adjusts image grayscale range, it is unbalanced to eliminate illumination on image
Problem enhances the image detail of dark space, while not losing the image detail in clear zone again.
Wherein, the concrete operations of the step are known to those skilled in the art, and the embodiment of the present invention does not repeat them here this.
Step 2: Face datection
It detects in pretreated image with the presence or absence of face.Using Haar feature and Adaboost cascade classifier to figure
As carrying out Face datection.Face datection uses image pyramid and sliding window technique, and image pyramid reduces ratio setting and exists
Between 1.1 to 1.5, minimum adjacent sliding window coupling number is arranged between 2 to 5, and smallest match is sized to (20,20).Detection
As a result the rectangle frame as marked in Fig. 4, Fig. 5, Fig. 6.
Step 3: face critical point detection
To the image after Face datection, tree method is returned using cascade and carries out face critical point detection.The people that will test
Face rectangle frame and original image, which are sent into, returns detector, and detector calibrates 68 key points of face, chooses from this 68 key points
5 representative face key points, 5 key points demarcated in testing result such as Fig. 4,5,6.
Step 4: In vivo detection
After calibrating face key point, need to carry out In vivo detection according to the variation of face key point.Steps are as follows:
1) acquisition user's face image is handled for the 5th step, and 5 face key point coordinates when recording positive face, such as figure
Shown in 4.
2) user's rotary head to the right detects the behavior of user's right-hand rotation head according to the relative positional relationship of key point, acquires user
Left side face image is for the processing of the 5th step, and 5 face key point coordinates when recording left side of the face, as shown in Figure 5.
3) user's rotary head to the left detects the behavior of user's left-hand rotation head according to the relative positional relationship of key point, acquires user
Right side face image for the 5th step processing, and record right side face when 5 face key point coordinates, as shown in Figure 6.
Step 5: positive face face identification
User's face image of 4th step acquisition is sent into convolutional neural networks model VGGFace, obtains similarity score
List, similarity score value range are [0,1], and score is higher, are indicated more similar.Positive face similarity score result is shown in Fig. 7 institute
Show the third column of table.The similarity score to rank the first is less than empirical value 0.8, carries out the 6th step multi-orientation Face identity and knows
Not.
Step 6: multi-orientation Face identification
The facial image of user's left side of the face of 4th step acquisition, right side face angle is respectively fed to convolutional neural networks model
VGGFace, obtains similarity score list, the secondary series of left side of the face similarity score result table as shown in Figure 7, right side face phase
Like property scores as shown in Figure 7 table the 4th column.Final face recognition result is by three angle face similarity scores
The result ballot to rank the first generates.
Bibliography
[1]Sun Y,Wang X,Tang X.Deep Learning Face Representation from
Predicting 10,000Classes[C]//Computer Vision and Pattern Recognition.IEEE,
2014:1891-1898.
[2]Taigman Y,Yang M,Ranzato M,et al.DeepFace:Closing the Gap to
Human-Level Performance in Face Verification[C]//IEEE Conference on Computer
Vision and Pattern Recognition.IEEE Computer Society,2014:1701-1708.
[3]Parkhi O M,Vedaldi A,Zisserman A.Deep face recognition[C]//
BMVC.2015,1(3):6.
[4]Schroff F,Kalenichenko D,Philbin J.FaceNet:A unified embedding for
face recognition and clustering[J].2015:815-823.
It will be appreciated by those skilled in the art that attached drawing is the schematic diagram of a preferred embodiment, the embodiments of the present invention
Serial number is for illustration only, does not represent the advantages or disadvantages of the embodiments.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and
Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (6)
1. a kind of personal identification method of living body faces multi-angle, which is characterized in that the described method comprises the following steps:
Learn the local binary feature of each key point by cascading regression tree, and combine, is detected and closed using linear regression
Key point finally obtains 5 key points；
If the coordinate of 5 key points, which calculates, meets the first formula, the face that is positive is judged；
If key point coordinate when positive face and side face meets the second formula or third formula, user's head left-hand rotation or right is judged
Turn, acquires user's left side of the face or right side face image；
The facial image of three the positive face of user, left side of the face, right side face angles is respectively fed to convolutional neural networks model, it is defeated respectively
The feature vector of 2622 dimensions out；Feature vector in this feature vector and face database does normalization dot product one by one, obtains similitude
Score list；
One of highest scoring in similarity list is taken as face recognition result, final face recognition result is by three angles
The recognition result of face, which is voted, to be generated, and poll is at most final recognition result.
2. a kind of personal identification method of living body faces multi-angle according to claim 1, which is characterized in that described 5
Key point specifically: the left and right tail of the eye, nose, the left and right corners of the mouth.
3. a kind of personal identification method of living body faces multi-angle according to claim 1, which is characterized in that described first
Formula specifically:
Wherein, (ax, ay), (bx, by), (cx, cy), (dx, dy), the transverse and longitudinal coordinate that (ex, ey) is 5 key points of face.
4. a kind of personal identification method of living body faces multi-angle according to claim 1, which is characterized in that described second
Formula specifically:
Wherein, the ordinate of face key point A1 and B1 when ay1, by1 are side face, face is crucial when ax1, bx1, cx1 are side face
The abscissa of point A1, B1, C1, ax0, bx0, cx0 be positive face when face key point A0, B0, C0 abscissa.
5. a kind of personal identification method of living body faces multi-angle according to claim 4, which is characterized in that the third
Formula specifically:
。
6. a kind of personal identification method of living body faces multi-angle described in any claim in -5 according to claim 1,
It is characterized in that, the method also includes:
If the recognition result of three angle faces is different, the face recognition result of positive face angle is known as final face
Other result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811537149.8A CN109800643B (en) | 2018-12-14 | 2018-12-14 | Identity recognition method for living human face in multiple angles |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811537149.8A CN109800643B (en) | 2018-12-14 | 2018-12-14 | Identity recognition method for living human face in multiple angles |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109800643A true CN109800643A (en) | 2019-05-24 |
| CN109800643B CN109800643B (en) | 2023-03-31 |
Family
ID=66556870
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811537149.8A Active CN109800643B (en) | 2018-12-14 | 2018-12-14 | Identity recognition method for living human face in multiple angles |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109800643B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110363091A (en) * | 2019-06-18 | 2019-10-22 | 广州杰赛科技股份有限公司 | Face identification method, device, equipment and storage medium in the case of side face |
| CN110443213A (en) * | 2019-08-12 | 2019-11-12 | 北京比特大陆科技有限公司 | Type of face detection method, object detection method and device |
| CN111680608A (en) * | 2020-06-03 | 2020-09-18 | 长春博立电子科技有限公司 | Intelligent sports auxiliary training system and training method based on video analysis |
| CN111931677A (en) * | 2020-08-19 | 2020-11-13 | 北京影谱科技股份有限公司 | Face detection method and device and face expression detection method and device |
| CN112149559A (en) * | 2020-09-22 | 2020-12-29 | 沈澈 | Face recognition method and device, readable storage medium and computer equipment |
| CN112307817A (en) * | 2019-07-29 | 2021-02-02 | 中国移动通信集团浙江有限公司 | Face living body detection method and device, computing equipment and computer storage medium |
| CN112801066A (en) * | 2021-04-12 | 2021-05-14 | 北京圣点云信息技术有限公司 | Identity recognition method and device based on multi-posture facial veins |
| CN113191322A (en) * | 2021-05-24 | 2021-07-30 | 口碑(上海)信息技术有限公司 | Method and device for detecting skin of human face, storage medium and computer equipment |
| CN113837009A (en) * | 2021-08-26 | 2021-12-24 | 张大艳 | Internet of things data acquisition and analysis system based on artificial intelligence |
| CN114120376A (en) * | 2021-11-18 | 2022-03-01 | 黑龙江大学 | Multi-mode image acquisition device and system |
| US20220343617A1 (en) * | 2019-09-12 | 2022-10-27 | Nec Corporation | Image analysis device, control method, and program |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011215843A (en) * | 2010-03-31 | 2011-10-27 | Secom Co Ltd | Facial image processing apparatus |
| CN106096560A (en) * | 2016-06-15 | 2016-11-09 | 广州尚云在线科技有限公司 | A kind of face alignment method |
| CN106503687A (en) * | 2016-11-09 | 2017-03-15 | 合肥工业大学 | The monitor video system for identifying figures of fusion face multi-angle feature and its method |
| CN106875191A (en) * | 2017-02-27 | 2017-06-20 | 努比亚技术有限公司 | One kind scanning payment processing method, device and terminal |
| WO2017106996A1 (en) * | 2015-12-21 | 2017-06-29 | 厦门中控生物识别信息技术有限公司 | Human facial recognition method and human facial recognition device |
| CN107480658A (en) * | 2017-09-19 | 2017-12-15 | 苏州大学 | Face identification device and method based on multi-angle video |
-
2018
- 2018-12-14 CN CN201811537149.8A patent/CN109800643B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011215843A (en) * | 2010-03-31 | 2011-10-27 | Secom Co Ltd | Facial image processing apparatus |
| WO2017106996A1 (en) * | 2015-12-21 | 2017-06-29 | 厦门中控生物识别信息技术有限公司 | Human facial recognition method and human facial recognition device |
| CN106096560A (en) * | 2016-06-15 | 2016-11-09 | 广州尚云在线科技有限公司 | A kind of face alignment method |
| CN106503687A (en) * | 2016-11-09 | 2017-03-15 | 合肥工业大学 | The monitor video system for identifying figures of fusion face multi-angle feature and its method |
| CN106875191A (en) * | 2017-02-27 | 2017-06-20 | 努比亚技术有限公司 | One kind scanning payment processing method, device and terminal |
| CN107480658A (en) * | 2017-09-19 | 2017-12-15 | 苏州大学 | Face identification device and method based on multi-angle video |
Non-Patent Citations (2)
| Title |
|---|
| GAYATHRI MAHALINGAM，ET AL: "Investigating the Periocular-Based Face Recognition Across Gender Transformation", 《IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY》 * |
| 周志华，等: "基于神经网络集成的多视角人脸识别", 《计算机研究与发展》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110363091A (en) * | 2019-06-18 | 2019-10-22 | 广州杰赛科技股份有限公司 | Face identification method, device, equipment and storage medium in the case of side face |
| CN110363091B (en) * | 2019-06-18 | 2021-08-10 | 广州杰赛科技股份有限公司 | Face recognition method, device and equipment under side face condition and storage medium |
| CN112307817A (en) * | 2019-07-29 | 2021-02-02 | 中国移动通信集团浙江有限公司 | Face living body detection method and device, computing equipment and computer storage medium |
| CN112307817B (en) * | 2019-07-29 | 2024-03-19 | 中国移动通信集团浙江有限公司 | Face living body detection method, device, computing equipment and computer storage medium |
| CN110443213A (en) * | 2019-08-12 | 2019-11-12 | 北京比特大陆科技有限公司 | Type of face detection method, object detection method and device |
| US20220343617A1 (en) * | 2019-09-12 | 2022-10-27 | Nec Corporation | Image analysis device, control method, and program |
| CN111680608A (en) * | 2020-06-03 | 2020-09-18 | 长春博立电子科技有限公司 | Intelligent sports auxiliary training system and training method based on video analysis |
| CN111680608B (en) * | 2020-06-03 | 2023-08-18 | 长春博立电子科技有限公司 | Intelligent sports auxiliary training system and training method based on video analysis |
| CN111931677A (en) * | 2020-08-19 | 2020-11-13 | 北京影谱科技股份有限公司 | Face detection method and device and face expression detection method and device |
| CN112149559A (en) * | 2020-09-22 | 2020-12-29 | 沈澈 | Face recognition method and device, readable storage medium and computer equipment |
| CN112801066A (en) * | 2021-04-12 | 2021-05-14 | 北京圣点云信息技术有限公司 | Identity recognition method and device based on multi-posture facial veins |
| CN112801066B (en) * | 2021-04-12 | 2022-05-17 | 北京圣点云信息技术有限公司 | Identity recognition method and device based on multi-posture facial veins |
| CN113191322A (en) * | 2021-05-24 | 2021-07-30 | 口碑(上海)信息技术有限公司 | Method and device for detecting skin of human face, storage medium and computer equipment |
| CN113837009A (en) * | 2021-08-26 | 2021-12-24 | 张大艳 | Internet of things data acquisition and analysis system based on artificial intelligence |
| CN114120376A (en) * | 2021-11-18 | 2022-03-01 | 黑龙江大学 | Multi-mode image acquisition device and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109800643B (en) | 2023-03-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109800643A (en) | A kind of personal identification method of living body faces multi-angle | |
| Alotaibi et al. | Improved gait recognition based on specialized deep convolutional neural network | |
| Balcoh et al. | Algorithm for efficient attendance management: Face recognition based approach | |
| US8320643B2 (en) | Face authentication device | |
| El Khiyari et al. | Age invariant face recognition using convolutional neural networks and set distances | |
| Jaafar et al. | A review of multibiometric system with fusion strategies and weighting factor | |
| CN105760815A (en) | Heterogeneous human face verification method based on portrait on second-generation identity card and video portrait | |
| Anthony et al. | A review of face anti-spoofing methods for face recognition systems | |
| CN111767877A (en) | Living body detection method based on infrared features | |
| Deriche | Trends and challenges in mono and multi biometrics | |
| Bouhabba et al. | Support vector machine for face emotion detection on real time basis | |
| Sudhakar et al. | Facial identification of twins based on fusion score method | |
| Yin et al. | Fusion of face recognition and facial expression detection for authentication: a proposed model | |
| Mohamed et al. | Automated face recogntion system: Multi-input databases | |
| Rose et al. | Deep learning based estimation of facial attributes on challenging mobile phone face datasets | |
| Sarode et al. | Review of iris recognition: an evolving biometrics identification technology | |
| Méndez-Llanes et al. | On the use of local fixations and quality measures for deep face recognition | |
| Jahromi et al. | Automatic access control based on face and hand biometrics in a non-cooperative context | |
| Singla et al. | Face Detection and Recognition Using Digital Image Processing:" State of the Art" | |
| Tan et al. | Face recognition algorithm based on open CV | |
| Muthukumaran et al. | Face and Iris based Human Authentication using Deep Learning | |
| Drosou et al. | Event-based unobtrusive authentication using multi-view image sequences | |
| Wei | Unconstrained face recognition with occlusions | |
| Gavkare et al. | Face recognition smart attendance system-a survey | |
| Gavrilova et al. | Fusing multiple matcher's outputs for secure human identification |
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 |