CN103824050B - A kind of face key independent positioning method returned based on cascade - Google Patents
A kind of face key independent positioning method returned based on cascade Download PDFInfo
- Publication number
- CN103824050B CN103824050B CN201410053323.7A CN201410053323A CN103824050B CN 103824050 B CN103824050 B CN 103824050B CN 201410053323 A CN201410053323 A CN 201410053323A CN 103824050 B CN103824050 B CN 103824050B
- Authority
- CN
- China
- Prior art keywords
- face
- cascade
- essence
- recurrence
- key point
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of based on the face key independent positioning method for cascading recurrence, its step includes:1)A large amount of face picture data are gathered, and marks initial key point position;2)By being trained to face picture data, study obtains robust regression device, and then using the output of robust regression device as input, study obtains essence and returns device;3)Given face picture data to be identified, are revert to the original shape of face near true shape by robust regression device, then using the output of robust regression device as input, return the accurate coordinates that device obtains face key point by essence.Proposed by the present invention by slightly to essence cascade homing method, by learning to great amount of samples, and multiple features fusion, the mode for returning device fusion more, drastically increase the speed and robustness of algorithm, blocking, face key point location is carried out under the attitude such as light difference and side face and all achieve extraordinary effect, can effectively increase the accuracy and speed of face key point location.
Description
Technical field
The invention belongs to Digital Image Processing and technical field of face recognition, and in particular to a kind of people returned based on cascade
Face key independent positioning method.
Background technology
Face key point is the strong key point of some sign abilities of face, such as eyes, nose, face and face mask
Deng.Key point is positioned at field of face identification to be played the role of critically important, and such as recognition of face, tracking, Expression analysis and 3D build
Mould all relies on the result of crucial point location.
Traditional face key independent positioning method, is the method based on parametric shape model, according to the table near key point
Feature is seen, is learnt a parameter model, is iteratively optimized the position of key point in use, finally obtain crucial point coordinates.
Above-mentioned face key independent positioning method, all depends critically upon the shooting quality of picture, human face posture.Blocking,
When illumination and larger attitudes vibration, accurate result can not be all obtained.Simultaneously as at present mobile terminal demand is developed rapidly,
Said method can not realize real-time processing by mobile terminals such as mobile phones.
Content of the invention
Existing face key independent positioning method to attitude, block, light very sensitive, accuracy and speed is poor,
It is difficult to accomplish real-time processing on the mobile terminals such as the mobile phone that image-forming condition is poor and computing resource is limited.The present invention proposes a kind of fast
Speed, accurate by thick to essence cascade homing method, by being learnt to great amount of samples and multiple features fusion, many recurrence devices melt
Close, solve the key point orientation problem in recognition of face well.
The technical solution used in the present invention is as follows:
A kind of face key independent positioning method returned based on cascade, its step are included:
1)A large amount of face picture data are gathered, and marks initial key point position(Handmarking can be passed through);
2)By being trained to a large amount of face picture data, study obtains a robust regression device, then with described thick
The output of device is returned as input, study obtains an essence and returns device, so as to obtain device being returned by the cascade slightly to essence;
3)Face picture to be detected and corresponding face location is given, by the robust regression device by the initial shape of face
Shape is revert near true shape, then using the output of the robust regression device as input, is returned device by the essence and is obtained people
The accurate coordinates of face key point.
Further, the robust regression device is designed to linear regressor, at all key points extracts SURF features.This time
Return device represent the relation between 3D attitudes and SURF features.
Further, linear regressor of the robust regression device comprising multi-stage cascade, it is preferred to use two-stage linear regressor,
Input of the output of the first order as the second level.The robust regression device consisted of this two-stage linear regressor, can obtain one
Rough key point position and 3D attitudes.
Further, the essence returns device using the output of robust regression device above as input, is cascaded back using random fern
Return device, using pixel value difference as feature.Device is returned by essence, the rough result that robust regression device is provided an essence can be returned into
True result.
Further, it is a double-layer structure that the essence returns device, and ground floor is a series of weak recurrence device { f1,f2,…,
ftCascade;The second layer, is a series of cascade of random ferns recurrence devices, constitutes a weak recurrence device f.
Further, the face key point includes the positions such as eyes, nose, face, face mask, more specifically, such as
The positions such as pupil, canthus, eyebrow angle, the corners of the mouth, lip edge.
Propose in the present invention a kind of by slightly to essence cascade regression algorithm, devise returning one more device fusion cascade
Return device.The cascade returns device and is divided into two parts:1. robust regression device, feature are that speed is fast, can quickly revert to the vicinity of normal solution;
2. essence returns device, and feature is that each amount for returning is less, but can obtain more accurately result.According to the recurrence device for designing
Feature, allows the different devices that return to complete different tasks(Linear regressor and cascade random fern return device), merged multiple spies
Levy(SURF and margin of image element feature).
Proposed by the present invention by slightly to the cascade regression algorithm of essence, by learning to great amount of samples, and multiple features
Fusion, the mode for returning device fusion, greatly improve the speed and robustness of algorithm more, blocking, light difference and side face etc.
Face key point location is carried out under attitude and all achieves extraordinary effect, can effectively increase the essence of face key point location
Degree and speed, hence it is evident that better than other algorithms existing.
Description of the drawings
The step of Fig. 1 is the independent positioning method crucial based on the face for cascading recurrence of present invention flow chart.
Fig. 2 is that the cascade of the present invention returns device schematic diagram.
Fig. 3 is to return the schematic diagram that initial value is revert to device true solution using cascade.
Fig. 4 is using the schematic diagram for carrying out face key point location using the method for the present invention.
Specific embodiment
Below by specific embodiments and the drawings, the present invention will be further described.
The face key independent positioning method returned based on cascade of the present invention, its steps flow chart is as shown in figure 1, mainly include
Two parts content, one is to set up to return the cascade recurrence device that device part constitutes by robust regression device part and essence, and two is using foundation
Cascade return device face image data is processed to recognize key point.
1. set up and device is returned by the cascade slightly to essence
The general frame of the present invention is that a cascade returns device.Our target is one regression function f of study, enables it to
Enough solution space is mapped to from initial sample space, enable to mean square deviation minimum.Run into the linear pass of higher dimensional space and complexity
When being, if simply learning a recurrence device to express this mapping relations unrealistic.Then, we have proposed using cascade
Method, by cascading multiple weak recurrence devices, they are constituted a higher strong recurrence device of regression capability.The present invention is adopted
Cascade homing method, the cascade { f that regression function f is divided into t simple regression function1,f2,…,ft, per one-level fk's
Input is all its previous stage fk‐1Output, as shown in Fig. 2 pass through f1,f2,…,ftCombine, the regression function energy for obtaining
Enough approximately go out original shape to the Nonlinear Mapping relation of the complexity of true shape.
The present invention return device follow by slightly to essence process, cascade recurrence device be divided into two parts, robust regression device and essence
Return device.
If simply according to above method, cascaded with several weak recurrence devices using simple, effect is paid no attention to first
Think, because the shooting condition of picture varies, attitude is different, shape to be returned also all is not quite similar, obtain perfect
Effect is too high to the requirement for returning device.Secondly, if cascade series is excessive, speed also can be very slow, can not meet to speed
Require.Innovatively propose in the present invention and mutually cascaded using different types of recurrence device, be allowed to that Each performs its own functions, mutually promote, raise
Length is kept away short.
Therefore, the recurrence device of cascade is divided into two parts by us, and Part I is robust regression device, and initial value is revert to very
The vicinity of real solution, completes big regressive object, but is indifferent to details.This part, the coarse regressive object for completing, speed are non-
Often fast, it is that Part II generates input.Part II returns device for essence, it is only necessary to be adjusted in detail, progressively to true
Solution is slowly approached, and whole process is as shown in Figure 3.Two parts, constitute one and return device by the cascade slightly to essence, in speed and
In effect, there is very big lifting.
For two-part different qualities, the present invention devises different graders and feature, can complete in maximum efficiency
Regressive object.
The target of Part I is to be quickly obtained coarse solution, and, using SURF features, study is out linearly returned for one for we
Return device, this part returns device, and rapidly initial value can be mapped near normal solution.Specific implementation step is as follows:
1. initial SURF features are extracted at each key point on original shape, is denoted as Φ0, true regressive object is designated as
ΔX*;
2. in the training process, as true shape X is, it is known that initial value X0It is known, then true regressive object
Δ X* is, it is known that Δ X*=X X0.Linear regressor can be expressed as Δ X0=R0*Φ0+b0, target allows recurrence to obtain
Estimator Δ X0With true regressive object Δ X* infinite approachs.Parameter required herein is exactly R0And b0.Can pass through to minimize following formula
Try to achieve:
Wherein, diFor i-th face picture, X0 iFor the original shape of i-th face, Δ X* iFor the true of i-th face
Look back target, Φ0 iIt is i-th face in original shape X0 iThe SURF characteristic vectors at place, this are the most young waiters in a wineshop or an inn of solution familiar to us
Problem is taken advantage of, R can be readily obtained0And b0.
3. according to the R for obtaining0And b0, just can obtain increment Delta X that estimates0=R0*Φ1+b0, X+ Δ X0As new instruction
Practice collection, be designated as X1.According to new training set, new SURF features Φ are extracted1, have Δ X1=R1*Φ1+b1, in the same manner, according to above-mentioned
Method, can readily try to achieve R1And b1.By that analogy, a lot of similar linear regressors can be learnt, in Part I, I
Learn two-layer linear regressor just much of that, solution X of estimation2Very close to true solution X.Part I obtain coarse solution it
Afterwards, as the input of Part II, remaining fine regressive object is given below to do.
Part II, the random fern that present invention employs cascade return device, and pixel value difference is used as feature.We are by first
Input of the output for dividing as this part, this value is apart from true solution very close to be done is in detail
Have adjusted, its Step wise approximation is truly solved.
Random fern returns the combination that device is 5 features and threshold value, and training sample is divided into 25Individual space.Each space
Corresponding output Δ Xbin, Δ XbinFor being divided into the mean value of all regressive objects in the space.
The random fern of the cascade of Part II returns device, is the recurrence device of a two-layer.Because if simply this time
Device is returned to be designed to the cascade that original random fern returns device, regression capability is too weak, so being designed into one two in the present invention
The structure of layer.Specific as follows, multiple original random ferns return device cascade, constitute a weak recurrence device f.Again by these weak recurrence devices
{f1,f2,…,ftCascade constitutes a strong recurrence device, that is, Part II essence mentioned above returns device.
Specific implementation step is as follows:
1. the pixel value difference feature of each sample is extracted:Take two key points at random, generate an interpolation coefficient at random, obtain
A position in 2 points of lines, the pixel value difference on two such positions is used as feature.In the present invention, extract altogether
The feature of 400 points.
2. selected characteristic:The feature of 400 points has been generated above, and one has 160000 combination of two.The present invention
Middle random fern returns 5 stack features used in device, in so big feature space, will select 5 groups out.Method is as follows,
A random column vector is firstly generated, true regressive object matrix is mapped on a direction, each is then calculated respectively
Characteristic vector and the coefficient correlation of this projection vector, choose coefficient correlation maximum 5 groups.
3. the weak generation for returning device:According to the feature that previous step is extracted, sample can be divided into original random fern and be returned
In certain space of device.Calculate average true shape increment Delta X of all samples in the spacebin, it is added in current spatial
Each estimation in shape, obtain new estimation shape.The estimation shape for obtaining is returned device as next original random fern
Input, pass to next original random fern and return device, keep feature invariant, obtain new random fern and return device.By such 10
Individual original random fern returns device cascade and constitutes a weak recurrence device.
4. the generation of device is returned by force:Through above-mentioned steps, learn to weak recurrence device fk, for an initial set Xk, can
To pass through fkObtain regression deltas and estimate Δ Xk, new initial set can be by calculating Xk+ΔXkObtain, in new estimation shape
On the basis of extract new feature, obtain next weak recurrence device according to the method described above, as shown in Fig. 2 by that analogy, in this reality
Apply in example, cascade 100 weak recurrence device { f1,f2,…,f100, constitute one two layers of strong recurrence device.
So far, obtain the complete cascade by slightly to essence and return device.
2. return device using cascade to process face image data, to recognize key point
This cascade recurrence device by slightly to essence in the present invention can be taken when face key point orientation problem is solved
Obtain extraordinary effect.
Specifically, in face key point location, an original shape is given(People can be snapped to by average shape
Lian Kuang centers), extract SURF operators in each key point as characteristic vector, by Part I, obtain one rough
As a result, as the original shape of Part II, in the position of key point, pixel value difference feature is extracted, is used step by step, most
After obtain accurate shape.Whole cascade recurrence device is followed and adjusts big attitude, such as by slightly to the process of essence, Part I
Side face, pitching, rotation angularly, face size, translational movement etc..The essence of Part II returns device, adjusts the details of shape, such as
Face shape, eye shape, eyebrow shape etc..Two parts, have reached into one and have returned device by the cascade slightly to essence, in speed and
In effect, there is very big lifting.
Fig. 4 is to return device using cascade to carry out the schematic diagram of face key point location, and wherein (a) is init state, (b)
Through the face key point position that Part I robust regression device is obtained, it is (c) to return the face that device is obtained through Part II essence
Key point position.As can be seen that adopting the method for the present invention, through Part I robust regression device, face key point is by initial value
It revert to vicinity of true solution, and then returns device and approach through Part II essence and truly solve, it was demonstrated that the method for the present invention can
To obtain good effect in terms of face key point location.
The method of the present invention, processes a face, in the computer of Intel (R) Core (TM) i3-4130CPU@3.4GHz
On, the used time is 8ms or so, speed be before method several times.On LFPW data sets, mean error is about the present invention
0.036(The mean error of each point is divided by interocular distance), reached industry advanced level.
Above example only in order to technical scheme to be described rather than be limited, the ordinary skill of this area
Personnel can modify to technical scheme or equivalent, without departing from the spirit and scope of the present invention, this
The protection domain of invention should be to be defined described in claim.
Claims (7)
1. a kind of based on the face key independent positioning method for cascading recurrence, its step includes:
1) a large amount of face picture data are gathered, and marks initial key point position;
2) by being trained to a large amount of face picture data, study obtains a robust regression device, then with the robust regression
Used as input, study obtains an essence and returns device for the output of device, so as to obtain returning device by the cascade slightly to essence;The robust regression device
Using linear regressor, at all key points, extract SURF features;The robust regression device is the linear regressor of a cascade,
Input of the output of previous stage as rear stage;The linear regressor is obtained using SURF feature learnings, concrete steps include:
1. initial SURF features are extracted at each key point on original shape, is denoted as Φ0, true regressive object is designated as Δ X*;
2. in the training process, due to crucial point coordinates X, initial value key point coordinates X0Known, then key point truly returns mesh
Mark Δ X*As, it is known that Δ X*=X-X0;Linear regressor is expressed as Δ X0=R0*Φ0+b0, parameter R therein0And b0By most
Littleization following formula is tried to achieve:
Wherein, diFor i-th face picture,For the original shape of i-th face,True review mesh for i-th face
Mark,It is i-th face in original shapeThe SURF characteristic vectors at place;
3. according to the R for obtaining0And b0, obtain increment Delta X that estimates0=R0*Φ1+b0, X+ Δ X0As new training set, it is designated as
X1;According to new training set, new SURF features Φ are extracted1, have Δ X1=R1*Φ1+b1, in the same manner, R is tried to achieve according to said method1
And b1;By that analogy, obtain multistage linear and return device;
3) face picture to be identified and corresponding face location is given, the original shape of face is returned by the robust regression device
It is grouped near true shape, then using the output of the robust regression device as input, device is returned by the essence and obtains face pass
The accurate coordinates of key point.
2. the method for claim 1, it is characterised in that:The robust regression device includes two-stage linear regressor.
3. the method for claim 1, it is characterised in that:The essence returns device and returns device using random fern cascade, with picture
Plain difference is used as feature.
4. method as claimed in claim 3, it is characterised in that:It is a double-layer structure that the essence returns device, and ground floor is one
The weak cascade for returning device of series;The second layer is a series of cascade that random ferns return device, constitutes a weak recurrence device.
5. method as claimed in claim 4, it is characterised in that:The step of essence recurrence device for generating the double-layer structure, includes:
1. the pixel value difference feature of each sample is extracted:Take two key points at random, generate an interpolation coefficient at random, obtain two
A position in point line, the pixel value difference on two such positions is used as feature;
2. selected characteristic:A random column vector is generated, true regressive object matrix is mapped on a direction, Ran Houfen
The coefficient correlation of each characteristic vector and this projection vector is not calculated, and device is returned using random fern and is chosen coefficient correlation maximum
Many stack features;
3. the weak generation for returning device:According to the feature that previous step is extracted, sample be divided into original random fern return device certain
In space, average true shape increment Delta X of all samples in the space is calculatedbin, each being added in current spatial estimate
Meter in shape, obtains new estimation shape, the input that the estimation shape for obtaining is returned device as next original random fern, passes
Device is returned to next original random fern, feature invariant is kept, new random fern is obtained and is returned device, multiple original random ferns are returned
Device cascade is returned to constitute a weak recurrence device;
4. the generation of device is returned by force:Through above-mentioned steps, learn to weak recurrence device fk, for an initial set Xk, by fk
Obtain regression deltas and estimate Δ Xk, new initial set is by calculating Xk+ΔXkObtain, extract on the basis of new estimation shape new
Feature, obtain next weak recurrence device according to the method described above, by that analogy, cascade multiple weak recurrence devices, constitute one two layers
Strong recurrence device.
6. method as claimed in claim 5, it is characterised in that:The essence returns the original of the weak recurrence device comprising 10 cascades of device
Beginning random fern returns device, and the essence returns the described weak recurrence device of the strong recurrence device comprising 100 cascades of device.
7. the method for claim 1, it is characterised in that:The face key point includes eyes, nose, face, face
The position of profile.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410053323.7A CN103824050B (en) | 2014-02-17 | 2014-02-17 | A kind of face key independent positioning method returned based on cascade |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410053323.7A CN103824050B (en) | 2014-02-17 | 2014-02-17 | A kind of face key independent positioning method returned based on cascade |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103824050A CN103824050A (en) | 2014-05-28 |
CN103824050B true CN103824050B (en) | 2017-03-15 |
Family
ID=50759102
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410053323.7A Active CN103824050B (en) | 2014-02-17 | 2014-02-17 | A kind of face key independent positioning method returned based on cascade |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103824050B (en) |
Families Citing this family (36)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2015180100A1 (en) * | 2014-05-29 | 2015-12-03 | Beijing Kuangshi Technology Co., Ltd. | Facial landmark localization using coarse-to-fine cascaded neural networks |
CN104361358A (en) * | 2014-11-21 | 2015-02-18 | 江苏刻维科技信息有限公司 | Automatic facial feature and facial contour identification method |
CN104361362A (en) * | 2014-11-21 | 2015-02-18 | 江苏刻维科技信息有限公司 | Method for obtaining locating model of human face outline |
CN105844206A (en) * | 2015-01-15 | 2016-08-10 | 北京市商汤科技开发有限公司 | Identity authentication method and identity authentication device |
CN105844284B (en) * | 2015-01-16 | 2019-07-09 | 阿里巴巴集团控股有限公司 | Man face characteristic point positioning method and device |
CN105868769A (en) * | 2015-01-23 | 2016-08-17 | 阿里巴巴集团控股有限公司 | Method and device for positioning face key points in image |
CN104615996B (en) * | 2015-02-15 | 2017-11-21 | 四川川大智胜软件股份有限公司 | A kind of various visual angles two-dimension human face automatic positioning method for characteristic point |
CN107615295B (en) * | 2015-05-21 | 2020-09-25 | 北京市商汤科技开发有限公司 | Apparatus and method for locating key features of face image |
CN106529374A (en) * | 2015-09-10 | 2017-03-22 | 大唐电信科技股份有限公司 | Cascaded face key point positioning method and system |
CN105469056A (en) * | 2015-11-26 | 2016-04-06 | 小米科技有限责任公司 | Face image processing method and device |
CN105678248B (en) * | 2015-12-31 | 2018-11-27 | 上海科技大学 | Face key point alignment algorithm based on deep learning |
CN105469081B (en) * | 2016-01-15 | 2019-03-22 | 成都品果科技有限公司 | A kind of face key independent positioning method and system for U.S. face |
CN107016319B (en) * | 2016-01-27 | 2021-03-05 | 北京三星通信技术研究有限公司 | Feature point positioning method and device |
CN107016323A (en) * | 2016-01-28 | 2017-08-04 | 厦门中控生物识别信息技术有限公司 | A kind of localization method and device of palm area-of-interest |
CN106056080B (en) * | 2016-05-30 | 2019-11-22 | 中控智慧科技股份有限公司 | A kind of visual biometric information acquisition device and method |
CN105956581B (en) * | 2016-06-08 | 2019-08-20 | 华南理工大学 | A kind of quick human face characteristic point initial method |
CN106200657B (en) * | 2016-07-09 | 2018-12-07 | 东莞市华睿电子科技有限公司 | A kind of unmanned aerial vehicle (UAV) control method |
CN106157372B (en) * | 2016-07-25 | 2019-01-25 | 深圳市唯特视科技有限公司 | A kind of 3D face grid reconstruction method based on video image |
CN107766867A (en) * | 2016-08-15 | 2018-03-06 | 佳能株式会社 | Object shapes detection means and method, image processing apparatus and system, monitoring system |
CN106650622B (en) * | 2016-11-18 | 2019-06-25 | 华南农业大学 | Working state of excavator recognition methods based on mixing LBF shape regression model |
CN108133221B (en) * | 2016-12-01 | 2022-06-07 | 佳能株式会社 | Object shape detection device, image processing device, object shape detection method, and monitoring system |
CN106845520B (en) * | 2016-12-23 | 2018-05-18 | 深圳云天励飞技术有限公司 | A kind of image processing method and terminal |
CN107274508A (en) * | 2017-07-26 | 2017-10-20 | 南京多伦科技股份有限公司 | A kind of vehicle-mounted timing have the records of distance by the log terminal and using the terminal recognition methods |
CN107644203B (en) * | 2017-09-12 | 2020-08-28 | 江南大学 | Feature point detection method for shape adaptive classification |
CN107729827A (en) * | 2017-09-29 | 2018-02-23 | 北京飞搜科技有限公司 | A kind of man face characteristic point positioning method and device |
CN108875492B (en) * | 2017-10-11 | 2020-12-22 | 北京旷视科技有限公司 | Face detection and key point positioning method, device, system and storage medium |
CN107766851A (en) * | 2017-12-06 | 2018-03-06 | 北京搜狐新媒体信息技术有限公司 | A kind of face key independent positioning method and positioner |
CN108196535B (en) * | 2017-12-12 | 2021-09-07 | 清华大学苏州汽车研究院(吴江) | Automatic driving system based on reinforcement learning and multi-sensor fusion |
CN108062545B (en) * | 2018-01-30 | 2020-08-28 | 北京搜狐新媒体信息技术有限公司 | Face alignment method and device |
CN108509862B (en) * | 2018-03-09 | 2022-03-25 | 华南理工大学 | Rapid face recognition method capable of resisting angle and shielding interference |
CN109063584B (en) * | 2018-07-11 | 2022-02-22 | 深圳大学 | Facial feature point positioning method, device, equipment and medium based on cascade regression |
CN109740426B (en) * | 2018-11-23 | 2020-11-06 | 成都品果科技有限公司 | Face key point detection method based on sampling convolution |
CN109934129B (en) * | 2019-02-27 | 2023-05-30 | 嘉兴学院 | Face feature point positioning method, device, computer equipment and storage medium |
CN110175558B (en) * | 2019-05-24 | 2021-02-05 | 北京达佳互联信息技术有限公司 | Face key point detection method and device, computing equipment and storage medium |
EP3973449A4 (en) * | 2019-06-11 | 2022-08-03 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | Method and system for facial landmark detection using facial component-specific local refinement |
CN110705437A (en) * | 2019-09-26 | 2020-01-17 | 中国科学技术大学 | Face key point detection method and system based on dynamic cascade regression |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102262724A (en) * | 2010-05-31 | 2011-11-30 | 汉王科技股份有限公司 | Object image characteristic points positioning method and object image characteristic points positioning system |
WO2013048160A1 (en) * | 2011-09-27 | 2013-04-04 | (주)올라웍스 | Face recognition method, apparatus, and computer-readable recording medium for executing the method |
-
2014
- 2014-02-17 CN CN201410053323.7A patent/CN103824050B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102262724A (en) * | 2010-05-31 | 2011-11-30 | 汉王科技股份有限公司 | Object image characteristic points positioning method and object image characteristic points positioning system |
WO2013048160A1 (en) * | 2011-09-27 | 2013-04-04 | (주)올라웍스 | Face recognition method, apparatus, and computer-readable recording medium for executing the method |
Non-Patent Citations (1)
Title |
---|
一个基于多层结构的快速眼睛定位算法;车昊等;《中国图象图形学报》;20080331;第13卷(第3期);第472-479页 * |
Also Published As
Publication number | Publication date |
---|---|
CN103824050A (en) | 2014-05-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103824050B (en) | A kind of face key independent positioning method returned based on cascade | |
CN103824089B (en) | Cascade regression-based face 3D pose recognition method | |
US11182644B2 (en) | Method and apparatus for pose planar constraining on the basis of planar feature extraction | |
US11763485B1 (en) | Deep learning based robot target recognition and motion detection method, storage medium and apparatus | |
WO2021082635A1 (en) | Region of interest detection method and apparatus, readable storage medium and terminal device | |
CN110309842B (en) | Object detection method and device based on convolutional neural network | |
CN107808129A (en) | A kind of facial multi-characteristic points localization method based on single convolutional neural networks | |
CN104392241B (en) | A kind of head pose estimation method returned based on mixing | |
CN106530340B (en) | A kind of specified object tracking | |
CN105760836A (en) | Multi-angle face alignment method based on deep learning and system thereof and photographing terminal | |
CN101964064A (en) | Human face comparison method | |
CN101968846A (en) | Face tracking method | |
CN102654903A (en) | Face comparison method | |
CN103514441A (en) | Facial feature point locating tracking method based on mobile platform | |
CN104462494A (en) | Remote sensing image retrieval method and system based on non-supervision characteristic learning | |
CN108109121A (en) | A kind of face based on convolutional neural networks obscures quick removing method | |
CN103279746A (en) | Method and system for identifying faces based on support vector machine | |
CN100383807C (en) | Feature point positioning method combined with active shape model and quick active appearance model | |
CN104091148B (en) | A kind of man face characteristic point positioning method and device | |
CN110060286A (en) | A kind of monocular depth estimation method | |
Zou et al. | Microarray camera image segmentation with Faster-RCNN | |
CN108961385A (en) | A kind of SLAM patterning process and device | |
CN116385660A (en) | Indoor single view scene semantic reconstruction method and system | |
CN107330363A (en) | A kind of quick Internet advertising board detection method | |
CN107729827A (en) | A kind of man face characteristic point positioning method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |