CN103793697B - The identity mask method and face personal identification method of a kind of facial image - Google Patents
The identity mask method and face personal identification method of a kind of facial image Download PDFInfo
- Publication number
- CN103793697B CN103793697B CN201410053879.6A CN201410053879A CN103793697B CN 103793697 B CN103793697 B CN 103793697B CN 201410053879 A CN201410053879 A CN 201410053879A CN 103793697 B CN103793697 B CN 103793697B
- Authority
- CN
- China
- Prior art keywords
- face
- picture
- identity
- name
- marked
- 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
Abstract
The invention discloses the identity mask method and face personal identification method of a kind of facial image.This recognition methods is：1）The identity of each face picture to be marked is labeled：Search for the facial image similar with the picture and corresponding webpage；Frequency according to name appeared in webpage is returned determines the identity of the picture；The identity of face technology platform and face identification the model inspection picture is respectively adopted；Summary recognition result determines the final identity of the picture and marks；2）A group picture piece and 1 for same name will be belonged to）Middle annotation results carry out matching filtering for the picture of the name；3）The feature vector of each identity mark picture after extraction filtering, is trained the face picture after mark using machine learning algorithm, generates face identification model；4）For two facial images to be detected, extract its feature vector and judge whether it belongs to same people using face identification model.The present invention greatly promotes annotating efficiency and recognition effect.
Description
Technical field
The present invention relates to the identity mask method and face body of a kind of face identification method, more particularly to a kind of facial image
Part recognition methods, belongs to image identification technical field.
Background technology
Face recognition technology is used widely in each field at present, becomes a current research hotspot, for example apply
Numbers 201210313721.9, the patent document of title " face identification method ", application number 201210310643.7, title are " a kind of
The patent document of face identification method and its system ".
Wherein, the extraction of human face characteristic point and mark are an essential job in Face datection recognition methods, than
As application number 201310115471.2, title " a kind of face automatic marking method and system " are detected from the video of interception first
Go out face, obtain face picture set, then filter out face picture set, meanwhile, the hsv color for obtaining consecutive frame picture is straight
Square figure difference, shot segmentation is carried out using the Scene Incision algorithm of spatial color histogram, to the face from consecutive frame,
Detect angle point in the target area of the first frame, and next frame is given by these angle points are deferred using the method for local matching, and carry out
Corresponding renewal, and statistical match number, according to the threshold value of matching number, go on obtain face sequence according to this.Then lead to
Cross lip and move detection module and moved according to the lip of speaker in face sequence and detect speaker and non-speaker, by speaker, speak
Content and the time three that speaks fusion are labeled；Finally, the face in each sequence is read in, is positioned one by one, further according to positioning
As a result affine transformation is carried out, and extracts after conversion the characteristic point grey scale pixel value in fixed size border circular areas nearby, is used as this
Face characteristic.
Application number 200610096709.1, title " man face characteristic point positioning method in face identification system " are directed to people
Man face characteristic point positioning method in face identifying system, using the statistical model of image gradient directional information, passes through statistical inference
Method determine human face characteristic point, comprise the following steps:(1) definition and locating human face's characteristic point, that is, utilize the side of image gradient
To the human face characteristic point for defining and positioning candidate;(2) feature vector (3) of human face characteristic point utilizes one in extraction step (1)
The statistical model of a feature and relativeness for considering human face characteristic point, using the method for statistical inference, marks face characteristic
Point, so that it is determined that the position of the human face characteristic point needed.
Face technology belongs to machine learning category, and technology and system are required for experience data training process, i.e., a large amount of people
Face image and corresponding mark are given to algorithm together as input, and algorithm can learn corresponding automatically according to these training datas
Model is for practical application.Since the characteristic attribute information detected required by current method for detecting human face requires increasingly
It is abundant, generally by there is the facial image of mark to be trained using machine learning algorithm to obtain identification model, so as to numerous
The facial image not marked is labeled and identifies.But the mask method on face identity is not solved effectively always,
It is very time-consuming if going to screen mark one by one simply by manual method.
The content of the invention
For problems of the prior art, it is an object of the invention to provide a kind of identity mark side of facial image
Method and the recognition methods based on the circulation of face picture big data.
The technical scheme is that：
A kind of identity mask method of facial image, its step are：
1）The facial image similar with face picture to be marked and corresponding webpage are searched for from image search engine；
2）Statistics returns to the frequency of name appeared in webpage, and primarily determines that the face figure to be marked according to the frequency
The identity of piece；
3）The identity of face technology platform and face identification the model inspection face picture to be marked is respectively adopted；
4）According to step 2）, recognition result 3) determine the final identity of the face picture to be marked, it is to be marked to mark this
The identity of face picture.
Further, the face location and key point information of face picture to be marked are extracted first, by face location standard
Change to a reference format on the face.
Further, the N for belonging to same name of search is opened images comparison chart two-by-two by the face identification model
The identity of face belongs to the confidence level of same person in piece；Then according to confidence level determine face picture identity to be marked whether be
The name.
A kind of face personal identification method of facial image, its step are：
1）Automatic data acquisition system obtains face picture and its contextual information from server；
2）Data automatic marking system is labeled the identity of each face picture to be marked of acquisition；Wherein mark side
Method is：
21）The facial image similar with face picture to be marked and corresponding webpage are searched for from image search engine；
22）Statistics returns to the frequency of name appeared in webpage, and primarily determines that the face to be marked according to the frequency
The identity of picture；
23）The identity of face technology platform and face identification the model inspection face picture to be marked is respectively adopted；
24）According to step 22）, recognition result 23) determine the final identity of the face picture to be marked, mark this and wait to mark
Note the identity of face picture；
3）By the group picture piece for belonging to same name gathered in automatic data acquisition system and step 2）Middle annotation results
Matching filtering is carried out for the picture of the name, removes the face picture that the name is not belonging in the group；
4）Extraction step 3）The feature vector of each identity mark picture retained after filtering, automatic algorithms training system profit
Face picture after being marked with machine learning algorithm to identity is trained, and generates a face identification model；
5）For two facial images to be detected, extract its feature vector and judge it using the face identification model
Whether same people is belonged to.
Further, the automatic data acquisition system obtains face picture and its method for contextual information from server
For：
51）The server is according to the corresponding face picture file of name keyword search of input and preserves；
52）Calculate Hash codes, color histogram, context and the label information of each face picture file；
53）Each face picture is compared with having deposited face picture progress Hash codes and color histogram, removes repetition
Image；
54）User's face detection algorithm module detecting step 53）The each face picture retained after processing, by face location
Information
It is saved in database；Using the key point information on face key point location algorithm locating human face and it is saved in data
Storehouse.
Further, step 21）The face location and key point information of face picture to be marked are first extracted before, by face
Location criteria to a reference format on the face.
As described in Figure 1, its detection method comprises the following steps detecting system of the present invention：
1）Automatic data acquisition system, automatically from search engine, social networks, and photograph album class application background server of taking pictures
Constantly excavate the required human face data of learning algorithm and related context information；
2）Data automatic marking system, by a small amount of manual intervention, the noise in automatic fitration gathered data, and using upper
The required face identity markup information of context information automatic mining learning algorithm；
3）Automatic algorithms training system, is obtaining human face data and the identity markup information that automatic mining goes out, the system
Data are automatically periodically sent into Algorithm Learning system and carry out Algorithm for Training, wait to build executable algorithm mould after the completion of training automatically
Block；
4）3）In obtained newest algoritic module can be recycled into 2）Subsystem so that help preferably arrive people
Face identity marks.
Compared with prior art, the positive effect of the present invention is：
The present invention can be realized carries out automatic marking to facial image identity, substantially increases the effect of facial image mark
Rate；The face identification method of the present invention can help to make full use of the advantage of big data, greatly promote recognition effect.
Brief description of the drawings
Fig. 1 overall system schematic diagrames；
Fig. 2 automatic data collection method schematic diagrames；
Fig. 3 data automatic marking method schematic diagrames；
Fig. 4 automatic algorithms train schematic diagram.
Embodiment
The technology of the present invention is explained in further detail below in conjunction with the accompanying drawings.
1）Automatic data acquisition system（As shown in Figure 2）
One key condition of the lifting each sport technique segment algorithm performance of face technology is the extensive of acquisition better quality
Human face data.Conventional method is manually to build collection environment, organizes volunteer's facial image, manually the face of mark collection
Data, such as the picture position of face, the image coordinate of face key point, the age of face, identity etc..Conventional method gathers
Time-consuming, the data collected are also very dull, for example are all, or some age brackets regional at one, certain illumination bar
Under part, the view data of certain human face posture, its multifarious shortage can not meet the Algorithm for Training of high performance face technology
It is required that.The appearance of search engine and internet provides big data and excavates and the possibility that utilizes, and substantial amounts of name personal data can be with
Efficiently obtained by name key search very much.While also there is the face of substantial amounts of same person on social networks and photograph album
View data, these all provide abundant face identity data source for lifting face recognition algorithms.How these numbers is utilized
It it is also the current the problem of of requiring study according to boosting algorithm performance.
In view of the above-mentioned problems, human face data is excavated in the collection that this method is automated using following steps：
1. system searches for name keyword on a search engine, key word library is obtained from all kinds of encyclopaedia data, such as body
Educate, name of performing art star etc..
2. system downloads the result images file of search engine offer automatically, it is saved in a temporary file system, text
Part face image data is simultaneously to be not belonging to the people with the matched facial image of the name of retrieval, remaining facial image in part
Name is, it is necessary to take data automatic marking system, i.e. system 2）Described in the step of filtered.
3. the Hash codes for the image file downloaded in calculation procedure 2（Such as use MD5 algorithms）And color histogram data
With context and label information（Such as data source web, timestamp, keyword in context etc.）, database is stored in, and establish rope
Draw.
4. the data obtained in pair step 3 carry out duplicate removal processing：Each pictures will be with the storage in database
Picture carry out Hash codes and color histogram and compare, remove the image of repetition.
5. remaining picture is saved into a lasting distributed file system after being screened in step 4.
6. the face in the image preserved in user's face detection algorithm module detecting step 5, face location information is protected
It is stored to database；Using the key point information on face key point location algorithm module locating human face and it is saved in database.
7. the final system produces a distributed file system for storing image file data and one is preserved respectively
The distributed data base of kind face and image metamessage.
2）Data automatic marking system（As shown in Figure 3）
1. for the face picture produced in acquisition system, the face location and key produced using acquisition system step 6
Face location is normalized into point information a fixed size and position preserves into image and facilitates subsequent step 4）Middle extraction feature
Vector.
2. searching for the facial image produced in image search engine in uploading step 1 in third party, search obtains similar
Face picture source web page corresponding with its connects.Name keyword is analyzed in obtained results web page, each name is counted and closes
The frequency that key word occurs, confidence level is converted into by its frequency statistics result with equation below（Assuming that there is M name, wherein pi tables
Show the frequency that i-th of name occurs, Xi represents that i-th of name is the confidence level of the facial image identity）.
Xi=pi/ (p1+p2+...+pM),
3. in third party's face technology API platforms（With reference to http://www.skybiometry.com/Demo；http://
www.lambdal.com/）The facial image produced in middle uploading step 1, obtains identification as a result, extraction returns the result row
The possible name identity gone out（Assuming that API, which is returned, K possible name）The confidence provided with third party face technology API platforms
Degree.The confidence level also specific name be the facial image identity confidence level.
4. by 2,3 result using weighted average obtain the data downloaded in acquisition system which and download name used
Keyword is matched, so that the image data after being filtered.
Experiment shows, this method can obtain extremely accurately face identity labeled data.Results of property is shown in Table 1.We
List the performance figures of traditional manual mask method as a comparison.Conventional method checks everyone in download pictures one by one
Whether face picture, which matches, is downloaded name keyword used so as to filter out the picture met.
Table 1 marks Contrast on effect table for identity
| Mark manually | Method proposed by the present invention | |
| Mark accuracy | 99.4% | 99.2% |
| Average every pictures label time | 12 seconds | 0.8 second |
3）Automatic algorithms training system（As shown in Figure 4）
Obtaining labeling system 2）Face image data after filtering, the system extract the spy of each identity mark picture
Sign vector（The LBP in open field, any one feature vector such as Gabor, HOG can be used）, automatic algorithms training system
Face picture after periodically being marked using machine learning algorithm to identity is trained, and generates a face identification model；So
The data for meeting screening conditions are imported into Algorithm for Training system so as to detect whether inputted picture belongs to same person afterwards.Its
Comprise the following steps that：
1. user is periodically according to demand by the human face data amount and screening conditions of needs（For example image derives from 2013
10000 famous persons in internet search for data, everyone 50 pictures）One job queue data storehouse of typing.
2. the timing of automatic algorithms training system reads task from job queue data storehouse.
It is required that 3. the target algorithm of the image in 2 and data in task is normalized into the Algorithm for Training by system
Storage format.
4. the data after the normalization in 3 are uploaded to learning training server and are trained by system, training objective be to
Fixed a pair of face image data algoritic module exports whether this pair of of people is same person, generates a face identification model；
For the facial image of same name retrieval, its feature vector is extracted；Then using the face identification model to its into
Row detection, identifies whether it belongs to same people.
The adaptive face machine learning algorithm training system based on big data that the present invention describes can be used for face skill
The modules of art, including but not limited to Face datection, face key point location, dividing property of face character（Gender, age, kind
Race, expression etc.）, and face identification.
Claims (6)
1. a kind of identity mask method of facial image, its step are：
1) face picture of the face to be marked obtained according to automatic data acquisition system from server, from image search engine
The middle search facial image similar with face picture to be marked and corresponding webpage；
2) statistics returns to the frequency of name appeared in webpage, by frequency statistics result formula Xi=pi/ (p1+p2+...+
PM confidence level) is converted into, determines the identity of the face picture to be marked；Wherein, M name is shared, pi represents that i-th of name goes out
Existing frequency, Xi represent that i-th of name is the confidence level of the facial image identity；
3) identity of face technology platform and face identification the model inspection face picture to be marked is respectively adopted；
4) recognition result according to step 2), 3) determines the final identity of the face picture to be marked using weighted average, mark
The identity of the face picture to be marked.
2. the method as described in claim 1, it is characterised in that extract the face location and key of face picture to be marked first
Point information, a reference format is normalized on the face by face location.
3. method as claimed in claim 1 or 2, it is characterised in that the face identification model is same by belonging to for search
The identity that the N of name images compare face in picture two-by-two belongs to the confidence level of same person；Then determined according to confidence level
Whether face picture identity to be marked is the name.
4. a kind of face personal identification method of facial image, its step are：
1) automatic data acquisition system obtains face picture and its contextual information from server；
2) data automatic marking system is labeled the identity of each face picture to be marked of acquisition；Wherein mask method
For：
21) face picture of the face to be marked obtained according to the automatic data acquisition system from server, is searched from image
Index holds up the middle search facial image similar with face picture to be marked and corresponding webpage；
22) statistics returns to the frequency of name appeared in webpage, by frequency statistics result formula Xi=pi/ (p1+p2+...
+ pM) confidence level is converted into, determine the identity of the face picture to be marked；Wherein, M name is shared, pi represents i-th of name
The frequency of appearance, Xi represent that i-th of name is the confidence level of the facial image identity；
23) identity of face technology platform and face identification the model inspection face picture to be marked is respectively adopted；
24) recognition result according to step 22), 23) determines the final identity of the face picture to be marked using weighted average, mark
Note the identity of the face picture to be marked；
3) being by annotation results in the group picture piece for belonging to same name gathered in automatic data acquisition system and step 2) should
The picture of name carries out matching filtering, removes the face picture that the name is not belonging in the group；
4) extraction step 3) feature vector of each identity mark picture that retains after filtering, automatic algorithms training system utilizes machine
Face picture after device learning algorithm marks identity is trained, and generates a face identification model；
5) for two facial images to be detected, extract whether its feature vector judges it using the face identification model
Belong to same people.
5. method as claimed in claim 4, it is characterised in that the automatic data acquisition system obtains face figure from server
The method of piece and its contextual information is：
51) server according to the corresponding face picture file of name keyword search of input and preserves；
52) Hash codes, color histogram, context and the label information of each face picture file are calculated；
53) each face picture is compared with having deposited face picture progress Hash codes and color histogram, removes the image of repetition；
54) user's face detection algorithm module detecting step 53) each face picture for retaining after processing, by face location information
It is saved in database；Using the key point information on face key point location algorithm locating human face and it is saved in database.
6. method as described in claim 4 or 5, it is characterised in that the people of face picture to be marked is first extracted before step 21)
Face position and key point information, a reference format is normalized on the face by face location.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410053879.6A CN103793697B (en) | 2014-02-17 | 2014-02-17 | The identity mask method and face personal identification method of a kind of facial image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410053879.6A CN103793697B (en) | 2014-02-17 | 2014-02-17 | The identity mask method and face personal identification method of a kind of facial image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103793697A CN103793697A (en) | 2014-05-14 |
| CN103793697B true CN103793697B (en) | 2018-05-01 |
Family
ID=50669342
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410053879.6A Active CN103793697B (en) | 2014-02-17 | 2014-02-17 | The identity mask method and face personal identification method of a kind of facial image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103793697B (en) |
Families Citing this family (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984738B (en) * | 2014-05-22 | 2017-05-24 | 中国科学院自动化研究所 | Role labelling method based on search matching |
| CN104091164A (en) * | 2014-07-28 | 2014-10-08 | 北京奇虎科技有限公司 | Face picture name recognition method and system |
| CN104281842A (en) * | 2014-10-13 | 2015-01-14 | 北京奇虎科技有限公司 | Face picture name identification method and device |
| CN104778481B (en) * | 2014-12-19 | 2018-04-27 | 五邑大学 | A kind of construction method and device of extensive face pattern analysis sample storehouse |
| CN104463177A (en) * | 2014-12-23 | 2015-03-25 | 北京奇虎科技有限公司 | Similar face image obtaining method and device |
| CN104537341B (en) * | 2014-12-23 | 2016-10-05 | 北京奇虎科技有限公司 | Face picture information getting method and device |
| US10489637B2 (en) | 2014-12-23 | 2019-11-26 | Beijing Qihoo Technology Company Limited | Method and device for obtaining similar face images and face image information |
| CN106462349B (en) * | 2015-03-27 | 2019-10-25 | 华为技术有限公司 | A kind of electronic photo display methods, device and mobile device |
| CN104850600B (en) * | 2015-04-29 | 2019-05-28 | 百度在线网络技术（北京）有限公司 | A kind of method and apparatus for searching for the picture comprising face |
| CN106326815B (en) * | 2015-06-30 | 2019-09-13 | 芋头科技（杭州）有限公司 | A kind of facial image recognition method |
| CN105138245B (en) * | 2015-09-30 | 2018-06-29 | 北京奇虎科技有限公司 | A kind of duplicate removal treatment method and device of intelligent terminal screenshot picture |
| CN106560840B (en) | 2015-09-30 | 2019-08-13 | 腾讯科技（深圳）有限公司 | A kind of image information identifying processing method and device |
| CN105468760B (en) * | 2015-12-01 | 2018-09-11 | 北京奇虎科技有限公司 | The method and apparatus that face picture is labeled |
| CN105979331A (en) * | 2015-12-01 | 2016-09-28 | 乐视致新电子科技（天津）有限公司 | Smart television data recommend method and device |
| CN105678622A (en) * | 2016-01-07 | 2016-06-15 | 平安科技（深圳）有限公司 | Analysis method and system for vehicle insurance claim-settlement photos |
| CN106203294B (en) * | 2016-06-30 | 2019-05-03 | 广东微模式软件股份有限公司 | The testimony of a witness based on face character analysis unifies auth method |
| CN106327546B (en) * | 2016-08-24 | 2020-12-08 | 北京旷视科技有限公司 | Method and device for testing face detection algorithm |
| CN106844412A (en) * | 2016-11-02 | 2017-06-13 | 厦门中控生物识别信息技术有限公司 | A kind of human face data collection method and device |
| CN106503691B (en) * | 2016-11-10 | 2019-12-20 | 广州视源电子科技股份有限公司 | Identity labeling method and device for face picture |
| CN106548162B (en) * | 2016-11-24 | 2019-03-29 | 中译语通科技股份有限公司 | A method of automatically extracting band name human face data from news pages |
| CN106649610A (en) * | 2016-11-29 | 2017-05-10 | 北京智能管家科技有限公司 | Image labeling method and apparatus |
| CN106934364A (en) * | 2017-03-09 | 2017-07-07 | 腾讯科技（上海）有限公司 | The recognition methods of face picture and device |
| CN108875453A (en) * | 2017-05-11 | 2018-11-23 | 北京旷视科技有限公司 | The method, apparatus and computer storage medium of face picture bottom library registration |
| CN108228872A (en) | 2017-07-21 | 2018-06-29 | 北京市商汤科技开发有限公司 | Facial image De-weight method and device, electronic equipment, storage medium, program |
| CN110058407A (en) * | 2018-01-19 | 2019-07-26 | 尹寅 | A kind of intelligent glasses system shooting photos and videos |
| CN108921204B (en) * | 2018-06-14 | 2023-12-26 | 平安科技（深圳）有限公司 | Electronic device, picture sample set generation method, and computer-readable storage medium |
| CN109583325B (en) * | 2018-11-12 | 2023-06-27 | 平安科技（深圳）有限公司 | Face sample picture labeling method and device, computer equipment and storage medium |
| CN109919754A (en) * | 2019-01-24 | 2019-06-21 | 北京迈格威科技有限公司 | A kind of data capture method, device, terminal and storage medium |
| CN110084289B (en) * | 2019-04-11 | 2021-07-27 | 北京百度网讯科技有限公司 | Image annotation method and device, electronic equipment and storage medium |
| CN110825808A (en) * | 2019-09-23 | 2020-02-21 | 重庆特斯联智慧科技股份有限公司 | Distributed human face database system based on edge calculation and generation method thereof |
| CN113094538A (en) * | 2019-12-23 | 2021-07-09 | 中国电信股份有限公司 | Image retrieval method, device and computer-readable storage medium |
| CN115129921B (en) * | 2022-06-30 | 2023-05-26 | 重庆紫光华山智安科技有限公司 | Picture retrieval method, apparatus, electronic device, and computer-readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101668176A (en) * | 2009-09-25 | 2010-03-10 | 北京酷联天下科技有限公司 | Multimedia content-on-demand and sharing method based on social interaction graph |
| CN102055932A (en) * | 2009-10-30 | 2011-05-11 | 深圳Tcl新技术有限公司 | Method for searching television program and television set using same |
| CN102629275A (en) * | 2012-03-21 | 2012-08-08 | 复旦大学 | Face and name aligning method and system facing to cross media news retrieval |
| CN102648462A (en) * | 2009-11-18 | 2012-08-22 | 高通股份有限公司 | Methods and systems for managing electronic messages |
| CN102938065A (en) * | 2012-11-28 | 2013-02-20 | 北京旷视科技有限公司 | Facial feature extraction method and face recognition method based on large-scale image data |
-
2014
- 2014-02-17 CN CN201410053879.6A patent/CN103793697B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101668176A (en) * | 2009-09-25 | 2010-03-10 | 北京酷联天下科技有限公司 | Multimedia content-on-demand and sharing method based on social interaction graph |
| CN102055932A (en) * | 2009-10-30 | 2011-05-11 | 深圳Tcl新技术有限公司 | Method for searching television program and television set using same |
| CN102648462A (en) * | 2009-11-18 | 2012-08-22 | 高通股份有限公司 | Methods and systems for managing electronic messages |
| CN102629275A (en) * | 2012-03-21 | 2012-08-08 | 复旦大学 | Face and name aligning method and system facing to cross media news retrieval |
| CN102938065A (en) * | 2012-11-28 | 2013-02-20 | 北京旷视科技有限公司 | Facial feature extraction method and face recognition method based on large-scale image data |
Non-Patent Citations (5)
| Title |
|---|
| Automatic Face Naming with Caotion-based Supervision;Matthieu Guillaumin et al.;《Computer Vision and Pattern Recognition,2008》;20081231;第1-8页 * |
| Cross-media alignment of names and faces;PHi The Pham et al.;《IEEE Transactions on Multimedia》;20100131;第12卷(第1期);第13-27页 * |
| 基于GraphCuts的图像标注工具及管理系统的设计与实现;付明;《中国优秀硕士学位论文全文数据库 信息科技辑》;20131215(第S2期);论文正文第1-63页 * |
| 找到你的脸_智能图片搜索技术;盛嘉;《互联网天地》;20090228(第02期);第70页 * |
| 网络新闻图像中人脸标注技术研究;刘胜宇;《中国优秀硕士学位论文全文数据库 信息科技辑》;20120415(第04期);论文正文第1-44页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103793697A (en) | 2014-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103793697B (en) | The identity mask method and face personal identification method of a kind of facial image | |
| CN103824053B (en) | The sex mask method and face gender detection method of a kind of facial image | |
| WO2020224424A1 (en) | Image processing method and apparatus, computer readable storage medium, and computer device | |
| US10755086B2 (en) | Picture ranking method, and terminal | |
| US10140508B2 (en) | Method and apparatus for annotating a video stream comprising a sequence of frames | |
| CN105678250B (en) | Face identification method and device in video | |
| US9436682B2 (en) | Techniques for machine language translation of text from an image based on non-textual context information from the image | |
| US9940577B2 (en) | Finding semantic parts in images | |
| CN106503691B (en) | Identity labeling method and device for face picture | |
| CN111191695A (en) | Website picture tampering detection method based on deep learning | |
| WO2019033571A1 (en) | Facial feature point detection method, apparatus and storage medium | |
| CN110008909B (en) | Real-name system business real-time auditing system based on AI | |
| CN105740758A (en) | Internet video face recognition method based on deep learning | |
| Tang et al. | View-independent facial action unit detection | |
| CN107909033A (en) | Suspect's fast track method based on monitor video | |
| CN103824052A (en) | Multilevel semantic feature-based face feature extraction method and recognition method | |
| CN107679070B (en) | Intelligent reading recommendation method and device and electronic equipment | |
| JP6787831B2 (en) | Target detection device, detection model generation device, program and method that can be learned by search results | |
| CN110287952A (en) | A kind of recognition methods and system for tieing up sonagram piece character | |
| Taverriti et al. | Real-time wearable computer vision system for improved museum experience | |
| CN109753962A (en) | Text filed processing method in natural scene image based on hybrid network | |
| CN106537387A (en) | Retrieving/storing images associated with events | |
| CN111144469A (en) | End-to-end multi-sequence text recognition method based on multi-dimensional correlation time sequence classification neural network | |
| Ren et al. | A novel scene text detection algorithm based on convolutional neural network | |
| Rankothge et al. | Plant recognition system based on neural networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |