CN109558833A - A kind of face recognition algorithms evaluating method and device - Google Patents
A kind of face recognition algorithms evaluating method and device Download PDFInfo
- Publication number
- CN109558833A CN109558833A CN201811430321.XA CN201811430321A CN109558833A CN 109558833 A CN109558833 A CN 109558833A CN 201811430321 A CN201811430321 A CN 201811430321A CN 109558833 A CN109558833 A CN 109558833A
- Authority
- CN
- China
- Prior art keywords
- facial image
- identified
- age
- face
- recognition algorithms
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of face recognition algorithms evaluating method and devices, are related to a kind of evaluating method.This method comprises: establish facial image database, building library average time and build Kucheng's power for every sample facial image is counted；Determine the quality difficulty score of facial image to be identified；Determine the age at the age of the facial image to be identified and the target facial image in the facial image database；According to the facial image to be identified, the face recognition algorithms to be evaluated identify the facial image of predetermined number from the facial image database, obtain recognition result；According to the recognition result, quality difficulty score and age difficulty difference built library average time and build Kucheng's power and the facial image to be identified, the face recognition algorithms to be evaluated are evaluated and tested.Technical solution of the present invention is adapted in various actual combat scenes, and evaluating method is quick and accurate.
Description
Technical field
The present invention relates to a kind of evaluating method, in particular to a kind of face recognition algorithms evaluating method and device.
Background technique
Recognition of face assessment method is relatively simple at present, and usage scenario also compares limitation, when different scenes will spend longer
Between do data preparation, and recognition of face assessment method also more difficult migration actual combat scene.
The manufacturer of recognition of face at present is more, and every advantage and disadvantage are variant, needs to propose that one kind can be rapidly and accurately to not
The method evaluated and tested with the face recognition algorithms of manufacturer.
Summary of the invention
In order to overcome technical problem as described above, the present invention proposes a kind of face recognition algorithms evaluating method and dress
It sets, this method library average time and builds Kucheng's power and the face figure to be identified according to building for face recognition algorithms to be evaluated
Quality difficulty score, age difficulty difference and the recognition result of picture, evaluate and test the face recognition algorithms to be evaluated, can fit
Should be in various actual combat scenes, and evaluating method is quick and accurate.
Specific technical solution of the present invention is as follows:
In a first aspect, the present invention proposes a kind of face recognition algorithms evaluating method, comprising:
It establishes facial image database: according to different application scene, establishing facial image database using face recognition algorithms to be evaluated,
And count building library average time and build Kucheng's power for every sample facial image；
It determines quality of human face image: determining the quality difficulty score of facial image to be identified；
It determines the facial image age: determining the age of the facial image to be identified and in the facial image database
The age of target facial image；
Obtain recognition result: according to the facial image to be identified, the face recognition algorithms to be evaluated are from the face
The facial image that predetermined number is identified in image library, obtains recognition result；
Comprehensive evaluating: according to the identification built library average time and build Kucheng's power and the facial image to be identified
As a result, quality difficulty score and age difficulty difference, evaluate and test the face recognition algorithms to be evaluated.
Further, the determining quality of human face image: the quality difficulty score for determining facial image to be identified includes:
Its quality difficulty point is determined according to human face posture angle, light conditions and the readability of facial image to be identified
Number.
Further, the determining quality of human face image: the quality difficulty score for determining facial image to be identified includes:
Quality difficulty prediction model is established using depth convolutional neural networks, to the quality of the facial image to be identified
Difficulty score is predicted.
Further, the determining facial image age: the age of the facial image to be identified is determined and described
The age of target facial image in facial image database includes:
Age prediction model is established using depth convolutional neural networks, age to the facial image to be identified and
The age of target facial image in the facial image database is predicted.
Second aspect, the present invention proposes a kind of face recognition algorithms evaluating apparatus, including processor and memory, described to deposit
Reservoir is stored with an at least Duan Chengxu, and described program is executed by the processor to realize recognition of face as described in relation to the first aspect
Algorithm evaluating method.
The third aspect, the present invention propose a kind of computer readable storage medium, which is characterized in that deposit in the storage medium
An at least Duan Chengxu is contained, at least one section of program is executed by the processor to realize that face as described in relation to the first aspect is known
Other algorithm evaluating method.
Technical solution provided by the invention has the benefit that
The present invention establishes facial image database: according to different application scene, establishing face using face recognition algorithms to be evaluated
Image library, and count building library average time and build Kucheng's power for every sample facial image；Determine quality of human face image:
Determine the quality difficulty score and age difficulty difference of facial image to be identified；Obtain recognition result: according to the people to be identified
Face image, the face recognition algorithms to be evaluated identify the facial image of predetermined number from the facial image database, obtain
Recognition result；Comprehensive evaluating: according to the matter built library average time and build Kucheng's power and the facial image to be identified
Difficulty score, age difficulty difference and recognition result are measured, the face recognition algorithms to be evaluated are evaluated and tested.The technology of the present invention
Scheme can be suitable for a variety of different actual combat scenes, and quickly and accurately evaluate and test to face recognition algorithms to be evaluated.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment
Attached drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill of field, without creative efforts, it can also be obtained according to these attached drawings others
Attached drawing.
Fig. 1 show a kind of face recognition algorithms evaluating method schematic diagram of the invention；
Fig. 2 shows the structural schematic diagrams of face recognition algorithms evaluating apparatus involved in the embodiment of the present invention.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention
Case is described in further detail.
It is as shown in Figure 1 a kind of face recognition algorithms evaluating method schematic diagram of the invention, shows the specific of this method
Implementation steps, comprising:
In a step 101, it establishes facial image database: according to different application scene, being built using face recognition algorithms to be evaluated
Vertical facial image database, and count building library average time and build Kucheng's power for every sample facial image；
Library personnel amount and picture quality is built to be assessed according to actual combat occasion to select.Actual combat scene differs greatly, public security field
Jing Jianku quantity reaches million ranks, and residential area builds library quantity and reaches ten thousand ranks；Some scenes foreigner is relatively more, it is necessary to
It is put in storage the personnel of different nationalities.
In this step, building library average time and building Kucheng's power for every sample facial image is counted, which will make
For one of the foundation evaluated and tested in subsequent step to face recognition algorithms to be evaluated.
In a step 102, it determines quality of human face image: determining the quality difficulty score of facial image to be identified；
Facial image to be identified in this step refers to the test image for testing face recognition algorithms to be evaluated, then
It is easily understood that the requirement to face recognition algorithms to be evaluated is relatively high when the quality of facial image to be identified is poor
, that is to say, that face recognition algorithms to be tested can identify the poor facial image of quality, illustrate that its recognition performance is
Better.Here the quality of facial image, optionally, human face posture angle, light conditions including facial image and clear
Degree, in a kind of possible manipulation of physical, note facial angle is set as parameter d, then can carry out for human face posture angle following
Classification: positive face: d < 5 °；Low-angle: 5 ° < d < 15 °；Certain angle: 15 ° < d < 30 °；Wide-angle: 30 ° < d.In a kind of possible reality
In the operation of border, can carry out following classification for the light conditions of facial image: light is excessive, light is excessively dark and light is moderate.
In a kind of possible realization operation, can carry out following classification for the readability of facial image: clear certificate photo, comparison are clear
It is clear, relatively fuzzyyer and very fuzzy.The above-mentioned classification standard for picture quality is only used as a kind of optional mode classification, this hair
It is not limited thereto in bright technical solution.A kind of quality of facial image to be identified point 10 in the present invention is shown in following table 1
A rank, and table the case where its determining quality difficulty score.
Table 1
It should be noted that in a kind of possible practical operation, depth nerve can be passed through in technical solution of the present invention
Convolution CNN quality model forecast image quality, by deep learning frame such as TensorFlow, Caffe, Keras, using network
Such as AlexNet, resnet, VGG, MobileNet, training obtain model parameter, just can predict the face figure of maximum probability
Image quality amount difficulty score.
In step 103, it determines the facial image age: determining the age of the facial image to be identified and in the people
The age of target facial image in face image library；
It is understood that should include the facial image to be identified by the facial image database that step 101 is established
The facial image of the corresponding same person, that is, target facial image as described herein.Facial image to be identified is corresponding
Age and the facial image database in age between may have age gap, when age gap is bigger, identify difficulty
Be it is bigger, this for face recognition algorithms to be evaluated recognition performance require be it is relatively high, therefore, age gap here can
Using as one of the foundation evaluated and tested for face recognition algorithms to be evaluated in subsequent step.
It should be first to facial image to be identified and in the people before being evaluated and tested to face recognition algorithms to be evaluated
Facial image in face image library carries out respective labels production, and label substance includes the quality difficulty score of above-mentioned facial image
And the age, it should also include that can indicate identity information corresponding to facial image, the corresponding identity letter of the image of the same person
Breath should be consistent, and optionally, identity information can be the ID card No. of the corresponding natural person of facial image.
It should be noted that in a kind of possible practical operation, depth nerve can be passed through in technical solution of the present invention
The convolution CNN age models forecast image age, by deep learning frame such as TensorFlow, Caffe, Keras, using network
Such as AlexNet, resnet, VGG, MobileNet, training obtain model parameter, just can predict the face figure of maximum probability
As the corresponding age.
At step 104, obtain recognition result: according to the facial image to be identified, the recognition of face to be evaluated is calculated
Method identifies the facial image of predetermined number from the facial image database, obtains recognition result；
Facial image to be identified is input to the face recognition algorithms to be evaluated to identify, and from the facial image
It is picked out in library and the facial image by the corresponding same natural person of facial image to be identified, and the face picked out
Picture number can be preset, and optionally, can be 5 or 10 etc..Here recognition result is needed comprising facial image
Relevant information may include optionally the identity information of the corresponding natural person of unique identification image, with facial image to be identified
Between similarity and similarity ranking, it is preferred that in recognition result ranking results according to similarity size carry out descending row
Column.
In step 105, Kucheng's power and the people to be identified and comprehensive evaluating: are built at library average time according to described build
Recognition result, quality difficulty score and the age difficulty difference of face image, evaluate and test the face recognition algorithms to be evaluated.
Facial image database is established using face recognition algorithms to be evaluated in a step 101, and is counted for every sample people
Building for face image and builds Kucheng's power at library average time, in this step will according to build library average time and build Kucheng's power for
Face recognition algorithms progress to be evaluated is corresponding to build library performance evaluating.In a kind of possible realization, with CPU monokaryon performance
3.4GHZ is as reference standard: average time-consuming (ms/) score of single thread storage is denoted as St, and table 2 shows one kind of the invention
The case where being evaluated and tested library average time table is built, table 3 shows one kind of the present invention and builds the case where Kucheng's power Ss is evaluated and tested table.
Table 2
| Storage is average time-consuming (ms/) | Evaluate and test score St |
| 0-300ms | 10 points |
| 300-500ms | 8 points |
| 500ms-1000ms | 5 points |
| Greater than 1000ms | 1 point |
Table 3
It is incorporated in quality difficulty score, age difficulty difference and the knowledge of the facial image to be identified that abovementioned steps obtain
It, in this step will be according to quality difficulty score for be evaluated not as a result, being evaluated and tested to the face recognition algorithms to be evaluated
Survey face recognition algorithms progress is corresponding to build library performance evaluating.It should be noted that the age difficulty difference is that basis exists
The age of the facial image to be identified determined in step 103 and target facial image in the facial image database
Age carries out what age difference was calculated.
In a kind of possible practical operation, the picture number of the recognition result of face recognition algorithms to be evaluated is preset as
10, quality difficulty score is classified and is given a mark according in table 1, and technical solution of the present invention proposes a kind of according to individual people to be identified
The identification situation of face image, the score calculation formula that face recognition algorithms to be evaluated are evaluated and tested:
Wherein, i=1,2,3 ... n, i indicate i-th image in n facial images to be identified, when the first in recognition result
When matching correct, k=1；When second to the 5th matching is correct, k=0.5；When 6th to the tenth matching is correct, k=0.2；
Other k=0, a and b respectively indicate the quality difficulty score and age difficulty difference of the facial image for being evaluated and tested, and need
It is noted that working as b > 18, b=18, when this allows for the degree that age difference reaches certain, this difference can not be considered
Bring influences, this meets practical situations.
According to (1) formula it is found that when there is the face of matched same natural person in recognition result in images to be recognized
Image, then Si=0, that is, do not need the quality difficulty score for considering further that the image and age difficulty difference.
It should be noted that (1) formula is the formula evaluated and tested according to the recognition result of individual facial image to be identified,
When considering that the recognition result of all facial images to be identified is evaluated and tested, technical solution of the present invention proposes that a kind of summation takes mean value
Mode, it may be assumed that
Wherein, n indicates the number of total facial image to be identified.
So, comprehensive face recognition algorithms to be evaluated establish facial image database build library average time evaluation and test score St and
Kucheng's power evaluation and test score Ss is built, the present invention proposes the calculation formula of comprehensive evaluating score S a kind of:
S=Sm+Ss+St (3)
Technical solution of the present invention is put into the evaluation and test effect in practical application invention additionally discloses a kind of, primary medium-sized
The exhibitions of system of real name deploys 2.3 ten thousand blacklist libraries, using three manufacturer's alignment algorithms.The photo for arranging 500 people, by everyone 3
~6 photos are as test picture.It is evaluated and tested according to the face recognition algorithms of above-mentioned formula (1) to (3) to three manufacturers,
The appraisal result of last evaluation and test are as follows: A manufacturer 81.56 divides；B manufacturer 77.04 divides；C manufacturer 58.93 divides；It is optimal to select
Vendor A disposes similar exhibitions scene.
It should be noted that table 1, table 2 and the corresponding classification situation of table 3 and fractional weight can according to the actual situation into
Row adjustment, it is only necessary to which the quality for following facial image to be identified is more bad, the original that the fractional weight after identification is correct should be bigger
Then；The age difficulty difference of target facial image of the facial image to be identified with identification after correct is bigger, and fractional weight is bigger；It builds
Vertical facial image database to build library average time shorter and build that library rate is bigger, the bigger evaluation and test of fractional weight is given a mark rule,
It does not need to be limited with concrete condition proposed by the invention, in addition, the configuration parameter and k etc. in formula (1) are all can
Appropriate adjustment is carried out according to practical situations to obtain more acurrate reliable evaluation result.
The present embodiment establishes facial image database: according to different application scene, establishing people using face recognition algorithms to be evaluated
Face image library, and count building library average time and build Kucheng's power for every sample facial image；Determine facial image matter
Amount: the quality difficulty score and age difficulty difference of facial image to be identified are determined；Obtain recognition result: according to described to be identified
Facial image, the face recognition algorithms to be evaluated identify the facial image of predetermined number from the facial image database, obtain
To recognition result；Comprehensive evaluating: it builds library average time according to described and builds Kucheng's power and the facial image to be identified
Quality difficulty score, age difficulty difference and recognition result, evaluate and test the face recognition algorithms to be evaluated.Skill of the present invention
Art scheme can be suitable for a variety of different actual combat scenes, and quickly and accurately comment face recognition algorithms to be evaluated
It surveys.
Fig. 2 shows the structural schematic diagram of face recognition algorithms evaluating apparatus involved in the embodiment of the present invention, the devices
Mainly include processor 201, memory 202 and bus 203, the memory is stored with an at least Duan Chengxu, described program by
The processor is executed to realize the face recognition algorithms evaluating method as described in above-described embodiment.
Processor 201 includes one or more processing cores, and processor 201 passes through bus 203 and 202 phase of memory
Even, memory 10 realizes the above method when executing the program instruction in memory 202 for storing program instruction, processor 201
The face recognition algorithms evaluating method that embodiment provides.
Optionally, memory 202 can be by any kind of volatibility or non-volatile memory device or their group
It closes and realizes, such as static to access memory (SRAM) at any time, electrically erasable programmable read-only memory (EEPROM) is erasable to compile
Journey read-only memory (EPROM), programmable read only memory (PROM), read-only memory (ROM), magnetic memory, flash
Device, disk or CD.
The present invention also provides a kind of computer readable storage medium, be stored in the storage medium at least one instruction,
At least a Duan Chengxu, code set or instruction set, at least one instruction, an at least Duan Chengxu, code set or instruction set are by institute
State processor load and execute with realize above method embodiment provide face recognition algorithms evaluating method.
Optionally, the present invention also provides a kind of computer program products comprising instruction, when it runs on computers
When, so that computer executes face recognition algorithms evaluating method described in above-mentioned various aspects.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, the program can store computer-readable with one kind
In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely presently preferred embodiments of the present invention, is not used to limit 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 (8)
1. a kind of face recognition algorithms evaluating method characterized by comprising
It establishes facial image database: according to different application scene, establishing facial image database using face recognition algorithms to be evaluated, and unite
Count building library average time and build Kucheng's power for every sample facial image；
It determines quality of human face image: determining the quality difficulty score of facial image to be identified；
Determine the facial image age: the age for determining the facial image to be identified and the target in the facial image database
The age of facial image；
Obtain recognition result: according to the facial image to be identified, the face recognition algorithms to be evaluated are from the facial image
The facial image that predetermined number is identified in library, obtains recognition result；
Comprehensive evaluating: according to the identification knot built library average time and build Kucheng's power and the facial image to be identified
Fruit, quality difficulty score and age difficulty difference, evaluate and test the face recognition algorithms to be evaluated.
2. face recognition algorithms evaluating method according to claim 1, which is characterized in that the determination face figure to be identified
The quality difficulty score of picture includes:
Its quality difficulty score is determined according to human face posture angle, light conditions and the readability of facial image to be identified.
3. face recognition algorithms evaluating method according to claim 1, which is characterized in that the determination face figure to be identified
The quality difficulty score of picture includes:
Quality difficulty score prediction model is established using depth convolutional neural networks, to the quality of the facial image to be identified
Difficulty score is predicted.
4. face recognition algorithms evaluating method according to claim 1, which is characterized in that the determination people to be identified
The age of face image and the age of the target facial image in the facial image database include:
Age prediction model is established using depth convolutional neural networks, age to the facial image to be identified and in institute
The age for stating the target facial image in facial image database is predicted.
5. face recognition algorithms evaluating method according to claim 1, which is characterized in that the score S of the comprehensive evaluating
Calculation formula:
S=Sm+Ss+St
Wherein, to build library average time evaluation and test score, Ss, to build, Kucheng's power evaluates and tests score to St and Sm is all faces to be identified
The recognition result of image evaluates and tests score.
6. face recognition algorithms evaluating method according to claim 5, which is characterized in that all face figures to be identified
The evaluation and test formula of the recognition result evaluation and test score Sm of picture are as follows:
Wherein, n indicates the number of total facial image to be identified,I indicates that n is opened wait know
I-th image in others' face image, k indicate preset evaluation and test coefficient, and a and b respectively indicate the face figure for being evaluated and tested
The quality difficulty score and age difficulty difference of picture.
7. a kind of face recognition algorithms evaluating apparatus, which is characterized in that including processor and memory, the memory is stored with
An at least Duan Chengxu, described program are executed by the processor to realize that the recognition of face as described in claim 1 to 6 is any is calculated
Method evaluating method.
8. a kind of computer readable storage medium, which is characterized in that an at least Duan Chengxu is stored in the storage medium, it is described
The face recognition algorithms evaluating method as described in claim 1 to 6 is any is executed when at least one section of program operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811430321.XA CN109558833A (en) | 2018-11-28 | 2018-11-28 | A kind of face recognition algorithms evaluating method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811430321.XA CN109558833A (en) | 2018-11-28 | 2018-11-28 | A kind of face recognition algorithms evaluating method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109558833A true CN109558833A (en) | 2019-04-02 |
Family
ID=65867429
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811430321.XA Pending CN109558833A (en) | 2018-11-28 | 2018-11-28 | A kind of face recognition algorithms evaluating method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109558833A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110647940A (en) * | 2019-09-25 | 2020-01-03 | 捻果科技(深圳)有限公司 | Airport apron foreign person monitoring method based on video analysis and deep learning |
| CN110838102A (en) * | 2019-10-24 | 2020-02-25 | 开望(杭州)科技有限公司 | Intelligent image uploading method |
| CN111310709A (en) * | 2020-03-02 | 2020-06-19 | 邓谊 | Image-text annual newspaper emotion calibration method and system |
| CN111598132A (en) * | 2020-04-21 | 2020-08-28 | 烽火通信科技股份有限公司 | Portrait recognition algorithm performance evaluation method and device |
| CN111654694A (en) * | 2020-06-18 | 2020-09-11 | 北京字节跳动网络技术有限公司 | Quality evaluation method and device of image processing algorithm and electronic equipment |
| WO2020228129A1 (en) * | 2019-05-16 | 2020-11-19 | 罗普特科技集团股份有限公司 | Method and system for evaluating and invoking facial recognition algorithm engine |
| CN112733717A (en) * | 2021-01-11 | 2021-04-30 | 广州市金其利信息科技有限公司 | Method for optimizing face recognition based on face attributes |
| CN113095672A (en) * | 2021-04-09 | 2021-07-09 | 公安部物证鉴定中心 | Method and system for evaluating face image comparison algorithm |
| CN113255599A (en) * | 2021-06-29 | 2021-08-13 | 成都考拉悠然科技有限公司 | System and method for user-defined human flow testing face distribution control rate |
| CN115601799A (en) * | 2022-09-09 | 2023-01-13 | 广州市盛通建设工程质量检测有限公司（Cn） | Evaluation method, system, equipment and storage medium based on face recognition |
| CN116483733A (en) * | 2023-06-12 | 2023-07-25 | 数据堂（北京）科技股份有限公司 | Multi-dimensional artificial intelligence product evaluation method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615196A (en) * | 2009-07-28 | 2009-12-30 | 公安部第一研究所 | The test macro and the method for testing of millions one-to-many face recognition products |
| US20140341422A1 (en) * | 2013-05-10 | 2014-11-20 | Tencent Technology (Shenzhen) Company Limited | Systems and Methods for Facial Property Identification |
| CN107977647A (en) * | 2017-12-20 | 2018-05-01 | 上海依图网络科技有限公司 | A kind of face recognition algorithms evaluating method of suitable public security actual combat |
-
2018
- 2018-11-28 CN CN201811430321.XA patent/CN109558833A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615196A (en) * | 2009-07-28 | 2009-12-30 | 公安部第一研究所 | The test macro and the method for testing of millions one-to-many face recognition products |
| US20140341422A1 (en) * | 2013-05-10 | 2014-11-20 | Tencent Technology (Shenzhen) Company Limited | Systems and Methods for Facial Property Identification |
| CN107977647A (en) * | 2017-12-20 | 2018-05-01 | 上海依图网络科技有限公司 | A kind of face recognition algorithms evaluating method of suitable public security actual combat |
Non-Patent Citations (1)
| Title |
|---|
| 范英等: "人口信息人像比对系统评测方法研究与实践", 《电子科学技术》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020228129A1 (en) * | 2019-05-16 | 2020-11-19 | 罗普特科技集团股份有限公司 | Method and system for evaluating and invoking facial recognition algorithm engine |
| CN110647940A (en) * | 2019-09-25 | 2020-01-03 | 捻果科技(深圳)有限公司 | Airport apron foreign person monitoring method based on video analysis and deep learning |
| CN110838102A (en) * | 2019-10-24 | 2020-02-25 | 开望(杭州)科技有限公司 | Intelligent image uploading method |
| CN111310709A (en) * | 2020-03-02 | 2020-06-19 | 邓谊 | Image-text annual newspaper emotion calibration method and system |
| CN111598132A (en) * | 2020-04-21 | 2020-08-28 | 烽火通信科技股份有限公司 | Portrait recognition algorithm performance evaluation method and device |
| CN111654694A (en) * | 2020-06-18 | 2020-09-11 | 北京字节跳动网络技术有限公司 | Quality evaluation method and device of image processing algorithm and electronic equipment |
| CN112733717A (en) * | 2021-01-11 | 2021-04-30 | 广州市金其利信息科技有限公司 | Method for optimizing face recognition based on face attributes |
| CN112733717B (en) * | 2021-01-11 | 2021-11-23 | 广州市金其利信息科技有限公司 | Method for optimizing face recognition based on face attributes |
| CN113095672A (en) * | 2021-04-09 | 2021-07-09 | 公安部物证鉴定中心 | Method and system for evaluating face image comparison algorithm |
| CN113095672B (en) * | 2021-04-09 | 2024-06-07 | 公安部物证鉴定中心 | Evaluation method and system for facial image comparison algorithm |
| CN113255599A (en) * | 2021-06-29 | 2021-08-13 | 成都考拉悠然科技有限公司 | System and method for user-defined human flow testing face distribution control rate |
| CN115601799A (en) * | 2022-09-09 | 2023-01-13 | 广州市盛通建设工程质量检测有限公司（Cn） | Evaluation method, system, equipment and storage medium based on face recognition |
| CN116483733A (en) * | 2023-06-12 | 2023-07-25 | 数据堂（北京）科技股份有限公司 | Multi-dimensional artificial intelligence product evaluation method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109558833A (en) | A kind of face recognition algorithms evaluating method and device | |
| US9076030B2 (en) | Liveness detection | |
| CN107742100B (en) | A kind of examinee's auth method and terminal device | |
| WO2020155939A1 (en) | Image recognition method and device, storage medium and processor | |
| Chen et al. | Schelling points on 3D surface meshes | |
| CN108229330A (en) | Face fusion recognition methods and device, electronic equipment and storage medium | |
| CN109583489A (en) | Defect classifying identification method, device, computer equipment and storage medium | |
| CN107977647B (en) | Face recognition algorithm evaluation method suitable for public security actual combat | |
| CN108875522A (en) | Face cluster methods, devices and systems and storage medium | |
| CN110097003A (en) | Check class attendance method, equipment, storage medium and device neural network based | |
| CN105740808B (en) | Face identification method and device | |
| CN103679160B (en) | Human-face identifying method and device | |
| CN109993021A (en) | The positive face detecting method of face, device and electronic equipment | |
| US20240169518A1 (en) | Method and apparatus for identifying body constitution in traditional chinese medicine, electronic device, storage medium and program | |
| CN109389002A (en) | Biopsy method and device | |
| CN109657542A (en) | Personality prediction technique, device, computer equipment and the computer storage medium of interviewee | |
| CN111368772A (en) | Identity recognition method, device, equipment and storage medium | |
| CN108875509A (en) | Biopsy method, device and system and storage medium | |
| CN110175990A (en) | Quality of human face image determination method, device and computer equipment | |
| CN110502694A (en) | Lawyer's recommended method and relevant device based on big data analysis | |
| Haindl et al. | A competition in unsupervised color image segmentation | |
| CN107392191B (en) | Method for judging identity, device and electronic equipment | |
| CN110532838A (en) | Object test equipment and method and storage medium | |
| Tamilkodi | Automation system software assisting educational institutes for attendance, fee dues, report generation through email and mobile phone using face recognition | |
| CN110490058A (en) | Training method, device, system and the computer-readable medium of pedestrian detection model |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190402 |
|
| RJ01 | Rejection of invention patent application after publication |