CN108062521A - Method for detecting human face, device, terminal and medium based on convolutional neural networks - Google Patents
Method for detecting human face, device, terminal and medium based on convolutional neural networks Download PDFInfo
- Publication number
- CN108062521A CN108062521A CN201711319230.4A CN201711319230A CN108062521A CN 108062521 A CN108062521 A CN 108062521A CN 201711319230 A CN201711319230 A CN 201711319230A CN 108062521 A CN108062521 A CN 108062521A
- Authority
- CN
- China
- Prior art keywords
- face
- image
- barycenter
- sample set
- depth characteristic
- 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
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
Abstract
The applicable field of computer technology of the present invention, providing a kind of method for detecting human face based on convolutional neural networks, device, terminal and storage medium, this method includes：When receiving Face datection request, image to be detected is obtained from Face datection request, Face datection is carried out at the same time to image to be detected by advance trained first default quantity individual's face posture grader, the face that first default quantity individual's face posture detection of classifier goes out is marked and exported, it is detected so as to carry out the differentiation of face according to the different postures of face in image to be detected, the detection that becomes more meticulous of human face posture is realized, improves the speed and accuracy of Face datection.
Description
Technical field
The invention belongs to technical field of face recognition more particularly to a kind of Face datection sides based on convolutional neural networks
Method, device, terminal and medium.
Background technology
Recognition of face is a kind of biological identification technology that facial feature information based on people carries out identification, at present by
Every field is widely applied to, for example, the functions such as human-computer interaction, robot, graphical analysis, face authentication or video monitoring.
The accuracy of recognition of face determines the application prospect of recognition of face, therefore, is related to the company of field of face identification and scientific research machine
Structure has all put into substantial amounts of manpower and materials the accuracy of recognition of face is continuously improved.
At present, based on the main or positive and static recognition of face of recognition of face, when face is not front in image
Or during in dynamic change, recognition of face is with regard to particularly difficult, at this point, general use manually carries out recognition of face to image,
This method is cumbersome and time consuming, and not only efficiency is low, and the accuracy rate of manual identified is very low when face has partial occlusion.
The content of the invention
It is an object of the invention to provide a kind of method for detecting human face based on convolutional neural networks, device, terminal and
Medium, it is intended to solve that due to the prior art a kind of effective method for detecting human face can not be provided, cause the inspection in Face datection
Survey the problem of accuracy is low and detection result is bad.
On the one hand, the present invention provides a kind of method for detecting human face based on convolutional neural networks, the described method includes under
State step：
When receiving Face datection request, image to be detected is obtained from Face datection request；
Described image to be detected is carried out at the same time by advance trained first default quantity individual's face posture grader
Face datection；
The face that described first default quantity individual's face posture detection of classifier goes out is marked and exported.
On the other hand, the present invention provides a kind of human face detection device based on convolutional neural networks, described device includes：
Image receiving unit needs to detect for when receiving Face datection request, receiving the Face datection request
Image to be detected；
Image detecting element is treated for passing through trained first default quantity individual's face posture grader in advance to described
Detection image is carried out at the same time Face datection；And
Mark output unit, for the face that described first default quantity individual's face posture detection of classifier goes out into rower
Remember and export.
On the other hand, the present invention also provides a kind of image processing terminal, including memory, processor and it is stored in institute
The computer program that can be run in memory and on the processor is stated, the processor performs real during the computer program
Now such as the step of above-mentioned method for detecting human face based on convolutional neural networks.
On the other hand, the present invention also provides a kind of computer readable storage medium, the computer readable storage mediums
Computer program is stored with, such as the above-mentioned face based on convolutional neural networks is realized when the computer program is executed by processor
The step of detection method.
The present invention obtains image to be detected, by pre- when receiving Face datection request from Face datection request
First trained first default quantity individual's face posture grader is carried out at the same time Face datection to image to be detected, default to first
The face that quantity human face posture detection of classifier goes out is marked and exports, so as to the difference according to face in image to be detected
Posture carry out face differentiation detection, realize the detection that becomes more meticulous of human face posture, improve Face datection speed and
Accuracy.
Description of the drawings
Fig. 1 is the realization flow chart for the method for detecting human face based on convolutional neural networks that the embodiment of the present invention one provides；
Fig. 2 is the structure diagram of the human face detection device provided by Embodiment 2 of the present invention based on convolutional neural networks；
Fig. 3 is the structure diagram for the terminal that the embodiment of the present invention three provides；And
Fig. 4 is a kind of structure diagram for image processing terminal that the embodiment of the present invention four provides.
Specific embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to the accompanying drawings and embodiments, it is right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
The specific implementation of the present invention is described in detail below in conjunction with specific embodiment：
Embodiment one：
Fig. 1 shows the realization stream for the method for detecting human face based on convolutional neural networks that the embodiment of the present invention one provides
Journey, for convenience of description, illustrate only with the relevant part of the embodiment of the present invention, details are as follows：
In step S101, when receiving Face datection request, image to be detected is obtained from Face datection request.
The embodiment of the present invention is suitable for image processing terminal, specifically, suitable for face detection system or platform, with defeated
The face of different postures in image to be detected is detected when entering image to be detected.In embodiments of the present invention, when receiving image
After processing terminal Face datection request input by user, image to be detected is obtained from the Face datection request received, wherein,
Image to be detected can be the photo of shooting, or video image is not limited herein.
In step s 102, by advance trained first default quantity individual's face posture grader to image to be detected
It is carried out at the same time Face datection.
In embodiments of the present invention, the first default quantity is by the pre-set people for needing to detect of image processing terminal user
Face posture quantity.
Preferably, before step S102, deep learning is carried out to the facial image in facial image sample set, to obtain
Sample set under different postures, then the human face posture grader built in advance is carried out by the sample set under different postures
Training finally obtains trained first default quantity individual's face posture grader, in this way, by facial image sample set into
Row deep learning, training obtain the human face posture grader under different postures, effectively reduce human face posture variation and detection is tied
The influence of fruit, so as to improve the accuracy rate of Face datection and speed, and can be according to the human face posture demand of user to figure to be detected
As being detected, to extract the human face posture that user needs in image, wherein, sample set is a kind of face sample of posture
Set.It is further preferred that when the human face posture grader to building in advance is trained, AdaBoost algorithms are used
Classifier training is carried out, so as to more quickly complete the training to each human face posture.
In embodiments of the present invention, when carrying out deep learning to the facial image in facial image sample set, it is preferable that
The depth characteristic of facial image is extracted from facial image sample set using depth convolutional neural networks, is calculated by default cluster
Method clusters the depth characteristic extracted, to obtain the sample set of various different posture servant's face images, thus will be each
Kind facial image sample is divided into the sample set of the facial image of several different postures so that Face datection more becomes more meticulous.
Preferably, the depth characteristic extracted is clustered by K-means (hard cluster) algorithm, specifically, first
Random selected second default quantity barycenter, according to the second default quantity barycenter and the position of depth characteristic, by depth characteristic
Position is distributed to depth characteristic apart from minimum barycenter, to obtain second default the first cluster of quantity, then calculates second in advance
If the average value of the first cluster of quantity, with the barycenter of default the first cluster of quantity of update second, the second present count is being updated
After the barycenter of the first cluster of amount, jump to and depth characteristic is distributed into the position barycenter minimum with depth characteristic distance, with
Barycenter is cyclically updated, until convergence, so as to simplify the cluster process of facial image sample set.
It is further preferred that when depth characteristic is distributed to the position barycenter minimum with depth characteristic distance, pass through mesh
Scalar functionsThe distance of barycenter and depth characteristic is obtained, then obtains most connecing in barycenter by minimizing object function
The index c of the barycenter of nearly depth characteristici：=j, wherein, i is the variable of depth characteristic, and f is the position of depth characteristic, and j is matter
Heart variable, u be barycenter position, DijIt is the distance between depth characteristic and barycenter, ciIt is closest to the barycenter rope of depth characteristic
Draw, depth characteristic is distributed to by the position barycenter minimum with depth characteristic distance by Euclidean distance, so as to more simply and quickly
Depth characteristic is distributed into the position barycenter minimum with depth characteristic distance.
In step s 103, the face gone out to first default quantity individual's face posture detection of classifier is marked and defeated
Go out.
In embodiments of the present invention, the human face posture grader of the first default quantity is carried out at the same time image to be detected analysis
After detection, the face detected is marked, most backward image processing terminal user shows output.
In embodiments of the present invention, face figure is extracted from facial image sample set using depth convolutional neural networks first
The depth characteristic of picture clusters the depth characteristic extracted, to obtain sample of various different posture servant's face images
Collection, then human face posture grader is trained by the sample set under different postures, to obtain trained human face posture
Grader, and then differentiation detection is carried out to the face in image by trained human face posture grader, realize face
The detection that becomes more meticulous of posture improves the speed and accuracy of Face datection.
Embodiment two：
Fig. 2 shows the structure of the human face detection device provided by Embodiment 2 of the present invention based on convolutional neural networks, is
Convenient for explanation, illustrate only with relevant part of the embodiment of the present invention, including：
Image receiving unit 21 trained first presets quantity individual's face posture grader to be checked in advance for passing through
Altimetric image is carried out at the same time Face datection；
Image detecting element 22 trained first presets quantity individual's face posture grader to described in advance for passing through
Image to be detected is carried out at the same time Face datection；And
Output unit 23 is marked, the face for going out to first default quantity individual's face posture detection of classifier is marked
And it exports.
In embodiments of the present invention, when receiving Face datection request, mapping to be checked is obtained from Face datection request
Picture is carried out at the same time Face datection by advance trained first default quantity individual's face posture grader to image to be detected,
The face that first default quantity individual's face posture detection of classifier goes out is marked and exported, so as to according in image to be detected
The different postures of face carry out the differentiation detection of face, realize the detection that becomes more meticulous of human face posture, improve Face datection
Speed and accuracy.
In embodiments of the present invention, each unit of the human face detection device based on convolutional neural networks can be by corresponding hardware
Or software unit is realized, each unit can be independent soft and hardware unit, can also be integrated into a soft and hardware unit, herein
Not limiting the present invention.The specific embodiment of each unit can refer to the description of embodiment one, and details are not described herein.
Embodiment three：
Fig. 3 shows the structure for the human face detection device based on convolutional neural networks that the embodiment of the present invention three provides, and is
Convenient for explanation, illustrate only with relevant part of the embodiment of the present invention, including：
Face unit 31, for carrying out deep learning to the facial image in facial image sample set, to obtain not
With the sample set under posture；
Classifier training unit 32, it is pre- to obtain for being trained by the sample set of different postures to grader
If the human face posture grader of quantity；
Image receiving unit 33 trained first presets quantity individual's face posture grader to be checked in advance for passing through
Altimetric image is carried out at the same time Face datection；
Image detecting element 34 trained first presets quantity individual's face posture grader to described in advance for passing through
Image to be detected is carried out at the same time Face datection；And
As a result output unit 35, the face for going out to first default quantity individual's face posture detection of classifier are marked
And it exports.
Wherein, face unit 31 further includes：
Feature extraction unit 311, for depth convolutional neural networks to be used to extract face figure from facial image sample set
The depth characteristic of picture；And
Feature clustering unit 312, for being clustered by default clustering algorithm to the depth characteristic extracted, with
To the sample set of various different posture servant's face images.
In embodiments of the present invention, face figure is extracted from facial image sample set using depth convolutional neural networks first
The depth characteristic of picture clusters the depth characteristic extracted, to obtain sample of various different posture servant's face images
Collection, then human face posture grader is trained by the sample set under different postures, to obtain trained human face posture
Grader, and then differentiation detection is carried out to the face in image by trained human face posture grader, realize face
The detection that becomes more meticulous of posture improves the speed and accuracy of Face datection.
In embodiments of the present invention, each unit of the human face detection device based on convolutional neural networks can be by corresponding hardware
Or software unit is realized, each unit can be independent soft and hardware unit, can also be integrated into a soft and hardware unit, herein
Not limiting the present invention.The specific embodiment of each unit can refer to the description of embodiment one, and details are not described herein.
Example IV：
Fig. 4 shows the structure for the image processing terminal that the embodiment of the present invention 4 provides, and for convenience of description, illustrates only
With the relevant part of the embodiment of the present invention, including：
The image processing terminal 4 of the embodiment of the present invention includes processor 41, memory 42 and is stored in memory 42
And the computer program 43 that can be run on processor 41.The processor 41 is realized above-mentioned based on volume when performing computer program 43
Step in the method for detecting human face embodiment of product neutral net, such as step S101 to S103 shown in FIG. 1.Alternatively, processing
Device 41 realizes each list in above-mentioned each human face detection device embodiment based on convolutional neural networks when performing computer program 43
The function of member, such as the function of unit 31 to 35 shown in unit 21 to 23 shown in Fig. 2 and Fig. 3.
In embodiments of the present invention, the processor perform computer program when, first using depth convolutional neural networks from
The depth characteristic of facial image is extracted in facial image sample set, the depth characteristic extracted is clustered, it is various to obtain
The sample set of different posture servant face images, then human face posture grader is instructed by the sample set under different postures
Practice, to obtain trained human face posture grader, so as to carry out face according to the different postures of face in image to be detected
Differentiation detects, and realizes the detection that becomes more meticulous of human face posture, improves the speed and accuracy of Face datection.
The processor realizes the above-mentioned method for detecting human face embodiment based on convolutional neural networks when performing computer program
In step can refer to the description of embodiment one, details are not described herein.
Embodiment five：
In embodiments of the present invention, a kind of computer readable storage medium is provided, which deposits
Computer program is contained, which realizes the above-mentioned Face datection side based on convolutional neural networks when being executed by processor
Step in method embodiment, for example, step S101 to S103 shown in FIG. 1.Alternatively, when the computer program is executed by processor
Realize the function of each unit in above-mentioned each human face detection device embodiment based on convolutional neural networks, such as list shown in Fig. 2
The function of member unit 31 to 35 shown in 21 to 23 and Fig. 3.
In embodiments of the present invention, after computer program is executed by processor, first using depth convolutional neural networks
The depth characteristic of facial image is extracted from facial image sample set, the depth characteristic extracted is clustered, it is each to obtain
The sample set of the different posture servant face images of kind, then human face posture grader is carried out by the sample set under different postures
Training, to obtain trained human face posture grader, so as to carry out face according to the different postures of face in image to be detected
Differentiation detection, realize the detection that becomes more meticulous of human face posture, improve the speed and accuracy of Face datection.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
All any modification, equivalent and improvement made within refreshing and principle etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of method for detecting human face based on convolutional neural networks, which is characterized in that the described method includes following step：
When receiving Face datection request, image to be detected is obtained from Face datection request；
Face is carried out at the same time to described image to be detected by advance trained first default quantity individual's face posture grader
Detection；
The face that described first default quantity individual's face posture detection of classifier goes out is marked and exported.
2. the method as described in claim 1, which is characterized in that described image to be detected is passed through into the human face posture of default quantity
Before the step of grader is carried out at the same time Face datection to described image to be detected, including：
Deep learning is carried out to the facial image in facial image sample set with deep learning method, to obtain under different postures
Sample set；
The human face posture grader built in advance is trained by the sample set under the different postures, to be trained
The default quantity individual's face posture grader of good described first.
3. method as claimed in claim 2, which is characterized in that depth is carried out to the facial image in facial image sample set
The step of habit, including：
The depth characteristic of facial image is extracted from the facial image sample set using depth convolutional neural networks；
The depth characteristic extracted is clustered by default clustering algorithm, to obtain under the various different postures
The sample set of facial image.
4. method as claimed in claim 3, which is characterized in that special to the depth extracted by default clustering algorithm
The step of sign is clustered, including：
Random selected second default quantity barycenter；
According to the described second default quantity barycenter and the position of the depth characteristic, the depth characteristic is distributed into institute's rheme
The barycenter minimum with depth characteristic distance is put, to obtain described second default the first cluster of quantity；
The average value of the described second default quantity first cluster is calculated, to update the described second default quantity described the
The barycenter of one cluster.
5. method as claimed in claim 4, which is characterized in that the depth characteristic is distributed into the position and the depth
The step of barycenter of characteristic distance minimum, including：
Pass through object functionObtain the distance of the barycenter and the depth characteristic；
The index c of the barycenter of the closest depth characteristic in the barycenter is obtained by minimizing the object functioni：=j,
Wherein, i is the variable of the depth characteristic, and f is the depth characteristic, and j is the barycenter variable, and u is the position of the barycenter,
ciIt is closest to the barycenter index of the depth characteristic, D among the barycenterijThe depth characteristic with the barycenter it
Between distance.
6. a kind of human face detection device based on convolutional neural networks, which is characterized in that described device includes：
Image receiving unit needs that detects to treat for when receiving Face datection request, receiving the Face datection request
Detection image；
Image detecting element trained first presets quantity individual's face posture grader to described to be detected in advance for passing through
Image is carried out at the same time Face datection；And
Output unit is marked, the face for going out to described first default quantity individual's face posture detection of classifier is marked simultaneously
Output.
7. device as claimed in claim 6, which is characterized in that described device further includes：
Face unit, for carrying out deep learning to the facial image in facial image sample set, to obtain different postures
Under sample set；And
Classifier training unit, it is described to obtain for being trained by the sample set of the different postures to grader
The human face posture grader of default quantity.
8. device as claimed in claim 7, which is characterized in that the data unit, including：
Feature extraction unit, for depth convolutional neural networks to be used to extract facial image from the facial image sample set
Depth characteristic；And
Feature clustering unit, for being clustered by default clustering algorithm to the depth characteristic extracted, to obtain
The sample set of various different posture servant's face images.
9. a kind of image processing terminal, including memory, processor and it is stored in the memory and can be in the processing
The computer program run on device, which is characterized in that the processor realizes such as claim 1 when performing the computer program
The step of to 5 the methods.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In realization is such as the step of claim 1 to 5 the method when the computer program is executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711319230.4A CN108062521A (en) | 2017-12-12 | 2017-12-12 | Method for detecting human face, device, terminal and medium based on convolutional neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711319230.4A CN108062521A (en) | 2017-12-12 | 2017-12-12 | Method for detecting human face, device, terminal and medium based on convolutional neural networks |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108062521A true CN108062521A (en) | 2018-05-22 |
Family
ID=62138399
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711319230.4A Pending CN108062521A (en) | 2017-12-12 | 2017-12-12 | Method for detecting human face, device, terminal and medium based on convolutional neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108062521A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109145765A (en) * | 2018-07-27 | 2019-01-04 | 华南理工大学 | Method for detecting human face, device, computer equipment and storage medium |
| CN109726648A (en) * | 2018-12-14 | 2019-05-07 | 深圳壹账通智能科技有限公司 | A kind of facial image recognition method and device based on machine learning |
| CN110458225A (en) * | 2019-08-08 | 2019-11-15 | 北京深醒科技有限公司 | A kind of vehicle detection and posture are classified joint recognition methods |
| WO2020233333A1 (en) * | 2019-05-21 | 2020-11-26 | 华为技术有限公司 | Image processing method and device |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102622589A (en) * | 2012-03-13 | 2012-08-01 | 辉路科技（北京）有限公司 | Multispectral face detection method based on graphics processing unit (GPU) |
| CN102831413A (en) * | 2012-09-11 | 2012-12-19 | 上海中原电子技术工程有限公司 | Face identification method and face identification system based on fusion of multiple classifiers |
| US20130254138A1 (en) * | 2012-03-22 | 2013-09-26 | Andrey KUDRITSKIY | System and method for viewing, modifying, storing, and running artificial neural network components |
| CN103336832A (en) * | 2013-07-10 | 2013-10-02 | 中国科学院自动化研究所 | Video classifier construction method based on quality metadata |
| CN105138956A (en) * | 2015-07-22 | 2015-12-09 | 小米科技有限责任公司 | Face detection method and device |
| CN105426963A (en) * | 2015-12-01 | 2016-03-23 | 北京天诚盛业科技有限公司 | Convolutional neural network Training method and apparatus for human face identification and application |
| CN105718857A (en) * | 2016-01-13 | 2016-06-29 | 兴唐通信科技有限公司 | Human body abnormal behavior detection method and system |
| CN105718868A (en) * | 2016-01-18 | 2016-06-29 | 中国科学院计算技术研究所 | Face detection system and method for multi-pose faces |
| CN106650694A (en) * | 2016-12-30 | 2017-05-10 | 江苏四点灵机器人有限公司 | Human face recognition method taking convolutional neural network as feature extractor |
| CN106845513A (en) * | 2016-12-05 | 2017-06-13 | 华中师范大学 | Staff detector and method based on condition random forest |
-
2017
- 2017-12-12 CN CN201711319230.4A patent/CN108062521A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102622589A (en) * | 2012-03-13 | 2012-08-01 | 辉路科技（北京）有限公司 | Multispectral face detection method based on graphics processing unit (GPU) |
| US20130254138A1 (en) * | 2012-03-22 | 2013-09-26 | Andrey KUDRITSKIY | System and method for viewing, modifying, storing, and running artificial neural network components |
| CN102831413A (en) * | 2012-09-11 | 2012-12-19 | 上海中原电子技术工程有限公司 | Face identification method and face identification system based on fusion of multiple classifiers |
| CN103336832A (en) * | 2013-07-10 | 2013-10-02 | 中国科学院自动化研究所 | Video classifier construction method based on quality metadata |
| CN105138956A (en) * | 2015-07-22 | 2015-12-09 | 小米科技有限责任公司 | Face detection method and device |
| CN105426963A (en) * | 2015-12-01 | 2016-03-23 | 北京天诚盛业科技有限公司 | Convolutional neural network Training method and apparatus for human face identification and application |
| CN105718857A (en) * | 2016-01-13 | 2016-06-29 | 兴唐通信科技有限公司 | Human body abnormal behavior detection method and system |
| CN105718868A (en) * | 2016-01-18 | 2016-06-29 | 中国科学院计算技术研究所 | Face detection system and method for multi-pose faces |
| CN106845513A (en) * | 2016-12-05 | 2017-06-13 | 华中师范大学 | Staff detector and method based on condition random forest |
| CN106650694A (en) * | 2016-12-30 | 2017-05-10 | 江苏四点灵机器人有限公司 | Human face recognition method taking convolutional neural network as feature extractor |
Non-Patent Citations (3)
| Title |
|---|
| 公安部第三研究所: "《多摄像机协同关注目标检测跟踪技术》", 30 June 2017 * |
| 胡德文 等: "《生物特征识别技术与方法》", 31 August 2013 * |
| 陈新泉: "《聚类算法中的优化方法应用》", 31 July 2014 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109145765A (en) * | 2018-07-27 | 2019-01-04 | 华南理工大学 | Method for detecting human face, device, computer equipment and storage medium |
| CN109726648A (en) * | 2018-12-14 | 2019-05-07 | 深圳壹账通智能科技有限公司 | A kind of facial image recognition method and device based on machine learning |
| WO2020233333A1 (en) * | 2019-05-21 | 2020-11-26 | 华为技术有限公司 | Image processing method and device |
| CN110458225A (en) * | 2019-08-08 | 2019-11-15 | 北京深醒科技有限公司 | A kind of vehicle detection and posture are classified joint recognition methods |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Triplet-based semantic relation learning for aerial remote sensing image change detection | |
| CN109697434B (en) | Behavior recognition method and device and storage medium | |
| KR102174595B1 (en) | System and method for identifying faces in unconstrained media | |
| CN108629306B (en) | Human body posture recognition method and device, electronic equipment and storage medium | |
| CN108960278A (en) | Use the novetly detection of the discriminator of production confrontation network | |
| CN108062521A (en) | Method for detecting human face, device, terminal and medium based on convolutional neural networks | |
| Bregonzio et al. | Fusing appearance and distribution information of interest points for action recognition | |
| CN109214273A (en) | Facial image comparison method, device, computer equipment and storage medium | |
| Faraji et al. | Face recognition under varying illuminations using logarithmic fractal dimension-based complete eight local directional patterns | |
| WO2018133791A1 (en) | Living body discrimination method and system based on video analysis, and storage medium | |
| CN109145717B (en) | Face recognition method for online learning | |
| CN102938065B (en) | Face feature extraction method and face identification method based on large-scale image data | |
| CN107784282A (en) | The recognition methods of object properties, apparatus and system | |
| CN112381011B (en) | Non-contact heart rate measurement method, system and device based on face image | |
| CN103020985B (en) | A kind of video image conspicuousness detection method based on field-quantity analysis | |
| KR20230107415A (en) | Method for identifying an object within an image and mobile device for executing the method | |
| WO2013091370A1 (en) | Human body part detection method based on parallel statistics learning of 3d depth image information | |
| CN105956570B (en) | Smiling face's recognition methods based on lip feature and deep learning | |
| CN111368672A (en) | Construction method and device for genetic disease facial recognition model | |
| CN111311635A (en) | Target positioning method, device and system | |
| JP6351243B2 (en) | Image processing apparatus and image processing method | |
| CN106203255B (en) | A kind of pedestrian based on time unifying recognition methods and system again | |
| CN107766806A (en) | A kind of method for detecting human face and system based on change of scale | |
| CN109670517A (en) | Object detection method, device, electronic equipment and target detection model | |
| CN108446586A (en) | A kind of train driver specific action detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180522 |