CN106897695A - A kind of image recognizing and processing equipment, system and method - Google Patents
A kind of image recognizing and processing equipment, system and method Download PDFInfo
- Publication number
- CN106897695A CN106897695A CN201710103695.XA CN201710103695A CN106897695A CN 106897695 A CN106897695 A CN 106897695A CN 201710103695 A CN201710103695 A CN 201710103695A CN 106897695 A CN106897695 A CN 106897695A
- Authority
- CN
- China
- Prior art keywords
- image
- face
- deep learning
- recognition
- processing equipment
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/94—Hardware or software architectures specially adapted for image or video understanding
- G06V10/95—Hardware or software architectures specially adapted for image or video understanding structured as a network, e.g. client-server architectures
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of image recognizing and processing equipment, system and method, the image recognizing and processing equipment is deployed with small-sized deep learning neutral net, image is obtained using image acquisition unit, and treatment is identified to the image for obtaining to recognize whether face using small-sized deep learning network, the image of acquisition is sent to cloud processor when identifying and there is facial image, the present invention can significantly mitigate the treatment load and network interface bandwidth of cloud processor.
Description
Technical field
The present invention relates to image identification technical field, more particularly to a kind of image recognizing and processing equipment, system and method.
Background technology
Current deep learning algorithm reaches accuracy rate very high in field of image recognition, but due to can accurately know
The deep learning network of other image needs substantial amounts of computing resource, so with current computer technology, it is necessary to deep learning
Beyond the clouds, and local device is limited due to CPU disposal abilities for network design, it is impossible to recognize image well.
But large-scale deep learning network is placed on high in the clouds, there are two big defects：Firstly, because high in the clouds and local hardware are one
To many relations, when local hardware has high in the clouds to process request simultaneously, then the computing resource requirement to high in the clouds is very high；Secondly, it is right
The network interface bandwidth in high in the clouds also require that it is very high because the transmission of image needs very big bandwidth.
The content of the invention
To overcome the shortcomings of that above-mentioned prior art is present, the purpose of the present invention is to provide a kind of image recognition processing dress
Put, system and method, to mitigate the treatment load and network interface bandwidth of cloud processor.
It is that, up to above-mentioned purpose, the present invention proposes a kind of image recognizing and processing equipment, and the image recognizing and processing equipment is deployed with
Small-sized deep learning neutral net, image is obtained using image acquisition unit, and using small-sized deep learning network to acquisition
Image is identified treatment to recognize whether face, and the image of acquisition is sent into cloud when identifying and there is facial image
End processor.
Further, the image recognizing and processing equipment includes：
Image acquisition unit, for obtaining image；
Face identification unit, for the image obtained to the image acquisition unit using the small-sized deep learning neutral net
Recognition of face treatment is carried out, whether there is face in the image for recognizing acquisition；
Judging unit, for judging the recognition result of the face identification unit with the presence or absence of face, and in judging the people
When the recognition result of face recognition unit has face, start transmission unit；
Transmission unit, the image of the presence face for the image acquisition unit to be obtained is sent to cloud processor.
Further, the small-sized deep learning neutral net uses two sorter networks, to recognize that image whether there is people
Face.
Further, the image recognizing and processing equipment is local intelligent hardware, and the image acquisition unit is camera.
To reach above-mentioned purpose, the present invention also provides a kind of image recognition processing system, including：
Image recognizing and processing equipment, is deployed with small-sized deep learning neutral net thereon, is obtained using image acquisition unit
Image, and treatment is identified to the image for obtaining to recognize whether face using small-sized deep learning network, in identification
The image of acquisition is sent to cloud processor when going out to exist facial image；
Cloud processor, the image of transmission is set for receiving the image recognition processing, using large-scale deep learning network
Image to obtaining carries out recognition of face, identifies face type.
Further, the cloud processor has deep learning network training module, and the face of training in advance typing is regarded
Frequency or image, be converted into human face data, human face data and the other information of the human face data are carried out it is corresponding, by the knot after correspondence
Fruit is characterized as label with a certain in information, and classification is preserved into personal information database, and deep learning is generated with above-mentioned label
Each level parameter in network, forms extraction characteristic parameter and is preserved.
Further, the cloud processor extracts face when acquisition image carries out recognition of face from the image for obtaining
Data, search in personal information database saved human face data and carry out match cognization, output matching recognition result.
To reach above-mentioned purpose, the present invention also provides a kind of recognition processing method, comprises the following steps：
Step one, image recognizing and processing equipment obtains image using image acquisition unit；
Step 2, image recognition is carried out using small-sized deep learning neutral net to the image for obtaining, to determine what is obtained
Whether there is face in image；
Step 3, when there is face in the image for judging to obtain, cloud processor is sent to by the image of acquisition；
Step 4, the cloud processor carries out recognition of face using large-scale deep learning neutral net to the image for obtaining,
Identify face type.
Further, the small-sized deep learning neutral net is two sorter networks, and only the image for obtaining is divided into face
With the class of no face two.
Further, the large-scale deep learning neutral net is complicated many sorter networks.
Compared with prior art, a kind of image recognizing and processing equipment of the invention, system and method are by local image
Small-sized deep learning network is disposed on recognition process unit, in large-scale deep learning network is disposed in cloud processor, using this
Small-sized deep learning network, to judge whether face, will just be obtained to the image recognition for obtaining after judging to have face
Image send to cloud processor, face type is gone out with the large-scale deep learning Network Recognition using cloud processor, significantly
Alleviate the treatment load and network interface bandwidth of cloud processor.
Brief description of the drawings
Fig. 1 is a kind of system architecture diagram of image recognizing and processing equipment of the invention；
Fig. 2 is a kind of system architecture diagram of image recognition processing system of the invention；
Fig. 3 is the system architecture diagram of the image recognition processing system that the specific embodiment of the invention is used；
Fig. 4 is the image recognition processing process schematic of the specific embodiment of the invention；
The step of Fig. 5 is a kind of recognition processing method of the invention flow chart.
Specific embodiment
Below by way of specific instantiation and embodiments of the present invention are described with reference to the drawings, those skilled in the art can
Further advantage of the invention and effect are understood by content disclosed in the present specification easily.The present invention also can be different by other
Instantiation implemented or applied, the various details in this specification also can based on different viewpoints with application, without departing substantially from
Various modifications and change are carried out under spirit of the invention.
Fig. 1 is a kind of system architecture diagram of image recognizing and processing equipment of the invention.A kind of image recognition processing dress of the present invention
Put, dispose a small-sized deep learning neutral net, image is obtained using image acquisition unit, and using small-sized deep learning nerve
Network to obtain image be identified treatment to recognize whether face, in identify there is facial image when by obtain
Image is sent to cloud processor, as shown in figure 1, the image recognizing and processing equipment of the present invention includes：Image acquisition unit 101,
Face identification unit 102, judging unit 103 and transmission unit 104.
Wherein, image acquisition unit 101, for obtaining image, in the specific embodiment of the invention, image acquisition unit
101 can be camera head, the pick-up lens of such as video camera or the camera of smart mobile phone；Face identification unit 102, using this
Small-sized deep learning neutral net carries out recognition of face treatment to the image that image acquisition unit 101 is obtained, to recognize acquisition
Whether there is face in image, because recognition of face here uses prior art, will not be described here；Judging unit
103, for judging the recognition result of face identification unit 102 with the presence or absence of face, and in judging face identification unit 102
When recognition result has face, start transmission unit 104；Transmission unit 104, for depositing of obtaining image acquisition unit 101
Cloud processor is sent in the image of face.
Fig. 2 is a kind of system architecture diagram of image recognition processing system of the invention.As shown in Fig. 2 a kind of image of the invention
Identification processing system, including：Image recognizing and processing equipment 20 and cloud processor 30.
Wherein, image recognition processing sets 20 and is arranged at local side, and a small-sized deep learning neutral net is deployed with thereon,
Image recognition processing sets 20 and obtains image using image acquisition unit, and utilizes the small-sized deep learning neutral net disposed
To obtain image carry out recognition of face process to recognize whether face, in identify there is facial image when by obtain
Image is sent to cloud processor 30.
Cloud processor 30, the image of 20 transmission is set for receiving image recognition processing, using large-scale deep learning god
Recognition of face is carried out to the image for obtaining through network, face type is identified.Specifically, cloud processor 30 is with depth
Network training module is practised, by the face video or image of training in advance typing, human face data is converted into, by human face data and the people
The other information (such as mother, father or grandfather, grandmother) of face data carries out correspondence, by the result after correspondence with certain in information
One is characterized as label, and classification is preserved into personal information database, with each in above-mentioned label generation deep learning neutral net
Level parameter, forms extraction characteristic parameter and is preserved, when cloud processor 30 obtains image carries out recognition of face, from acquisition
Image in extract the data of face, search in personal information database saved human face data and carry out match cognization,
Output matching recognition result.
Hereinafter the present invention will be further illustrated by a specific embodiment：In the specific embodiment of the invention, its solution
Scene be face to be identified in actual video and is positioned, for example say as the intelligent camera of image recognizing and processing equipment
Machine is put at home, it is necessary to recognize father, mother, grandfather, grandmother, son automatically in the video flowing that camera is got, in order to
The real-time of guarantee system, it is assumed that require that the disposal ability of hardware reaches 2 frames/second in the present embodiment.
According to the scheme of prior art, intelligent camera captures 2 two field pictures each second, is then transmitted by Internet
To cloud processor, the type for identifying face is processed by cloud processor, be possible to face in certain image, it is also possible to do not have
Have face, if any, must identify father, mother or other people.Assuming that cloud processor must be simultaneously 10,000
Platform Intelligent hardware is serviced, then his disposal ability must reach 20000 frames/second.
Fig. 3 is the system architecture diagram of the image recognition processing system that the specific embodiment of the invention is used.Of the invention specific
In embodiment, in small-sized deep learning neutral net is disposed on intelligent camera, large-scale depth is disposed on processor beyond the clouds
Practise neutral net, small-sized deep learning neutral net here can only be distinguished in image either with or without face.Divide the image into two
Class, has face and without face, i.e., simple two sorter network；Whether it is father that large-scale deep learning neutral net can be identified
Father or mother or other people face, i.e., complicated many sorter networks.It is below at the image recognition of the specific embodiment of the invention
Reason process (as shown in Figure 4)：
1st, the image acquisition unit of intelligent camera collects view data with real-time 30frame/s；
2nd, view data is transmitted to the face identification unit of intelligent camera；
3rd, miniature neural network is had on intelligent camera, is two sorter networks, people or inhuman can be distinguished.(due to intelligence
CPU disposal abilities on video camera are limited, only set up miniature neural network)
4th, when face identification unit is inhuman using the obtained image of two sorter networks determination, differentiates and terminate, no longer pass
Further differentiation is done to cloud processor.
5th, when face identification unit is people using the obtained image of two sorter networks determination, further it is transmitted at high in the clouds
Reason device does further differentiation,.
6th, cloud processor is entered when the image of intelligent camera transmission is received using large-scale deep learning neutral net
One step carries out recognition of face, and whom likes with determination pair, such as the people in image is father, mother, grandfather, grandmother or son.
In the specific embodiment of the invention, the image that camera is obtained is 30frames/s, so at miniature neural network
The speed of reason is 30fps, because most images are all " inhuman ", so the image speed of the large-scale Processing with Neural Network on cloud
Rate will be far smaller than 30fps.
It can be seen that, in the present invention, when Intelligent hardware (such as video camera) capture images, Intelligent hardware (such as video camera) is first
First pass through local small-sized deep learning neutral net to discern whether with the presence of face, if it has, then sending the images at high in the clouds
Reason device, if it is not, image is abandoned, because in normal video, the scene for having face to occur is simultaneously few, is so sent to cloud
Hold the image of processor then to greatly reduce, can significantly mitigate the treatment load and network interface bandwidth of cloud processor.
The step of Fig. 5 is a kind of recognition processing method of the invention flow chart.As shown in figure 5, a kind of image of the invention
Identifying processing method, comprises the following steps：
Step 501, image recognizing and processing equipment obtains image, in the present invention, image recognition using image acquisition unit
Processing unit is local device, and image acquisition unit uses camera.
Step 502, image recognition is carried out using small-sized deep learning neutral net to the image for obtaining, to determine what is obtained
Whether there is face in image.In the specific embodiment of the invention, the small-sized deep learning neutral net is simple two classification
Network, i.e., only divide the image into two classes, has face and without face.
Step 503, when there is face in the image for judging to obtain, cloud processor is sent to by the image of acquisition.
Step 504, cloud processor carries out recognition of face using large-scale deep learning neutral net to the image for obtaining, and knows
Do not go out face type.In the specific embodiment of the invention, the large-scale deep learning neutral net is complicated many sorter networks, you can
To identify the type of the face in image, e.g. father or mother or other people face.
In sum, a kind of image recognizing and processing equipment of the invention, system and method are by local image recognition
Small-sized deep learning network is disposed on reason device, in large-scale deep learning network is disposed in cloud processor, using the small moldeed depth
Learning network is spent to the image recognition for obtaining to judge whether face, the image that will be just obtained after judging to have face
Send to cloud processor, face type is gone out with the large-scale deep learning Network Recognition using cloud processor, significantly reduce
The treatment load and network interface bandwidth of cloud processor.
Any those skilled in the art can repair under without prejudice to spirit and scope of the invention to above-described embodiment
Decorations and change.Therefore, the scope of the present invention, should be as listed by claims.
Claims (10)
1. a kind of image recognizing and processing equipment, it is characterised in that：The image recognizing and processing equipment is deployed with small-sized deep learning god
Through network, image is obtained using image acquisition unit, and place is identified to the image for obtaining using small-sized deep learning network
The image of acquisition is sent to cloud processor by reason to recognize whether face when identifying and there is facial image.
2. a kind of image recognizing and processing equipment as claimed in claim 1, it is characterised in that the image recognizing and processing equipment bag
Include：
Image acquisition unit, for obtaining image；
Face identification unit, for being carried out to the image that the image acquisition unit is obtained using the small-sized deep learning neutral net
Recognition of face is processed, whether there is face in the image for recognizing acquisition；
Judging unit, for judging the recognition result of the face identification unit with the presence or absence of face, and in judging that the face knows
When the recognition result of other unit has face, start transmission unit；
Transmission unit, the image of the presence face for the image acquisition unit to be obtained is sent to cloud processor.
3. a kind of image recognizing and processing equipment as claimed in claim 2, it is characterised in that：The small-sized deep learning nerve net
Network uses two sorter networks, to recognize that image whether there is face.
4. a kind of image recognizing and processing equipment as claimed in claim 3, it is characterised in that：The image recognizing and processing equipment is this
Ground Intelligent hardware, the image acquisition unit is camera.
5. a kind of image recognition processing system, including：
Image recognizing and processing equipment, is deployed with small-sized deep learning neutral net thereon, and image is obtained using image acquisition unit,
And treatment is identified to the image for obtaining to recognize whether face using small-sized deep learning network, in identifying presence
The image of acquisition is sent to cloud processor during facial image；
Cloud processor, the image of transmission is set for receiving the image recognition processing, using large-scale deep learning network to obtaining
The image for taking carries out recognition of face, identifies face type.
6. a kind of image recognition processing system as claimed in claim 5, it is characterised in that：The cloud processor has depth
Network training module is practised, by the face video or image of training in advance typing, human face data is converted into, by human face data and the people
The other information of face data carries out correspondence, and result after correspondence is characterized as into label with a certain in information, and classification is preserved to people
In member's information database, with each level parameter in above-mentioned label generation deep learning network, form extraction characteristic parameter and give
Preserve.
7. a kind of image recognition processing system as claimed in claim 6, it is characterised in that：The cloud processor is in acquisition image
When carrying out recognition of face, the data of face are extracted from the image for obtaining, saved people in lookup personal information database
Face data simultaneously carry out match cognization, output matching recognition result.
8. a kind of recognition processing method, comprises the following steps：
Step one, image recognizing and processing equipment obtains image using image acquisition unit；
Step 2, image recognition is carried out using small-sized deep learning neutral net to the image for obtaining, with the image for determining to obtain
In whether there is face；
Step 3, when there is face in the image for judging to obtain, cloud processor is sent to by the image of acquisition；
Step 4, the cloud processor carries out recognition of face using large-scale deep learning neutral net to the image for obtaining, and recognizes
Go out face type.
9. a kind of recognition processing method as claimed in claim 8, it is characterised in that：The small-sized deep learning neutral net
It is two sorter networks, only the image for obtaining is divided into has face and without the class of face two.
10. a kind of recognition processing method as claimed in claim 8, it is characterised in that：The large-scale deep learning nerve net
Network is complicated many sorter networks.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710103695.XA CN106897695A (en) | 2017-02-24 | 2017-02-24 | A kind of image recognizing and processing equipment, system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710103695.XA CN106897695A (en) | 2017-02-24 | 2017-02-24 | A kind of image recognizing and processing equipment, system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106897695A true CN106897695A (en) | 2017-06-27 |
Family
ID=59184101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710103695.XA Pending CN106897695A (en) | 2017-02-24 | 2017-02-24 | A kind of image recognizing and processing equipment, system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106897695A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108712630A (en) * | 2018-04-19 | 2018-10-26 | 安凯(广州)微电子技术有限公司 | A kind of internet camera system and its implementation based on deep learning |
| CN108985276A (en) * | 2018-08-21 | 2018-12-11 | 盯盯拍（深圳）技术股份有限公司 | Vision AI algorithmic system design method, vision AI algorithmic system design device and vision AI algorithmic system |
| CN109413152A (en) * | 2018-09-25 | 2019-03-01 | 上海瑾盛通信科技有限公司 | Image processing method, device, storage medium and electronic equipment |
| WO2019042230A1 (en) * | 2017-08-31 | 2019-03-07 | 深圳市商汤科技有限公司 | Method, system, photographing device, and computer storage medium for facial image search |
| CN109685088A (en) * | 2017-10-18 | 2019-04-26 | 上海仪电（集团）有限公司中央研究院 | Narrow band communication intelligent image analysis system based on cloud separation convolutional neural networks |
| CN109919244A (en) * | 2019-03-18 | 2019-06-21 | 北京字节跳动网络技术有限公司 | Method and apparatus for generating scene Recognition model |
| CN111698418A (en) * | 2020-04-17 | 2020-09-22 | 广州市讯思视控科技有限公司 | Industrial intelligent camera system based on deep learning configuration cloud platform |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140079297A1 (en) * | 2012-09-17 | 2014-03-20 | Saied Tadayon | Application of Z-Webs and Z-factors to Analytics, Search Engine, Learning, Recognition, Natural Language, and Other Utilities |
| CN105117624A (en) * | 2015-06-03 | 2015-12-02 | 深圳市美贝壳科技有限公司 | Recognition system and method of user system based on face recognition |
| CN105426875A (en) * | 2015-12-18 | 2016-03-23 | 武汉科技大学 | Face identification method and attendance system based on deep convolution neural network |
| CN105426857A (en) * | 2015-11-25 | 2016-03-23 | 小米科技有限责任公司 | Training method and device of face recognition model |
| CN105787437A (en) * | 2016-02-03 | 2016-07-20 | 东南大学 | Vehicle brand type identification method based on cascading integrated classifier |
| CN106204780A (en) * | 2016-07-04 | 2016-12-07 | 武汉理工大学 | A kind of based on degree of depth study and the human face identification work-attendance checking system and method for cloud service |
| CN106250832A (en) * | 2016-07-25 | 2016-12-21 | 华南理工大学 | A kind of national recognition methods based on integrated convolutional neural networks |
-
2017
- 2017-02-24 CN CN201710103695.XA patent/CN106897695A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140079297A1 (en) * | 2012-09-17 | 2014-03-20 | Saied Tadayon | Application of Z-Webs and Z-factors to Analytics, Search Engine, Learning, Recognition, Natural Language, and Other Utilities |
| CN105117624A (en) * | 2015-06-03 | 2015-12-02 | 深圳市美贝壳科技有限公司 | Recognition system and method of user system based on face recognition |
| CN105426857A (en) * | 2015-11-25 | 2016-03-23 | 小米科技有限责任公司 | Training method and device of face recognition model |
| CN105426875A (en) * | 2015-12-18 | 2016-03-23 | 武汉科技大学 | Face identification method and attendance system based on deep convolution neural network |
| CN105787437A (en) * | 2016-02-03 | 2016-07-20 | 东南大学 | Vehicle brand type identification method based on cascading integrated classifier |
| CN106204780A (en) * | 2016-07-04 | 2016-12-07 | 武汉理工大学 | A kind of based on degree of depth study and the human face identification work-attendance checking system and method for cloud service |
| CN106250832A (en) * | 2016-07-25 | 2016-12-21 | 华南理工大学 | A kind of national recognition methods based on integrated convolutional neural networks |
Non-Patent Citations (1)
| Title |
|---|
| 张瑞: "基于卷积神经网络的Logo检测与识别研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019042230A1 (en) * | 2017-08-31 | 2019-03-07 | 深圳市商汤科技有限公司 | Method, system, photographing device, and computer storage medium for facial image search |
| US11182594B2 (en) | 2017-08-31 | 2021-11-23 | Shenzhen Sensetime Technology Co., Ltd. | Face image retrieval methods and systems, photographing apparatuses, and computer storage media |
| CN109685088A (en) * | 2017-10-18 | 2019-04-26 | 上海仪电（集团）有限公司中央研究院 | Narrow band communication intelligent image analysis system based on cloud separation convolutional neural networks |
| CN108712630A (en) * | 2018-04-19 | 2018-10-26 | 安凯(广州)微电子技术有限公司 | A kind of internet camera system and its implementation based on deep learning |
| CN108985276A (en) * | 2018-08-21 | 2018-12-11 | 盯盯拍（深圳）技术股份有限公司 | Vision AI algorithmic system design method, vision AI algorithmic system design device and vision AI algorithmic system |
| CN109413152A (en) * | 2018-09-25 | 2019-03-01 | 上海瑾盛通信科技有限公司 | Image processing method, device, storage medium and electronic equipment |
| CN109413152B (en) * | 2018-09-25 | 2021-02-26 | 上海瑾盛通信科技有限公司 | Image processing method, image processing device, storage medium and electronic equipment |
| CN109919244A (en) * | 2019-03-18 | 2019-06-21 | 北京字节跳动网络技术有限公司 | Method and apparatus for generating scene Recognition model |
| CN111698418A (en) * | 2020-04-17 | 2020-09-22 | 广州市讯思视控科技有限公司 | Industrial intelligent camera system based on deep learning configuration cloud platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106897695A (en) | A kind of image recognizing and processing equipment, system and method | |
| WO2020151489A1 (en) | Living body detection method based on facial recognition, and electronic device and storage medium | |
| CN108596041A (en) | A kind of human face in-vivo detection method based on video | |
| CN111144284B (en) | Method and device for generating depth face image, electronic equipment and medium | |
| CN108513110A (en) | Recognition of face monitoring camera | |
| CN107463887A (en) | Train driver gesture intelligence inspection system and intelligent inspection method | |
| CN113191216B (en) | Multi-user real-time action recognition method and system based on posture recognition and C3D network | |
| EP3823267A1 (en) | Static video recognition | |
| CN106407906B (en) | Face face recognition method | |
| CN110222647A (en) | A kind of human face in-vivo detection method based on convolutional neural networks | |
| CN107665328A (en) | Face face recognition method | |
| WO2018121552A1 (en) | Palmprint data based service processing method, apparatus and program, and medium | |
| CN110348272A (en) | Method, apparatus, system and the medium of dynamic human face identification | |
| CN114173158B (en) | Face recognition method, cloud device, client device, electronic device and medium | |
| CN210721506U (en) | Dynamic face recognition terminal based on 3D camera | |
| CN114332983A (en) | Face image definition detection method, face image definition detection device, electronic equipment and medium | |
| CN111461005B (en) | Gesture recognition method and device, computer equipment and storage medium | |
| CN110163043B (en) | Face detection method, device, storage medium and electronic device | |
| CN109214354B (en) | Dynamic face retrieval and identification method based on 3D camera | |
| JP6836985B2 (en) | Programs, devices and methods for estimating the context of human behavior from captured images | |
| CN208522852U (en) | Recognition of face crime scene investigation device | |
| CN208691422U (en) | Recognition of face monitoring camera | |
| CN112784631A (en) | Method for recognizing face emotion based on deep neural network | |
| CN110119706A (en) | One kind being based on cloud iris identification method and device | |
| CN110348285B (en) | Social relationship identification method and device based on semantic enhanced network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170627 |