CN104899610A - Picture classification method and device - Google Patents
Picture classification method and device Download PDFInfo
- Publication number
- CN104899610A CN104899610A CN201510363919.1A CN201510363919A CN104899610A CN 104899610 A CN104899610 A CN 104899610A CN 201510363919 A CN201510363919 A CN 201510363919A CN 104899610 A CN104899610 A CN 104899610A
- Authority
- CN
- China
- Prior art keywords
- support vector
- vector machine
- convolutional neural
- neural networks
- original image
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Abstract
The disclosure relates to a picture classification method and device and is used for improving accuracy of picture classification. The method comprises a step of inputting original pictures into a trained convolutional neural network; a step of determining characteristic parameters of the original pictures through a full connection layer of the convolutional neural network; and a step of inputting the characteristic parameters into a trained support vector machine, and classifying the original pictures through the support vector machine. The technical scheme of the disclosure can prevent adoption of a single softmax classifier, enables pictures to have better identification effect, and improves the accuracy of picture classification.
Description
Technical field
The disclosure relates to image identification technical field, particularly relates to a kind of picture classification method and device.
Background technology
Convolutional neural networks is a kind of efficient identification algorithm being widely used in the fields such as image procossing in recent years, is a kind of structure of neural network.After the structure of convolutional neural networks is put up, the training of convolutional neural networks is identical with multilayer neural network training patterns.If directly application convolutional neural networks carries out Classification and Identification to picture, the feature obtained after last full articulamentum of convolutional neural networks is directly inputted to the softmax layer of convolutional neural networks, the sorter of softmax layer is applied the recurrence of many sorted logics and is classified, mode classification is comparatively single, causes the classification degree of accuracy of picture not high.
Summary of the invention
For overcoming Problems existing in correlation technique, disclosure embodiment provides a kind of picture classification method and device, in order to improve the degree of accuracy to picture classification.
According to the first aspect of disclosure embodiment, a kind of picture classification method is provided, comprises:
Original image is input to the convolutional neural networks after training;
The characteristic parameter of described original image is determined by the full articulamentum of described convolutional neural networks;
Described characteristic parameter is input in the support vector machine after training, by described support vector machine, described original image is classified.
In one embodiment, described described characteristic parameter is input in support vector machine, by described support vector machine, described original image is classified, can comprise:
The kernel function corresponding with described characteristic parameter is determined in described support vector machine;
According to described kernel function, described original image is classified.
In one embodiment, described method also can comprise:
Trained described convolutional neural networks by a picture group sheet, determine the full articulamentum parameter of described convolutional neural networks, described full articulamentum parameter is as the input parameter of described support vector machine;
By described full articulamentum parameter, described support vector machine is trained.
According to the second aspect of disclosure embodiment, a kind of picture classifier is provided, comprises:
Load module, is configured to original image to be input to the convolutional neural networks after training;
Determination module, is configured to the characteristic parameter being determined the described original image that described load module inputs by the full articulamentum of described convolutional neural networks;
Sort module, the described characteristic parameter being configured to be determined by described determination module is input in the support vector machine after training, is classified to described original image by described support vector machine.
In one embodiment, described sort module can comprise:
Determine submodule, be configured in described support vector machine, determine the kernel function corresponding with the described characteristic parameter that described determination module is determined;
Classification submodule, is configured to classify to described original image according to the described described kernel function determining that submodule is determined.
In one embodiment, described device also can comprise:
First training module, be configured to be trained the described convolutional neural networks that described load module needs by a picture group sheet, determine the full articulamentum parameter of described convolutional neural networks, described full articulamentum parameter is as the input parameter of described support vector machine;
Second training module, is configured to be trained described support vector machine by the described full articulamentum parameter obtained of described first training module training.
According to the third aspect of disclosure embodiment, a kind of picture classifier is provided, comprises:
Processor;
For the storer of storage of processor executable instruction;
Wherein, described processor is configured to:
Original image is input to the convolutional neural networks after training;
The characteristic parameter of described original image is determined by the full articulamentum of described convolutional neural networks;
Described characteristic parameter is input in the support vector machine after training, by described support vector machine, described original image is classified.
The technical scheme that embodiment of the present disclosure provides can comprise following beneficial effect: be input in the support vector machine after training by the characteristic parameter of the full articulamentum by convolutional neural networks, by support vector machine, original image is classified, because support vector machine has various kernel function optional, therefore more suitably kernel function can be chosen for inhomogeneous characteristic parameter, avoid and adopt single softmax sorter, thus picture can be made to have better recognition effect, improve the degree of accuracy of picture classification.
Should be understood that, it is only exemplary and explanatory that above general description and details hereinafter describe, and can not limit the disclosure.
Accompanying drawing explanation
Accompanying drawing to be herein merged in instructions and to form the part of this instructions, shows embodiment according to the invention, and is used from instructions one and explains principle of the present invention.
Figure 1A is the process flow diagram of the picture classification method according to an exemplary embodiment.
Figure 1B is the schematic diagram of convolutional neural networks according to an exemplary embodiment and support vector machine.
Fig. 2 is the process flow diagram of the picture classification method according to an exemplary embodiment one.
Fig. 3 is the process flow diagram of the picture classification method according to an exemplary embodiment two.
Fig. 4 is the block diagram of a kind of picture classifier according to an exemplary embodiment.
Fig. 5 is the block diagram of the another kind of picture classifier according to an exemplary embodiment.
Fig. 6 is a kind of block diagram being applicable to picture classifier according to an exemplary embodiment.
Embodiment
Here will be described exemplary embodiment in detail, its sample table shows in the accompanying drawings.When description below relates to accompanying drawing, unless otherwise indicated, the same numbers in different accompanying drawing represents same or analogous key element.Embodiment described in following exemplary embodiment does not represent all embodiments consistent with the present invention.On the contrary, they only with as in appended claims describe in detail, the example of apparatus and method that aspects more of the present invention are consistent.
Figure 1A is the process flow diagram of the picture classification method according to an exemplary embodiment, and Figure 1B is the schematic diagram of convolutional neural networks according to an exemplary embodiment and support vector machine; This picture classification method can be applied on terminal device (such as: smart mobile phone, panel computer, desk-top computer), can be realized by the mode of the mode or mounting software on the desktop of installing application on smart mobile phone or panel computer, as shown in Figure 1A, this picture classification method comprises the following steps S101-S 103:
In step S101, original image is input to the convolutional neural networks after training.
In one embodiment, can be trained convolutional neural networks by a picture group sheet, obtain the convolutional neural networks after training.
In step s 102, by the characteristic parameter of the full articulamentum determination original image of convolutional neural networks.
In one embodiment, full articulamentum is positioned at the layer second from the bottom of convolutional neural networks, after convolutional neural networks trains, obtains the characteristic parameter extracted through convolutional neural networks after full articulamentum.
In step s 103, characteristic parameter is input in the support vector machine after training, by support vector machine, original image is classified.
In one embodiment, after having trained, all nodes of the layer second from the bottom of convolutional neural networks are input in support vector machine as characteristic parameter, in one embodiment, the characteristic parameter being input to support vector machine is the vector of the value composition of all nodes on full articulamentum.
As shown in Figure 1B, the network that the disclosure uses comprises: input layer 11, convolutional neural networks 12 and support vector connect 13, wherein, input layer 11 can input the pixel value of original image, all nodes of the layer second from the bottom of convolutional neural networks 12 are input in support vector machine 13 as characteristic parameter, it will be understood by those skilled in the art that, Figure 1B is only illustratively unrestricted, contained as long as all nodes of the layer second from the bottom of convolutional neural networks 12 can be input to support vector machine 13 as characteristic parameter thus realize being the disclosure to the implementation that original image is classified.
In the present embodiment, be input in the support vector machine after training by the characteristic parameter of the full articulamentum by convolutional neural networks, by support vector machine, original image is classified, because support vector machine has various kernel function optional, therefore more suitably kernel function can be chosen for inhomogeneous characteristic parameter, avoid and adopt single softmax sorter, thus picture can be made to have better recognition effect, improve the degree of accuracy of picture classification.
In one embodiment, characteristic parameter is input in support vector machine, by support vector machine, original image is classified, can comprise:
The kernel function corresponding with characteristic parameter is determined in support vector machine;
According to kernel function, original image is classified.
In one embodiment, method also can comprise:
Trained convolutional neural networks by a picture group sheet, determine the full articulamentum parameter of convolutional neural networks, full articulamentum parameter is as the input parameter of support vector machine;
By full articulamentum parameter, support vector machine is trained.
Specifically how picture is classified, please refer to subsequent embodiment.
So far, the said method that disclosure embodiment provides, can avoid adopting single softmax sorter, make picture have better recognition effect, improve the degree of accuracy of picture classification.
With specific embodiment, the technical scheme that disclosure embodiment provides is described below.
Fig. 2 is the process flow diagram of the picture classification method according to an exemplary embodiment one; The said method that the present embodiment utilizes disclosure embodiment to provide, how to determine that the kernel function in support vector machine carries out exemplary illustration, as shown in Figure 2, comprises the steps:
In step s 201, original image is input to the convolutional neural networks after training.
In step S202, by the characteristic parameter of the full articulamentum determination original image of convolutional neural networks.
The description of step S201 and step S202 refers to the description of above-mentioned steps S101 and step S102, no longer describes in detail secondary,
In step S203, in support vector machine, determine the kernel function corresponding with characteristic parameter.
In step S204, according to kernel function, original image is classified.
In one embodiment, in order to ensure original image, there is better discrimination, because support vector machine comprises linear kernel function and RBF kernel function, therefore the disclosure can be determined and characteristic parameter more suitably kernel function for inhomogeneous characteristic parameter, avoids and adopts single softmax sorter.Such as, if the more sparse or obvious linear separability of characteristic parameter itself, linear kernel function then both can have been adopted also can to adopt RBF kernel function, linear kernel to be classified the situation of general situation or obvious characteristic linearly inseparable, then can adopt RBF kernel function, thus guarantee, to the classification of original image, there is more significant effect.
In the present embodiment, by the kernel function corresponding with characteristic parameter, original image is classified, avoid and adopt single softmax sorter, make picture have better recognition effect, improve the degree of accuracy of picture classification.
Fig. 3 is the process flow diagram of the picture classification method according to an exemplary embodiment two; The said method that the present embodiment utilizes disclosure embodiment to provide, carries out exemplary illustration how to be trained for example to convolutional neural networks and support vector machine, as shown in Figure 3, comprises the steps:
In step S301, trained convolutional neural networks by a picture group sheet, determine the full articulamentum parameter of convolutional neural networks, full articulamentum parameter is as the input parameter of support vector machine.
In step s 302, by full articulamentum parameter, support vector machine is trained.
In one embodiment, the training patterns in correlation technique can be adopted to train a picture group sheet by convolutional neural networks, after convolutional neural networks has been trained, convergence result is the value of each parameter in convolutional neural networks, by the layer second from the bottom of convolutional neural networks (namely, full articulamentum) all nodes be input in support vector machine as characteristic parameter, and then support vector machine is trained, because the layer second from the bottom of convolutional neural networks is as the input feature vector of support vector machine, and the label (label) of training set corresponding to input feature vector, thus the parameter of the sorter of support vector machine can be obtained by training.
In the present embodiment, by training convolutional neural networks and support vector machine, thus convolutional neural networks and support vector machine are combined, realize the classification to original image, because the characteristic parameter obtained after last full articulamentum of convolutional neural networks can directly input in support vector machine, improve the discriminator intensity to original image, avoid and the characteristic parameter that full articulamentum obtains is input to softmax layer, adopt many sorted logics to return to classify, improve the recognition effect of original image thus, improve the degree of accuracy of picture classification.
Fig. 4 is the block diagram of a kind of picture classifier according to an exemplary embodiment, and as shown in Figure 4, picture classifier comprises:
Load module 41, is configured to original image to be input to the convolutional neural networks after training;
Determination module 42, is configured to the characteristic parameter of the original image of full articulamentum determination load module 41 input by convolutional neural networks;
Sort module 43, is configured to the characteristic parameter that determination module 42 is determined to be input in the support vector machine after training, is classified by support vector machine to original image.
Fig. 5 is the block diagram of the another kind of picture classifier according to an exemplary embodiment, and as shown in Figure 5, on above-mentioned basis embodiment illustrated in fig. 4, in one embodiment, sort module 43 can comprise:
Determine submodule 431, be configured in support vector machine, determine the kernel function corresponding with the characteristic parameter that determination module 42 is determined;
Classification submodule 432, is configured to according to determining that the kernel function that submodule 431 is determined is classified to original image.
In one embodiment, device also can comprise:
First training module 44, be configured to be trained the convolutional neural networks that load module 41 needs by a picture group sheet, determine the full articulamentum parameter of convolutional neural networks, full articulamentum parameter is as the input parameter of support vector machine;
Second training module 45, the full articulamentum parameter obtained being configured to be trained by the first training module 44 is trained support vector machine.
About the device in above-described embodiment, wherein the concrete mode of modules executable operations has been described in detail in about the embodiment of the method, will not elaborate explanation herein.
Fig. 6 is a kind of block diagram being applicable to picture classifier according to an exemplary embodiment.Such as, device 600 can be mobile phone, computing machine, digital broadcast terminal, messaging devices, game console, tablet device, Medical Devices, body-building equipment, personal digital assistant etc.
With reference to Fig. 6, device 600 can comprise following one or more assembly: processing components 602, storer 604, power supply module 606, multimedia groupware 608, audio-frequency assembly 610, the interface 612 of I/O (I/O), sensor module 614, and communications component 616.
The integrated operation of the usual control device 600 of processing components 602, such as with display, call, data communication, camera operation and record operate the operation be associated.Treatment element 602 can comprise one or more processor 620 to perform instruction, to complete all or part of step of above-mentioned method.In addition, processing components 602 can comprise one or more module, and what be convenient between processing components 602 and other assemblies is mutual.Such as, processing element 602 can comprise multi-media module, mutual with what facilitate between multimedia groupware 608 and processing components 602.
Storer 604 is configured to store various types of data to be supported in the operation of equipment 600.The example of these data comprises for any application program of operation on device 600 or the instruction of method, contact data, telephone book data, message, picture, video etc.Storer 604 can be realized by the volatibility of any type or non-volatile memory device or their combination, as static RAM (SRAM), Electrically Erasable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory EPROM (EPROM), programmable read only memory (PROM), ROM (read-only memory) (ROM), magnetic store, flash memory, disk or CD.
The various assemblies that electric power assembly 606 is device 600 provide electric power.Electric power assembly 606 can comprise power-supply management system, one or more power supply, and other and the assembly generating, manage and distribute electric power for device 600 and be associated.
Multimedia groupware 608 is included in the screen providing an output interface between described device 600 and user.In certain embodiments, screen can comprise liquid crystal display (LCD) and touch panel (TP).If screen comprises touch panel, screen may be implemented as touch-screen, to receive the input signal from user.Touch panel comprises one or more touch sensor with the gesture on sensing touch, slip and touch panel.Described touch sensor can the border of not only sensing touch or sliding action, but also detects the duration relevant to described touch or slide and pressure.In certain embodiments, multimedia groupware 608 comprises a front-facing camera and/or post-positioned pick-up head.When equipment 600 is in operator scheme, during as screening-mode or video mode, front-facing camera and/or post-positioned pick-up head can receive outside multi-medium data.Each front-facing camera and post-positioned pick-up head can be fixing optical lens systems or have focal length and optical zoom ability.
Audio-frequency assembly 610 is configured to export and/or input audio signal.Such as, audio-frequency assembly 610 comprises a microphone (MIC), and when device 600 is in operator scheme, during as call model, logging mode and speech recognition mode, microphone is configured to receive external audio signal.The sound signal received can be stored in storer 604 further or be sent via communications component 616.In certain embodiments, audio-frequency assembly 610 also comprises a loudspeaker, for output audio signal.
I/O interface 612 is for providing interface between processing components 602 and peripheral interface module, and above-mentioned peripheral interface module can be keyboard, some striking wheel, button etc.These buttons can include but not limited to: home button, volume button, start button and locking press button.
Sensor module 614 comprises one or more sensor, for providing the state estimation of various aspects for device 600.Such as, sensor module 614 can detect the opening/closing state of equipment 600, the relative positioning of assembly, such as described assembly is display and the keypad of device 600, the position of all right pick-up unit 600 of sensor module 614 or device 600 1 assemblies changes, the presence or absence that user contacts with device 600, the temperature variation of device 600 orientation or acceleration/deceleration and device 600.Sensor module 614 can comprise proximity transducer, be configured to without any physical contact time detect near the existence of object.Sensor module 614 can also comprise optical sensor, as CMOS or ccd image sensor, for using in imaging applications.In certain embodiments, this sensor module 614 can also comprise acceleration transducer, gyro sensor, Magnetic Sensor, pressure transducer or temperature sensor.
Communications component 616 is configured to the communication being convenient to wired or wireless mode between device 600 and other equipment.Device 600 can access the wireless network based on communication standard, as WiFi, 2G or 3G, or their combination.In one exemplary embodiment, communication component 616 receives from the broadcast singal of external broadcasting management system or broadcast related information via broadcast channel.In one exemplary embodiment, described communication component 616 also comprises near-field communication (NFC) module, to promote junction service.Such as, can based on radio-frequency (RF) identification (RFID) technology in NFC module, Infrared Data Association (IrDA) technology, ultra broadband (UWB) technology, bluetooth (BT) technology and other technologies realize.
In the exemplary embodiment, device 600 can be realized, for performing said method by one or more application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing appts (DSPD), programmable logic device (PLD) (PLD), field programmable gate array (FPGA), controller, microcontroller, microprocessor or other electronic components.
In the exemplary embodiment, additionally provide a kind of non-transitory computer-readable recording medium comprising instruction, such as, comprise the storer 604 of instruction, above-mentioned instruction can perform said method by the processor 620 of device 600.Such as, described non-transitory computer-readable recording medium can be ROM, random access memory (RAM), CD-ROM, tape, floppy disk and optical data storage devices etc.
Those skilled in the art, at consideration instructions and after putting into practice disclosed herein disclosing, will easily expect other embodiment of the present disclosure.The application is intended to contain any modification of the present disclosure, purposes or adaptations, and these modification, purposes or adaptations are followed general principle of the present disclosure and comprised the undocumented common practise in the art of the disclosure or conventional techniques means.Instructions and embodiment are only regarded as exemplary, and true scope of the present disclosure and spirit are pointed out by claim below.
Should be understood that, the disclosure is not limited to precision architecture described above and illustrated in the accompanying drawings, and can carry out various amendment and change not departing from its scope.The scope of the present disclosure is only limited by appended claim.
Claims (7)
1. a picture classification method, is characterized in that, described method comprises:
Original image is input to the convolutional neural networks after training;
The characteristic parameter of described original image is determined by the full articulamentum of described convolutional neural networks;
Described characteristic parameter is input in the support vector machine after training, by described support vector machine, described original image is classified.
2. method according to claim 1, is characterized in that, is describedly input in support vector machine by described characteristic parameter, is classified, comprising by described support vector machine to described original image:
The kernel function corresponding with described characteristic parameter is determined in described support vector machine;
According to described kernel function, described original image is classified.
3. method according to claim 1, is characterized in that, described method also comprises:
Trained described convolutional neural networks by a picture group sheet, determine the full articulamentum parameter of described convolutional neural networks, described full articulamentum parameter is as the input parameter of described support vector machine;
By described full articulamentum parameter, described support vector machine is trained.
4. a picture classifier, is characterized in that, described device comprises:
Load module, is configured to original image to be input to the convolutional neural networks after training;
Determination module, is configured to the characteristic parameter being determined the described original image that described load module inputs by the full articulamentum of described convolutional neural networks;
Sort module, the described characteristic parameter being configured to be determined by described determination module is input in the support vector machine after training, is classified to described original image by described support vector machine.
5. device according to claim 4, is characterized in that, described sort module comprises:
Determine submodule, be configured in described support vector machine, determine the kernel function corresponding with the described characteristic parameter that described determination module is determined;
Classification submodule, is configured to classify to described original image according to the described described kernel function determining that submodule is determined.
6. device according to claim 4, is characterized in that, described device also comprises:
First training module, be configured to be trained the described convolutional neural networks that described load module needs by a picture group sheet, determine the full articulamentum parameter of described convolutional neural networks, described full articulamentum parameter is as the input parameter of described support vector machine;
Second training module, is configured to be trained described support vector machine by the described full articulamentum parameter obtained of described first training module training.
7. a picture classifier, is characterized in that, described device comprises:
Processor;
For the storer of storage of processor executable instruction;
Wherein, described processor is configured to:
Original image is input to the convolutional neural networks after training;
The characteristic parameter of described original image is determined by the full articulamentum of described convolutional neural networks;
Described characteristic parameter is input in the support vector machine after training, by described support vector machine, described original image is classified.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510363919.1A CN104899610A (en) | 2015-06-26 | 2015-06-26 | Picture classification method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510363919.1A CN104899610A (en) | 2015-06-26 | 2015-06-26 | Picture classification method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104899610A true CN104899610A (en) | 2015-09-09 |
Family
ID=54032264
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510363919.1A Pending CN104899610A (en) | 2015-06-26 | 2015-06-26 | Picture classification method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104899610A (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105335712A (en) * | 2015-10-26 | 2016-02-17 | 小米科技有限责任公司 | Image recognition method, device and terminal |
CN105335754A (en) * | 2015-10-29 | 2016-02-17 | 小米科技有限责任公司 | Character recognition method and device |
CN106326931A (en) * | 2016-08-25 | 2017-01-11 | 南京信息工程大学 | Mammary gland molybdenum target image automatic classification method based on deep learning |
CN106650795A (en) * | 2016-12-01 | 2017-05-10 | 携程计算机技术(上海)有限公司 | Sorting method of hotel room type images |
CN106778918A (en) * | 2017-01-22 | 2017-05-31 | 北京飞搜科技有限公司 | A kind of deep learning image identification system and implementation method for being applied to mobile phone terminal |
CN106815598A (en) * | 2016-12-15 | 2017-06-09 | 歌尔科技有限公司 | A kind of recognition methods of 360 degree of panoramic pictures and device |
CN107292353A (en) * | 2017-08-09 | 2017-10-24 | 广东工业大学 | A kind of fruit tree classification method and system |
CN107341190A (en) * | 2017-06-09 | 2017-11-10 | 努比亚技术有限公司 | Picture screening technique, terminal and computer-readable recording medium |
CN109781268A (en) * | 2019-03-16 | 2019-05-21 | 福州大学 | Keypoint part temperature monitoring system in a kind of switchgear based on the infrared thermovision technology of low cost |
WO2019127927A1 (en) * | 2017-12-29 | 2019-07-04 | 深圳云天励飞技术有限公司 | Neural network chip, method of using neural network chip to implement de-convolution operation, electronic device, and computer readable storage medium |
CN110321936A (en) * | 2019-06-14 | 2019-10-11 | 浙江鹏信信息科技股份有限公司 | A method of realizing that picture two is classified based on VGG16 and SVM |
US20230080164A1 (en) * | 2017-12-20 | 2023-03-16 | Alpvision S.A. | Authentication machine learning from multiple digital presentations |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7171042B2 (en) * | 2000-12-04 | 2007-01-30 | Intel Corporation | System and method for classification of images and videos |
CN101859377A (en) * | 2010-06-08 | 2010-10-13 | 杭州电子科技大学 | Electromyographic signal classification method based on multi-kernel support vector machine |
CN103679185A (en) * | 2012-08-31 | 2014-03-26 | 富士通株式会社 | Convolutional neural network classifier system as well as training method, classifying method and application thereof |
CN104318252A (en) * | 2014-11-02 | 2015-01-28 | 西安电子科技大学 | Hyperspectral image classification method based on stratified probability model |
-
2015
- 2015-06-26 CN CN201510363919.1A patent/CN104899610A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7171042B2 (en) * | 2000-12-04 | 2007-01-30 | Intel Corporation | System and method for classification of images and videos |
CN101859377A (en) * | 2010-06-08 | 2010-10-13 | 杭州电子科技大学 | Electromyographic signal classification method based on multi-kernel support vector machine |
CN103679185A (en) * | 2012-08-31 | 2014-03-26 | 富士通株式会社 | Convolutional neural network classifier system as well as training method, classifying method and application thereof |
CN104318252A (en) * | 2014-11-02 | 2015-01-28 | 西安电子科技大学 | Hyperspectral image classification method based on stratified probability model |
Non-Patent Citations (5)
Title |
---|
XIAO-XIAO NIU ET AL: "A novel hybrid CNN–SVM classifier for recognizing handwritten digits", 《PATTERN RECOGNITION》 * |
何小萍: "改进的支持向量机分类算法在语音识别中的应用研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
张吉斌: "基于图像处理及支持向量机的车牌识别技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
张艳岩: "基于支持向量机的网络舆情危机预警研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
程然: "最小二乘支持向量机的研究和应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105335712A (en) * | 2015-10-26 | 2016-02-17 | 小米科技有限责任公司 | Image recognition method, device and terminal |
CN105335754A (en) * | 2015-10-29 | 2016-02-17 | 小米科技有限责任公司 | Character recognition method and device |
CN106326931A (en) * | 2016-08-25 | 2017-01-11 | 南京信息工程大学 | Mammary gland molybdenum target image automatic classification method based on deep learning |
CN106650795B (en) * | 2016-12-01 | 2020-06-12 | 携程计算机技术(上海)有限公司 | Hotel room type image sorting method |
CN106650795A (en) * | 2016-12-01 | 2017-05-10 | 携程计算机技术(上海)有限公司 | Sorting method of hotel room type images |
CN106815598A (en) * | 2016-12-15 | 2017-06-09 | 歌尔科技有限公司 | A kind of recognition methods of 360 degree of panoramic pictures and device |
CN106778918B (en) * | 2017-01-22 | 2020-10-30 | 苏州飞搜科技有限公司 | Deep learning image recognition system applied to mobile phone terminal and implementation method |
CN106778918A (en) * | 2017-01-22 | 2017-05-31 | 北京飞搜科技有限公司 | A kind of deep learning image identification system and implementation method for being applied to mobile phone terminal |
CN107341190A (en) * | 2017-06-09 | 2017-11-10 | 努比亚技术有限公司 | Picture screening technique, terminal and computer-readable recording medium |
CN107292353A (en) * | 2017-08-09 | 2017-10-24 | 广东工业大学 | A kind of fruit tree classification method and system |
US20230080164A1 (en) * | 2017-12-20 | 2023-03-16 | Alpvision S.A. | Authentication machine learning from multiple digital presentations |
WO2019127927A1 (en) * | 2017-12-29 | 2019-07-04 | 深圳云天励飞技术有限公司 | Neural network chip, method of using neural network chip to implement de-convolution operation, electronic device, and computer readable storage medium |
US11010661B2 (en) | 2017-12-29 | 2021-05-18 | Shenzhen Intellifusion Technologies Co., Ltd. | Neural network chip, method of using neural network chip to implement de-convolution operation, electronic device, and computer readable storage medium |
CN109781268A (en) * | 2019-03-16 | 2019-05-21 | 福州大学 | Keypoint part temperature monitoring system in a kind of switchgear based on the infrared thermovision technology of low cost |
CN109781268B (en) * | 2019-03-16 | 2021-06-01 | 福州大学 | Switch cabinet interior key part temperature monitoring system based on low-cost infrared hot-vision technology |
CN110321936A (en) * | 2019-06-14 | 2019-10-11 | 浙江鹏信信息科技股份有限公司 | A method of realizing that picture two is classified based on VGG16 and SVM |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104899610A (en) | Picture classification method and device | |
CN105069786A (en) | Straight line detection method and straight line detection device | |
CN105528606A (en) | Region identification method and device | |
CN105119812A (en) | Method and apparatus for changing emoticons at chat interface, and terminal equipment | |
CN105512685A (en) | Object identification method and apparatus | |
CN104881104A (en) | Intelligent equipment, intelligent equipment performance improvement method and intelligent equipment performance improvement device | |
CN104298547A (en) | Terminal setting method and device | |
CN104238912A (en) | Application control method and application control device | |
CN105427233A (en) | Method and device for removing watermark | |
CN104918107A (en) | Video file identification processing method and device | |
CN105139378A (en) | Card boundary detection method and apparatus | |
CN105242942A (en) | Application control method and apparatus | |
CN105354560A (en) | Fingerprint identification method and device | |
CN105260360A (en) | Named entity identification method and device | |
CN104537380A (en) | Clustering method and device | |
CN104331228A (en) | Screen locking method and device | |
CN105139033A (en) | Classifier construction method and device and image processing method and device | |
CN105159496A (en) | Touch event response method and mobile terminal | |
CN105653165A (en) | Method and device for regulating character display | |
CN105354017A (en) | Information processing method and apparatus | |
CN105120155A (en) | Panoramic photograph generation method and device | |
CN105430715A (en) | Method and apparatus for controlling WIFI scanning | |
CN105426878A (en) | Method and device for face clustering | |
CN104484867A (en) | Picture processing method and device | |
CN105488829A (en) | Method and device for generating head portrait |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150909 |
|
RJ01 | Rejection of invention patent application after publication |