CN112116008A - Target detection model processing method based on intelligent decision and related equipment thereof - Google Patents
Target detection model processing method based on intelligent decision and related equipment thereof Download PDFInfo
- Publication number
- CN112116008A CN112116008A CN202010990029.4A CN202010990029A CN112116008A CN 112116008 A CN112116008 A CN 112116008A CN 202010990029 A CN202010990029 A CN 202010990029A CN 112116008 A CN112116008 A CN 112116008A
- Authority
- CN
- China
- Prior art keywords
- target detection
- detection model
- model
- model parameters
- central server
- 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.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 307
- 238000003672 processing method Methods 0.000 title claims abstract description 22
- 238000012549 training Methods 0.000 claims abstract description 89
- 239000002131 composite material Substances 0.000 claims abstract description 71
- 238000000034 method Methods 0.000 claims abstract description 31
- 238000003860 storage Methods 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000001960 triggered effect Effects 0.000 claims description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 238000004891 communication Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000004590 computer program Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 239000000284 extract Substances 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 239000003086 colorant Substances 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/602—Providing cryptographic facilities or services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Bioethics (AREA)
- General Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Computer Security & Cryptography (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the application belongs to the field of artificial intelligence, and relates to a target detection model processing method and device based on intelligent decision, computer equipment and a storage medium, wherein the method comprises the following steps: training the initial multi-target detection model according to the obtained local data set to obtain a relay multi-target detection model; calculating composite model parameters according to the generated additional random numbers and the model parameters of the relay multi-target detection model; sending the composite model parameters to a central server to instruct the central server to generate global model parameters according to the composite model parameters of each node; receiving global model parameters to update the relay multi-target detection model; and performing iterative training by taking the updated relay multi-target detection model as an initial multi-target detection model of next training until the model converges to obtain the multi-target detection model. In addition, the present application also relates to a blockchain technique, and the local data set can be stored in the blockchain. The target detection accuracy is improved.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a target detection model processing method based on intelligent decision and related equipment thereof.
Background
With the development of artificial intelligence, the target detection is more and more widely applied in life and production. The target detection relates to a detection model in intelligent decision-making, and can input an image into the target detection model, process the image by the target detection model and output a target object in the image. For example, in a garbage classification application, a garbage picture can be input into a target detection model to identify garbage, so as to guide people to recycle and classify the identified garbage.
In the conventional target detection technology, a target detection model can only focus on a single target for detection during detection, and when the target detection model is trained, because the data volume of a local data set is usually limited, the target detection model is difficult to be trained sufficiently, so that the detection accuracy of the target detection model is low.
Disclosure of Invention
An object of the embodiments of the present application is to provide a method and an apparatus for processing a target detection model based on an intelligent decision, a computer device, and a storage medium, so as to solve the problem of low detection accuracy of the target detection model.
In order to solve the foregoing technical problem, an embodiment of the present application provides a processing method for a target detection model based on an intelligent decision, where the target detection model is a multi-target detection model, and the following technical solutions are adopted:
acquiring a local data set and an initial multi-target detection model;
training the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model;
generating an additional random number, and calculating a composite model parameter according to the additional random number and the model parameter of the relay multi-target detection model;
sending the composite model parameters to a central server to instruct the central server to generate global model parameters according to the composite model parameters of each node;
receiving the global model parameters from the central server to update the relay multi-target detection model;
and performing iterative training by taking the updated relay multi-target detection model as an initial multi-target detection model of next training until the model converges to obtain the multi-target detection model.
Further, before the step of obtaining the local data set and the initial multi-target detection model, the method further includes:
obtaining global model parameters from a central server;
and constructing an initial multi-target detection model according to the global model parameters.
Further, the step of training the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model includes:
inputting the target image in the local data set into the initial multi-target detection model to obtain a target prediction result;
determining a prediction error according to the target prediction result and the image label in the local data set;
performing parameter adjustment on the initial multi-target detection model based on the prediction error;
and performing iterative training by taking the initial multi-target detection model after parameter adjustment as an initial multi-target detection model for next training until the iteration times reach a preset value, thereby obtaining a relay multi-target detection model.
Further, the step of generating an additional random number and calculating a composite model parameter according to the additional random number and the model parameter of the relay multi-target detection model includes:
generating an additional random number; wherein, the added value of the additional random numbers generated by each node in the alliance network is zero;
and performing linear operation on the generated additional random number and the model parameters of the relay multi-target detection model to obtain composite model parameters.
Further, the step of sending the composite model parameters to a central server to instruct the central server to generate global model parameters according to the composite model parameters of each node includes:
communicating with a central server to determine an encryption key;
encrypting the composite model parameter according to the encryption key to obtain an encryption model parameter;
and sending the encryption model parameters to the central server to instruct the central server to decrypt the encryption model parameters of each node, and calculating according to the decrypted composite model parameters to generate global model parameters.
Further, after the step of performing iterative training by using the updated relay multi-target detection model as an initial multi-target detection model for next training until the model converges to obtain the multi-target detection model, the method further includes:
acquiring an image to be detected;
inputting the image to be detected into the multi-target detection model to obtain a target object in the image to be detected;
and displaying the detected target object.
Further, after the step of displaying the detected target object, the method further includes:
when a triggered calibration instruction is received, displaying a calibration information input page;
acquiring an image to be calibrated and calibration indication information which are input in the calibration information input page;
generating a calibration data set according to the image to be calibrated and the calibration indication information;
and carrying out calibration training on the multi-target detection model through the calibration data set.
In order to solve the foregoing technical problem, an embodiment of the present application further provides a processing apparatus for a target detection model based on an intelligent decision, where the target detection model is a multi-target detection model, and the following technical solutions are adopted:
the acquisition module is used for acquiring a local data set and an initial multi-target detection model;
the model training module is used for training the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model;
the parameter calculation module is used for generating an additional random number and calculating a composite model parameter according to the additional random number and the model parameter of the relay multi-target detection model;
the parameter sending module is used for sending the composite model parameters to a central server so as to instruct the central server to generate global model parameters according to the composite model parameters of each node;
a model updating module for receiving the global model parameters from the central server to update the relay multi-target detection model;
and the iterative training module is used for performing iterative training by taking the updated relay multi-target detection model as an initial multi-target detection model for next training until the model converges to obtain the multi-target detection model.
In order to solve the above technical problem, an embodiment of the present application further provides a computer device, which includes a memory and a processor, where the memory stores a computer program, and the processor implements the steps of the processing method for an intelligent decision-based object detection model described above when executing the computer program.
In order to solve the above technical problem, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and the computer program, when executed by a processor, implements the steps of the processing method for an intelligent decision-based object detection model described above.
Compared with the prior art, the embodiment of the application mainly has the following beneficial effects: firstly, training an initial multi-target detection model according to a local data set to obtain a relay multi-target detection model; generating an additional random number, wherein the additional random number is used for calculating the model parameters of the relay multi-target detection model so as to encrypt the model parameters; the encrypted conforming model parameters are sent to a central server in the alliance network, the central server generates global model parameters according to the composite model parameters of all the nodes, and the global model parameters are issued to a terminal so as to update a relay multi-target detection model and carry out iterative training; the global model parameters are obtained on the basis of the local data sets of all the terminals, the characteristics of the local data sets are fused, the training of the multi-target detection model by the data sets in a larger scale is realized on the basis of protecting the local data sets, and the accuracy of the multi-target detection model detection obtained after the training is finished is improved.
Drawings
In order to more clearly illustrate the solution of the present application, the drawings needed for describing the embodiments of the present application will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and that other drawings can be obtained by those skilled in the art without inventive effort.
FIG. 1 is an exemplary system architecture diagram in which the present application may be applied;
FIG. 2 is a flow diagram of one embodiment of a method for intelligent decision-based object detection model processing according to the present application;
FIG. 3 is a schematic block diagram of one embodiment of an intelligent decision-based object detection model processing apparatus according to the present application;
FIG. 4 is a schematic block diagram of one embodiment of a computer device according to the present application.
Detailed Description
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "including" and "having," and any variations thereof, in the description and claims of this application and the description of the above figures are intended to cover non-exclusive inclusions. The terms "first," "second," and the like in the description and claims of this application or in the above-described drawings are used for distinguishing between different objects and not for describing a particular order.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.
As shown in fig. 1, the system architecture 100 may include terminals 101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the terminals 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.
The user may use the terminals 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminals 101, 102, 103 may have various communication client applications installed thereon, such as a web browser application, a shopping application, a search application, an instant messaging tool, a mailbox client, social platform software, and the like.
The terminals 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to a smart phone, a tablet computer, an e-book reader, an MP3 player (Moving Picture experts Group Audio Layer III, motion Picture experts compression standard Audio Layer 3), an MP4 player (Moving Picture experts Group Audio Layer IV, motion Picture experts compression standard Audio Layer 4), a laptop portable computer, a desktop computer, and the like.
The server 105 may be a server providing various services, such as a background server providing support for pages displayed on the terminals 101, 102, 103.
It should be noted that, the processing method of the target detection model based on the intelligent decision provided in the embodiment of the present application is generally executed by a terminal, and accordingly, the processing device of the target detection model based on the intelligent decision is generally disposed in the terminal.
It should be understood that the number of terminals, networks, and servers in fig. 1 are merely illustrative. There may be any number of terminals, networks, and servers, as desired for an implementation.
With continued reference to FIG. 2, a flow diagram of one embodiment of a method of processing an intelligent decision-based object detection model in accordance with the present application is shown. The target detection model is a multi-target detection model, and the processing method of the target detection model based on the intelligent decision comprises the following steps:
step S201, a local data set and an initial multi-target detection model are obtained.
In this embodiment, an electronic device (for example, the terminal shown in fig. 1) on which the processing method of the target detection model based on intelligent decision is operated may communicate with other terminals or servers through a wired connection manner or a wireless connection manner. It should be noted that the wireless connection means may include, but is not limited to, a 3G/4G connection, a WiFi connection, a bluetooth connection, a WiMAX connection, a Zigbee connection, a uwb (ultra wideband) connection, and other wireless connection means now known or developed in the future.
Wherein the local data set may be a data set stored in the terminal; the initial multi-target detection model may be an initial multi-target detection model.
Specifically, the terminal may be installed with an application program for target detection, and after the application program is started, the terminal loads the initial multi-target detection model located locally and obtains a local data set stored in the terminal.
The multi-target detection model is used, and the multi-target detection model can detect a plurality of targets at one time, so that the target detection efficiency is improved. In one embodiment, the multi-target detection model may be a RetinaNet network.
The multi-target detection model is a lightweight model, can be deployed in various terminals besides a server, and a holder of the terminal can expand a local data set according to the own needs, for example, the local data set can be expanded by taking pictures or acquiring an image expansion data set from the internet, so that the expansion difficulty of the local data set is reduced, and the data volume of the local data set is enriched.
In one embodiment, the processing method of the target detection model based on intelligent decision-making can also be executed by a server, and the server loads the initial multi-target detection model and acquires a local data set. And after the training of the server is finished, a target detection interface is provided, and a user can call the target detection interface at the terminal to perform target detection.
It is emphasized that, to further ensure the privacy and security of the local data set, the local data set may also be stored in a node of a blockchain.
The block chain referred by the application is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
And S202, training the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model.
Specifically, the terminal trains the initial multi-target detection model according to the local data set, and stops training when the training stop condition is met, so that the relay multi-target detection model is obtained. The training stopping condition may be that the number of iterations in the training reaches a preset value, or that a prediction error obtained in the training is smaller than a preset error threshold.
The relay multi-target detection model obtained by the terminal can be convergent or non-convergent.
And step S203, generating an additional random number, and calculating a composite model parameter according to the additional random number and the model parameter of the relay multi-target detection model.
The additional random number can be a parameter for realizing an encryption function on a model parameter of the relay multi-target detection model.
Specifically, the terminal performs federated learning, and the terminal may be one node in a federated network. Each node in the alliance network generates an additional random number, and the added value of the obtained additional random numbers is zero. And the terminal extracts the model parameters of the relay multi-target detection model, calculates the additional random number and the model parameters to obtain composite model parameters so as to protect the data privacy of the model parameters in the local relay multi-target detection model.
And step S204, sending the composite model parameters to a central server to instruct the central server to generate global model parameters according to the composite model parameters of each node.
The central server may be a server playing a role in central control in federal learning, and is used for instructing each node to perform federal learning.
Specifically, the terminal sends the composite model parameters to the central server. After the central server receives the composite model parameters of each node, linear operation can be performed on the composite model parameters of each model.
And step S205, receiving the global model parameters from the central server to update the relay multi-target detection model.
Specifically, the central server sends the global model parameters to the nodes. And after receiving the global model parameters, the terminal updates the local relay multi-target detection model according to the global model parameters, specifically, the model parameters in the relay multi-target detection model are replaced by the global model parameters.
And S206, performing iterative training by taking the updated relay multi-target detection model as an initial multi-target detection model for next training until the model converges to obtain the multi-target detection model.
Specifically, after the terminal updates the relay multi-target detection model, the relay multi-target detection model is used as an initial multi-target detection model, and the initial multi-target detection model obtained is continuously trained according to the local data set, namely, the steps S202 to S206 are iterated until the model converges, and the terminal stops training to obtain the multi-target detection model. The condition of model convergence may be that a prediction error obtained in the training is smaller than a preset error threshold.
In one embodiment, gradient information is communicated between the terminal and the central server. And the terminal calculates the additional random number and the model gradient of the relay multi-target detection model to obtain a conforming model gradient, and sends the conforming model gradient to the central server. And the central server accumulates the gradients conforming to the model, then calculates an average value, and issues the average value serving as a global average gradient to each node so as to update the relay multi-target detection model.
In one embodiment, when each node in the alliance network achieves model convergence, each node stops training, and the relay multi-target detection model when the training is stopped is used as the multi-target detection model of each node.
In the embodiment, the initial multi-target detection model is trained according to the local data set to obtain a relay multi-target detection model; generating an additional random number, wherein the additional random number is used for calculating the model parameters of the relay multi-target detection model so as to encrypt the model parameters; the encrypted conforming model parameters are sent to a central server in the alliance network, the central server generates global model parameters according to the composite model parameters of all the nodes, and the global model parameters are issued to a terminal so as to update a relay multi-target detection model and carry out iterative training; the global model parameters are obtained on the basis of the local data sets of all the terminals, the characteristics of the local data sets are fused, the training of the multi-target detection model by the data sets in a larger scale is realized on the basis of protecting the local data sets, and the accuracy of the multi-target detection model detection obtained after the training is finished is improved.
Further, before step S201, the method may further include: obtaining global model parameters from a central server; and constructing an initial multi-target detection model according to the global model parameters.
Specifically, the terminal needs to be initialized before training. During initialization, the terminal receives the global model parameters from the central server, and the global model parameters obtained at this time are the initialized global model parameters, and may be randomly generated by the central server, for example. And the terminal replaces the model parameters stored in the local multi-target detection model with the obtained global model parameters, so that the initial multi-target detection model is obtained.
In this embodiment, the terminal constructs an initial multi-target detection model according to the global model parameters issued by the central server, thereby implementing model initialization.
Further, the step S202 may include: inputting the target image in the local data set into an initial multi-target detection model to obtain a target prediction result; determining a prediction error according to a target prediction result and an image label in a local data set; adjusting parameters of the initial multi-target detection model based on the prediction error; and performing iterative training by taking the initial multi-target detection model after parameter adjustment as an initial multi-target detection model for next training until the iteration times reach a preset value, thereby obtaining a relay multi-target detection model.
Wherein the target image may be an image about the target object.
Specifically, the terminal extracts a target image and an image label from the local data set respectively, inputs the target image into the initial multi-target detection model, and obtains a target prediction result. And the terminal calculates a prediction error according to the target prediction result and the image label so as to reduce the prediction error and adjust the model parameters in the initial multi-target detection model as a target.
And the terminal carries out iterative training on the initial multi-target detection model after parameter adjustment according to the local data set, once model parameters are adjusted by the terminal, one iteration is realized, and the terminal stops iteration until the iteration times reach a preset value to obtain the relay multi-target detection model.
In one embodiment, the terminal adopts a Focal local Loss function when calculating the prediction error, and the Focal local Loss function is as follows:
where y is the image label and γ is the adjustment factor.
In the embodiment, iterative training is performed on the initial multi-target detection model for the preset times according to the target images and the image labels in the local data set to obtain the relay multi-target detection model, and the relay multi-target detection model is used for federal learning, so that realization of federal learning is guaranteed.
Further, the step S203 may include: generating an additional random number; wherein, the added value of the additional random numbers generated by each node in the alliance network is zero; and performing linear operation on the generated additional random number and the model parameters of the relay multi-target detection model to obtain composite model parameters.
Specifically, the terminal generates an additional random number. The terminal is a node in the alliance network, each node in the alliance network generates an additional random number, and the added random numbers of the nodes are zero. And the terminal extracts the model parameters of the relay multi-target detection model, and performs linear operation on the model parameters and the generated additional random number to obtain the composite model parameters. The composite model parameters are different from the model parameters of the relay multi-target detection model, and the data confidentiality effect on the relay multi-target detection model can be achieved.
In one embodiment, the terminal performs addition and subtraction operation on the generated additional random number and the model parameters of the relay multi-target detection model, and the generated composite model parameters are as follows:
wherein, ω iskIn order to relay the model parameters of the multi-target detection model,
for composite model parameters, SuvTo append the random number, K is the set of all nodes in the federated network, and u and v represent nodes in the federated network, both being elements in K.
In this embodiment, after the additional random number is generated, the additional random number and the model parameter of the relay multi-target detection model are subjected to linear operation to obtain a composite model parameter, so that data encryption of the model parameter is realized.
Further, the step S204 may include: communicating with a central server to determine an encryption key; encrypting the composite model parameter according to the encryption key to obtain an encryption model parameter; and sending the encryption model parameters to a central server to instruct the central server to decrypt the encryption model parameters of each node, and calculating according to the decrypted composite model parameters to generate global model parameters.
The encryption key may be a key for encrypting the composite model parameter.
Specifically, the terminal may communicate with the central server in advance, and determine the encryption Key with the central server based on a DH Key Exchange protocol/Algorithm (Diffie-Hellman Key Exchange protocol/Algorithm, Diffie-Hellman Key Exchange/encryption Algorithm, which may enable two parties requiring secure communication to determine a shared Key by the method).
When the terminal sends the composite model parameters to the central server, the composite model parameters can be encrypted again according to the encryption key to obtain encryption model parameters, and the encryption model parameters are sent to the central server, so that data privacy safety in communication with the central server is guaranteed.
And after the central server obtains the encryption model parameters of each node, decrypting the encryption model parameters according to the decryption key to obtain the composite model parameters of each node. The central server performs addition operation on the composite model parameters of each node, so that the additional random numbers in each group of composite model parameters return to zero, namely:
wherein, ω iskIn order to relay the model parameters of the multi-target detection model,
for the composite model parameter, K is the set of all nodes in the federation network.
And the central server performs weighted linear operation on the composite model parameters to obtain global model parameters.
When calculating the global model parameters, the weights of all groups of composite model parameters can be the same or different; when the global model parameters are not the same, the central server can calculate the global model parameters based on the safety aggregation algorithm of the FedAvg, and the formula is as follows:
wherein, ω iskModel parameters for the relay multi-target detection model, omega being a global model parameter, nkThe data volume of the kth node is, n is the total data volume of each node in the alliance network, K is the set of all nodes in the alliance network, K represents the kth node, and t represents the tth update.
In the embodiment, the composite model parameter is encrypted according to the encryption key to obtain the encryption model parameter, so that the data privacy in federal learning is further protected; after the encryption model parameters are sent to the central server, the central server decrypts the encryption model parameters and generates global model parameters according to the composite model parameters obtained after decryption, and the global model parameters are used for updating the relay multi-target detection model in each node, so that realization of federal learning is guaranteed.
Further, step S206 may further include: acquiring an image to be detected; inputting an image to be detected into a multi-target detection model to obtain a target object in the image to be detected; and displaying the detected target object.
The image to be detected can be an image used for inputting a multi-target detection model to carry out target detection.
Specifically, when the multi-target detection model is applied, a user can operate the terminal, acquire an image to be detected through an image acquisition device of the terminal, or select an image stored in the terminal as the image to be detected, and instruct the terminal to perform target detection on the image to be detected.
The terminal inputs the image to be detected into the multi-target detection model, processes the image to be detected through the multi-target detection model, identifies a target object in the image to be detected, and displays the target object through a screen. When the target object is displayed, the multi-target detection model can add a detection frame and an object description to the target object, the object description is used for displaying the category of the target object, and the detection frame can select different colors according to the category of the target object so as to more clearly display the information of the target object.
In one embodiment, when the user clicks the displayed target object, the terminal may further display the introduction and related information of the category of target object. For example, when the multi-target detection model is applied to garbage classification, the terminal can display garbage articles in an image to be detected, a user clicks the detected garbage articles, the terminal can introduce garbage categories to which the garbage articles belong, and provides classification suggestions of the garbage articles in the categories, so that the user can better classify the garbage.
In the embodiment, the multi-target detection model for performing the target detection on the image to be detected is obtained through federal learning, and a rich data set is used for training in the federal learning, so that the accuracy of the target detection is improved.
Further, after the displaying the detected target object, the method may further include: when a triggered calibration instruction is received, displaying a calibration information input page; acquiring an image to be calibrated and calibration indication information which are input in a calibration information input page; generating a calibration data set according to the image to be calibrated and the calibration indication information; and carrying out calibration training on the multi-target detection model through the calibration data set.
Specifically, the terminal provides a model calibration function, and when a user thinks that the detection result of the multi-target detection model is inaccurate, the user can click a virtual calibration button in a display page to trigger a calibration instruction. And after receiving the calibration instruction, the terminal displays a calibration information input page and indicates a user to input an image to be calibrated and calibration indication information in the calibration information input page. The terminal can also directly take the image to be detected in the target detection as the image to be calibrated. The user can input characters to describe information such as the position, shape, color, size, name and the like of the target object, or the target object is directly circled in the image to be calibrated in a frame recognition mode through the screen to obtain calibration indicating information.
The terminal generates a calibration data set according to the image to be calibrated and the calibration indication information, the calibration data set comprises the image to be calibrated and the corresponding image label, and the calibration data set is used for performing calibration training on the multi-target detection model so as to improve the detection accuracy of the multi-target detection model.
The terminal can only carry out calibration training on the multi-target detection model locally; the multi-target detection model can be calibrated and trained immediately in a federal learning mode; or after the preset time or after the calibration training is performed locally for the preset times, the multi-target detection model can be trained in a federal learning mode.
In the embodiment, when the calibration instruction is received, the calibration information input page is displayed, and the calibration data set is generated according to the image to be calibrated and the calibration indication information input in the calibration information input page, so that the multi-target detection model is calibrated and trained, and the detection accuracy of the multi-target detection model is improved.
The following describes a processing method of the target detection model based on intelligent decision making according to a specific embodiment. Specifically, a garbage classification application is installed in the terminal, and a user can take garbage pictures through the terminal to expand the local data set. And the terminal acquires the global model parameters from the central server to obtain an initialized multi-target detection model. And training the initialized multi-target detection model for N (N is an integer larger than zero) rounds according to the local data set to obtain the relay multi-target detection model.
The terminal determines the keys required for communication with the central server by means of the DH key exchange protocol. And the terminal generates an additional random number, and adds the additional random number and the model parameters of the relay multi-target detection model to obtain the composite model parameters. And the terminal encrypts the composite model parameters by using the encryption key to obtain encryption model parameters, and sends the encryption model parameters to the central server.
The central server decrypts the encrypted model parameters to obtain composite model parameters of each node, and adds the composite model parameters to eliminate the influence of the additional random number. The central server can calculate the global model parameters according to the safety aggregation algorithm of the FedAvg, and sends the global model parameters to each node, so that each node updates the relay multi-target detection model according to the global model parameters.
And the terminal carries out iterative training on the updated relay multi-target detection model according to the local data set until the model converges to obtain the multi-target detection model.
When the garbage detection system is applied, a user can photograph garbage articles through the terminal to obtain an image to be detected, the multi-target detection image can identify a plurality of garbage articles in the image to be detected at one time, and the identified garbage articles are displayed. When displaying the garbage articles, the garbage articles can be marked by the detection frames, and the categories of the garbage articles are displayed, for example, the garbage articles are displayed to be recyclable or harmful garbage, and different types of garbage articles can be distinguished by using the detection frames with different colors.
When the user thinks that the target detection is accurate, namely the garbage classification is correct, the user can click the detected garbage articles, the terminal introduces the garbage articles of the category and displays the garbage classification description so as to better classify the garbage.
When the user considers that the target detection is inaccurate, namely the garbage classification is wrong, the user can click the virtual calibration button, an image which is considered by the user to be inaccurate in detection is uploaded on a calibration information input page and serves as an image to be calibrated, explanatory words are input and serve as calibration indication information, and the terminal conducts calibration training on the multi-target detection model according to the image to be calibrated and the calibration indication information.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware associated with computer readable instructions, which can be stored in a computer readable storage medium, and when executed, the processes of the embodiments of the methods described above can be included. The storage medium may be a non-volatile storage medium such as a magnetic disk, an optical disk, a Read-Only Memory (ROM), or a Random Access Memory (RAM).
It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless explicitly stated herein. Moreover, at least a portion of the steps in the flow chart of the figure may include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed alternately or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
With further reference to fig. 3, as an implementation of the method shown in fig. 2, the present application provides an embodiment of an intelligent decision-based target detection model processing apparatus, where the target detection model is a multi-target detection model, and the embodiment of the apparatus corresponds to the embodiment of the method shown in fig. 2, and the apparatus can be applied in various electronic devices.
As shown in fig. 3, the processing apparatus 300 for an intelligent decision-based target detection model according to this embodiment includes: an obtaining module 301, a model training module 302, a parameter calculating module 303, a parameter sending module 304, a model updating module 305, and an iterative training module 306, wherein:
an obtaining module 301, configured to obtain a local data set and an initial multi-target detection model.
And the model training module 302 is configured to train the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model.
And the parameter calculation module 303 is configured to generate an additional random number, and calculate a composite model parameter according to the additional random number and the model parameter of the relay multi-target detection model.
And a parameter sending module 304, configured to send the composite model parameters to the central server, so as to instruct the central server to generate global model parameters according to the composite model parameters of each node.
A model update module 305 for receiving global model parameters from the central server to update the relay multi-target detection model.
And the iterative training module 306 is configured to perform iterative training by using the updated relay multi-target detection model as an initial multi-target detection model for next training until the model converges to obtain a multi-target detection model.
In the embodiment, the initial multi-target detection model is trained according to the local data set to obtain a relay multi-target detection model; generating an additional random number, wherein the additional random number is used for calculating the model parameters of the relay multi-target detection model so as to encrypt the model parameters; the encrypted conforming model parameters are sent to a central server in the alliance network, the central server generates global model parameters according to the composite model parameters of all the nodes, and the global model parameters are issued to a terminal so as to update a relay multi-target detection model and carry out iterative training; the global model parameters are obtained on the basis of the local data sets of all the terminals, the characteristics of the local data sets are fused, the training of the multi-target detection model by the data sets in a larger scale is realized on the basis of protecting the local data sets, and the accuracy of the multi-target detection model detection obtained after the training is finished is improved.
In some optional implementations of the present embodiment, the processing device 300 for the intelligent decision-based target detection model further includes: parameter acquisition module and model construction module, wherein:
and the parameter acquisition module is used for acquiring the global model parameters from the central server.
And the model construction module is used for constructing an initial multi-target detection model according to the global model parameters.
In this embodiment, the terminal constructs an initial multi-target detection model according to the global model parameters issued by the central server, thereby implementing model initialization.
In some optional implementations of this embodiment, the model training module 302 includes: the device comprises an image input submodule, an error determination submodule, a parameter adjustment submodule and an iterative training submodule, wherein:
and the image input submodule is used for inputting the target images in the local data set into the initial multi-target detection model to obtain a target prediction result.
And the error determining submodule is used for determining a prediction error according to the target prediction result and the image label in the local data set.
And the parameter adjusting submodule is used for adjusting parameters of the initial multi-target detection model based on the prediction error.
And the iterative training submodule is used for performing iterative training by taking the initial multi-target detection model after the parameters are adjusted as the initial multi-target detection model of the next training until the iteration times reach a preset value, so as to obtain the relay multi-target detection model.
In the embodiment, iterative training is performed on the initial multi-target detection model for the preset times according to the target images and the image labels in the local data set to obtain the relay multi-target detection model, and the relay multi-target detection model is used for federal learning, so that realization of federal learning is guaranteed.
In some optional implementations of this embodiment, the parameter calculating module 303 includes: an additional generation submodule and a parameter operation submodule, wherein:
an additional generation submodule for generating an additional random number; and adding the additional random numbers generated by each node in the alliance network to obtain a value of zero.
And the parameter operation submodule is used for performing linear operation on the generated additional random number and the model parameters of the relay multi-target detection model to obtain the parameters of the composite model.
In this embodiment, after the additional random number is generated, the additional random number and the model parameter of the relay multi-target detection model are subjected to linear operation to obtain a composite model parameter, so that data encryption of the model parameter is realized.
In some optional implementations of this embodiment, the parameter sending module 304 includes: the device comprises a key determining submodule, a parameter encrypting submodule and a parameter sending submodule, wherein:
and the key determining submodule is used for communicating with the central server to determine the encryption key.
And the parameter encryption submodule is used for encrypting the composite model parameter according to the encryption key to obtain an encryption model parameter.
And the parameter sending submodule is used for sending the encryption model parameters to the central server so as to instruct the central server to decrypt the encryption model parameters of each node, and calculating according to the decrypted composite model parameters to generate global model parameters.
In the embodiment, the composite model parameter is encrypted according to the encryption key to obtain the encryption model parameter, so that the data privacy in federal learning is further protected; after the encryption model parameters are sent to the central server, the central server decrypts the encryption model parameters and generates global model parameters according to the composite model parameters obtained after decryption, and the global model parameters are used for updating the relay multi-target detection model in each node, so that realization of federal learning is guaranteed.
In some optional implementations of the present embodiment, the processing device 300 for the intelligent decision-based target detection model further includes: image acquisition module, image input module and target show module, wherein:
and the image acquisition module is used for acquiring an image to be detected.
And the image input module is used for inputting the image to be detected into the multi-target detection model to obtain the target object in the image to be detected.
And the target display module is used for displaying the detected target object.
In the embodiment, the multi-target detection model for performing the target detection on the image to be detected is obtained through federal learning, and a rich data set is used for training in the federal learning, so that the accuracy of the target detection is improved.
In some optional implementations of the present embodiment, the processing device 300 for the intelligent decision-based target detection model further includes: the page display module, the calibration acquisition module, the calibration generation module and the calibration training module, wherein:
and the page display module is used for displaying the calibration information input page when the triggered calibration instruction is received.
And the calibration acquisition module is used for acquiring the image to be calibrated and the calibration indication information which are input in the calibration information input page.
And the calibration generating module is used for generating a calibration data set according to the image to be calibrated and the calibration indication information.
And the calibration training module is used for carrying out calibration training on the multi-target detection model through the calibration data set.
In the embodiment, when the calibration instruction is received, the calibration information input page is displayed, and the calibration data set is generated according to the image to be calibrated and the calibration indication information input in the calibration information input page, so that the multi-target detection model is calibrated and trained, and the detection accuracy of the multi-target detection model is improved.
In order to solve the technical problem, an embodiment of the present application further provides a computer device. Referring to fig. 4, fig. 4 is a block diagram of a basic structure of a computer device according to the present embodiment.
The computer device 4 comprises a memory 41, a processor 42, a network interface 43 communicatively connected to each other via a system bus. It is noted that only computer device 4 having components 41-43 is shown, but it is understood that not all of the shown components are required to be implemented, and that more or fewer components may be implemented instead. As will be understood by those skilled in the art, the computer device is a device capable of automatically performing numerical calculation and/or information processing according to a preset or stored instruction, and the hardware includes, but is not limited to, a microprocessor, an Application Specific Integrated Circuit (ASIC), a Programmable Gate Array (FPGA), a Digital Signal Processor (DSP), an embedded device, and the like.
The computer device can be a desktop computer, a notebook, a palm computer, a cloud server and other computing devices. The computer equipment can carry out man-machine interaction with a user through a keyboard, a mouse, a remote controller, a touch panel or voice control equipment and the like.
The memory 41 includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a Read Only Memory (ROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), a Programmable Read Only Memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as a hard disk or a memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are provided on the computer device 4. Of course, the memory 41 may also include both internal and external storage devices of the computer device 4. In this embodiment, the memory 41 is generally used for storing an operating system installed in the computer device 4 and various types of application software, such as computer readable instructions of an intelligent decision-based target detection model processing method. Further, the memory 41 may also be used to temporarily store various types of data that have been output or are to be output.
The processor 42 may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor, or other data Processing chip in some embodiments. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is configured to execute computer readable instructions or processing data stored in the memory 41, for example, computer readable instructions for executing the intelligent decision-based target detection model processing method.
The network interface 43 may comprise a wireless network interface or a wired network interface, and the network interface 43 is generally used for establishing communication connection between the computer device 4 and other electronic devices.
The computer device provided in this embodiment may execute the above target detection model processing method based on intelligent decision. The target detection model processing method based on intelligent decision here may be the target detection model processing method based on intelligent decision of the above embodiments.
In the embodiment, the initial multi-target detection model is trained according to the local data set to obtain a relay multi-target detection model; generating an additional random number, wherein the additional random number is used for calculating the model parameters of the relay multi-target detection model so as to encrypt the model parameters; the encrypted conforming model parameters are sent to a central server in the alliance network, the central server generates global model parameters according to the composite model parameters of all the nodes, and the global model parameters are issued to a terminal so as to update a relay multi-target detection model and carry out iterative training; the global model parameters are obtained on the basis of the local data sets of all the terminals, the characteristics of the local data sets are fused, the training of the multi-target detection model by the data sets in a larger scale is realized on the basis of protecting the local data sets, and the accuracy of the multi-target detection model detection obtained after the training is finished is improved.
The present application further provides another embodiment, which is to provide a computer-readable storage medium storing computer-readable instructions executable by at least one processor to cause the at least one processor to perform the steps of the intelligent decision-based object detection model processing method as described above.
In the embodiment, the initial multi-target detection model is trained according to the local data set to obtain a relay multi-target detection model; generating an additional random number, wherein the additional random number is used for calculating the model parameters of the relay multi-target detection model so as to encrypt the model parameters; the encrypted conforming model parameters are sent to a central server in the alliance network, the central server generates global model parameters according to the composite model parameters of all the nodes, and the global model parameters are issued to a terminal so as to update a relay multi-target detection model and carry out iterative training; the global model parameters are obtained on the basis of the local data sets of all the terminals, the characteristics of the local data sets are fused, the training of the multi-target detection model by the data sets in a larger scale is realized on the basis of protecting the local data sets, and the accuracy of the multi-target detection model detection obtained after the training is finished is improved.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present application may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present application.
It is to be understood that the above-described embodiments are merely illustrative of some, but not restrictive, of the broad invention, and that the appended drawings illustrate preferred embodiments of the invention and do not limit the scope of the invention. This application is capable of embodiments in many different forms and is provided for the purpose of enabling a thorough understanding of the disclosure of the application. Although the present application has been described in detail with reference to the foregoing embodiments, it will be apparent to one skilled in the art that the present application may be practiced without modification or with equivalents of some of the features described in the foregoing embodiments. All equivalent structures made by using the contents of the specification and the drawings of the present application are directly or indirectly applied to other related technical fields and are within the protection scope of the present application.
Claims (10)
1. A processing method of an object detection model based on intelligent decision is characterized in that the object detection model is a multi-object detection model, and the method comprises the following steps:
acquiring a local data set and an initial multi-target detection model;
training the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model;
generating an additional random number, and calculating a composite model parameter according to the additional random number and the model parameter of the relay multi-target detection model;
sending the composite model parameters to a central server to instruct the central server to generate global model parameters according to the composite model parameters of each node;
receiving the global model parameters from the central server to update the relay multi-target detection model;
and performing iterative training by taking the updated relay multi-target detection model as an initial multi-target detection model of next training until the model converges to obtain the multi-target detection model.
2. A method of processing an intelligent decision-making based object detection model as claimed in claim 1, wherein the steps of obtaining a local data set and an initial multi-object detection model are preceded by the method further comprising:
obtaining global model parameters from a central server;
and constructing an initial multi-target detection model according to the global model parameters.
3. The intelligent decision-making based target detection model processing method of claim 1, wherein the step of training the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model comprises:
inputting the target image in the local data set into the initial multi-target detection model to obtain a target prediction result;
determining a prediction error according to the target prediction result and the image label in the local data set;
performing parameter adjustment on the initial multi-target detection model based on the prediction error;
and performing iterative training by taking the initial multi-target detection model after parameter adjustment as an initial multi-target detection model for next training until the iteration times reach a preset value, thereby obtaining a relay multi-target detection model.
4. The method of claim 1, wherein the step of generating an additional random number and calculating composite model parameters based on the additional random number and the model parameters of the relay multi-objective detection model comprises:
generating an additional random number; wherein, the added value of the additional random numbers generated by each node in the alliance network is zero;
and performing linear operation on the generated additional random number and the model parameters of the relay multi-target detection model to obtain composite model parameters.
5. An intelligent decision-making based processing method for a target detection model according to claim 1, wherein the step of sending the composite model parameters to a central server to instruct the central server to generate global model parameters from the composite model parameters of each node comprises:
communicating with a central server to determine an encryption key;
encrypting the composite model parameter according to the encryption key to obtain an encryption model parameter;
and sending the encryption model parameters to the central server to instruct the central server to decrypt the encryption model parameters of each node, and calculating according to the decrypted composite model parameters to generate global model parameters.
6. The method of claim 1, wherein after the step of iteratively training the updated relay multi-target detection model as an initial multi-target detection model for a next round of training until the model converges to obtain a multi-target detection model, the method further comprises:
acquiring an image to be detected;
inputting the image to be detected into the multi-target detection model to obtain a target object in the image to be detected;
and displaying the detected target object.
7. The intelligent decision-making based target detection model processing method according to claim 6, wherein after the step of presenting the detected target object, the method further comprises:
when a triggered calibration instruction is received, displaying a calibration information input page;
acquiring an image to be calibrated and calibration indication information which are input in the calibration information input page;
generating a calibration data set according to the image to be calibrated and the calibration indication information;
and carrying out calibration training on the multi-target detection model through the calibration data set.
8. An apparatus for processing an object detection model based on intelligent decision making, wherein the object detection model is a multi-object detection model, the apparatus comprising:
the acquisition module is used for acquiring a local data set and an initial multi-target detection model;
the model training module is used for training the initial multi-target detection model according to the local data set to obtain a relay multi-target detection model;
the parameter calculation module is used for generating an additional random number and calculating a composite model parameter according to the additional random number and the model parameter of the relay multi-target detection model;
the parameter sending module is used for sending the composite model parameters to a central server so as to instruct the central server to generate global model parameters according to the composite model parameters of each node;
a model updating module for receiving the global model parameters from the central server to update the relay multi-target detection model;
and the iterative training module is used for performing iterative training by taking the updated relay multi-target detection model as an initial multi-target detection model for next training until the model converges to obtain the multi-target detection model.
9. A computer device comprising a memory having computer readable instructions stored therein and a processor which when executed implements the steps of a method of processing an intelligent decision-based object detection model as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium, having stored thereon computer-readable instructions which, when executed by a processor, implement the steps of a method of processing an intelligent decision-based object detection model as claimed in any one of claims 1 to 7.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010990029.4A CN112116008B (en) | 2020-09-18 | Processing method of target detection model based on intelligent decision and related equipment thereof | |
| PCT/CN2020/124439 WO2021174877A1 (en) | 2020-09-18 | 2020-10-28 | Processing method for smart decision-based target detection model, and related device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010990029.4A CN112116008B (en) | 2020-09-18 | Processing method of target detection model based on intelligent decision and related equipment thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112116008A true CN112116008A (en) | 2020-12-22 |
| CN112116008B CN112116008B (en) | 2024-07-05 |
Family
ID=
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114358316A (en) * | 2022-01-14 | 2022-04-15 | 中国人民解放军总医院 | Federal learning system and large-scale image training method and device thereof |
| CN117671385A (en) * | 2023-12-20 | 2024-03-08 | 北京斯年智驾科技有限公司 | Training method, system, device and storage medium for target recognition model |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5860556B1 (en) * | 2015-02-06 | 2016-02-16 | 日本電信電話株式会社 | Inconsistency detection method, inconsistency detection system, inconsistency detection apparatus, and program |
| CN110572253A (en) * | 2019-09-16 | 2019-12-13 | 济南大学 | Method and system for enhancing privacy of federated learning training data |
| CN110704860A (en) * | 2019-11-18 | 2020-01-17 | 深圳前海微众银行股份有限公司 | Longitudinal federal learning method, device and system for improving safety and storage medium |
| CN111462174A (en) * | 2020-03-06 | 2020-07-28 | 北京百度网讯科技有限公司 | Multi-target tracking method and device and electronic equipment |
| CN111611610A (en) * | 2020-04-12 | 2020-09-01 | 西安电子科技大学 | Federal learning information processing method, system, storage medium, program, and terminal |
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5860556B1 (en) * | 2015-02-06 | 2016-02-16 | 日本電信電話株式会社 | Inconsistency detection method, inconsistency detection system, inconsistency detection apparatus, and program |
| CN110572253A (en) * | 2019-09-16 | 2019-12-13 | 济南大学 | Method and system for enhancing privacy of federated learning training data |
| CN110704860A (en) * | 2019-11-18 | 2020-01-17 | 深圳前海微众银行股份有限公司 | Longitudinal federal learning method, device and system for improving safety and storage medium |
| CN111462174A (en) * | 2020-03-06 | 2020-07-28 | 北京百度网讯科技有限公司 | Multi-target tracking method and device and electronic equipment |
| CN111611610A (en) * | 2020-04-12 | 2020-09-01 | 西安电子科技大学 | Federal learning information processing method, system, storage medium, program, and terminal |
Non-Patent Citations (1)
| Title |
|---|
| 杜兰;刘彬;王燕;刘宏伟;代慧;: "基于卷积神经网络的SAR图像目标检测算法", 电子与信息学报, vol. 38, no. 12, pages 3018 - 3025 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114358316A (en) * | 2022-01-14 | 2022-04-15 | 中国人民解放军总医院 | Federal learning system and large-scale image training method and device thereof |
| CN117671385A (en) * | 2023-12-20 | 2024-03-08 | 北京斯年智驾科技有限公司 | Training method, system, device and storage medium for target recognition model |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2021174877A1 (en) | 2021-09-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021120676A1 (en) | Model training method for federated learning network, and related device | |
| CN110189192B (en) | Information recommendation model generation method and device | |
| US20210232974A1 (en) | Federated-learning based method of acquiring model parameters, system and readable storage medium | |
| WO2021174877A1 (en) | Processing method for smart decision-based target detection model, and related device | |
| Xu et al. | A lightweight RFID mutual authentication protocol based on physical unclonable function | |
| CN113159327A (en) | Model training method and device based on federal learning system, and electronic equipment | |
| CN112529101B (en) | Classification model training method and device, electronic equipment and storage medium | |
| CN113435583B (en) | Federal learning-based countermeasure generation network model training method and related equipment thereof | |
| CN112101437A (en) | Fine-grained classification model processing method based on image detection and related equipment thereof | |
| CN111831675A (en) | Storage model training method and device, computer equipment and storage medium | |
| CN114186256B (en) | Training method, device, equipment and storage medium of neural network model | |
| CN110837653A (en) | Label prediction method, device and computer readable storage medium | |
| CN115017107A (en) | Data retrieval method and device based on privacy protection, computer equipment and medium | |
| CN111368196A (en) | Model parameter updating method, device, equipment and readable storage medium | |
| CN113934983A (en) | Characteristic variable analysis method and device, computer equipment and storage medium | |
| CN111651749A (en) | Method and device for finding account based on password, computer equipment and storage medium | |
| CN114881247A (en) | Longitudinal federal feature derivation method, device and medium based on privacy computation | |
| CN113033717B (en) | Model generation method and device for model generation | |
| WO2022001233A1 (en) | Pre-labeling method based on hierarchical transfer learning and related device | |
| CN107665307A (en) | Application identification method and device, electronic equipment and storage medium | |
| CN111641517A (en) | Community division method and device for homogeneous network, computer equipment and storage medium | |
| CN112116008B (en) | Processing method of target detection model based on intelligent decision and related equipment thereof | |
| CN112085469B (en) | Data approval method, device, equipment and storage medium based on vector machine model | |
| CN115203391A (en) | Information retrieval method and device, computer equipment and storage medium | |
| CN109409111A (en) | It is a kind of to search for method generally towards encrypted image |
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 | ||
| GR01 | Patent grant |