Data processing method, apparatus and device

Abstract

Embodiments of the present description provide a data processing method, apparatus and device. The method comprises: on the basis of picture data corresponding to execution of a target service triggered by a target user, by means of a pre-trained block prediction model, determining a target number of blocks corresponding to the picture data, wherein the pre-trained block prediction model is obtained by training, on the basis of picture sample data and a first number of blocks of the picture sample data, a prediction model constructed by a preset machine learning algorithm, and the first number of blocks of the picture sample data is the number of blocks selected from among a plurality of second numbers of blocks on the basis of prediction accuracy scores corresponding to the second numbers of blocks of the picture sample data; performing block processing on the picture data on the basis of the target number of blocks to obtain a plurality of pieces of target sub-picture data; and on the basis of the plurality of pieces of target sub-picture data, by means of a pre-trained classification model, determining a classification category corresponding to the picture data.

Description

Data processing method, device and equipment Technical Field

[0001] This document relates to the field of data processing technology, and in particular to a data processing method, device and equipment. Background Art

[0002] With the rapid development of the internet industry, network operators are providing an increasing number of services to users. Classifying the image data used by users when using these services, and providing better services based on the corresponding classifications of the image data, has become a key concern for network operators. This includes quickly and accurately performing identity authentication based on the corresponding classifications of the image data to protect user privacy from being leaked.

[0003] The image data can be divided into multiple sub-image data, and then the sub-image data is input into the classification model to determine the classification category corresponding to the image data. The more the number of image block divisions, the better the classification effect of the classification model. Similarly, as the number of image block divisions increases, the resource consumption of the classification model also increases. Therefore, a solution is needed that can reduce the resource consumption of the classification model while ensuring the model performance of the classification model.

[0004] Summary of the Invention

[0005] The purpose of the embodiments of this specification is to provide a solution that can reduce the resource consumption of a classification model while ensuring the model performance of the classification model.

[0006] In order to realize the above technical solution, the embodiments of this specification are implemented as follows.

[0007] In a first aspect, an embodiment of the present specification provides a data processing method, comprising: obtaining image data corresponding to a target service triggered by a target user; determining, based on the image data, a target number of blocks corresponding to the image data through a pre-trained block prediction model, wherein the pre-trained block prediction model is obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, the first number of blocks of the image sample data being a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks selected from a plurality of second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data being determined by sub-image data obtained by block processing of the image sample data according to the second number of blocks, and a pre-trained classification model, wherein the classification model is an image classification model constructed based on a preset deep learning algorithm; block processing of the image data based on the target number of blocks to obtain multiple target sub-image data; and determining, based on the multiple target sub-image data, the classification category corresponding to the image data through the pre-trained classification model.

[0008] In a second aspect, an embodiment of the present specification provides a data processing device, the device comprising: a first acquisition module for acquiring image data corresponding to the target service triggered by the target user; a first determination module for determining, based on the image data, a target number of blocks corresponding to the image data through a pre-trained block prediction model, the pre-trained block prediction model being obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, the first number of blocks of the image sample data being a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks being selected from a plurality of the second number of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data being sub-image data obtained by block processing of the image sample data based on the second number of blocks, And a pre-trained classification model is determined, the classification model is a picture classification model constructed based on a preset deep learning algorithm; a first blocking module is used to block the picture data based on the target number of blocks to obtain multiple target sub-picture data; a category determination module is used to determine the classification category corresponding to the picture data based on the multiple target sub-picture data through the pre-trained classification model.

[0009] In a third aspect, an embodiment of the present specification provides a data processing device, comprising: a processor; and a memory arranged to store computer-executable instructions, wherein the executable instructions, when executed, cause the processor to: obtain image data corresponding to the target user triggering the execution of a target service; based on the image data, determine the target number of blocks corresponding to the image data through a pre-trained block prediction model, wherein the pre-trained block prediction model is obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, wherein the first number of blocks of the image sample data is based on the The prediction accuracy score corresponding to the second number of blocks of the image sample data, the number of blocks selected from multiple second number of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data, is determined based on the sub-image data obtained by block processing of the image sample data by the second number of blocks, and a pre-trained classification model, the classification model is an image classification model constructed based on a preset deep learning algorithm; the image data is block-processed based on the target number of blocks to obtain multiple target sub-image data; based on the multiple target sub-image data, the classification category corresponding to the image data is determined through the pre-trained classification model.

[0010] In a fourth aspect, an embodiment of the present specification provides a storage medium, which is used to store computer-executable instructions, which implement the following process when executed: obtaining image data corresponding to the target service triggered by the target user; based on the image data, determining the target number of blocks corresponding to the image data through a pre-trained block prediction model, the pre-trained block prediction model is obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, the first number of blocks of the image sample data is a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks selected from multiple second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data is determined by sub-image data obtained by block processing of the image sample data according to the second number of blocks, and a pre-trained classification model, the classification model is an image classification model constructed based on a preset deep learning algorithm; block processing is performed on the image data based on the target number of blocks to obtain multiple target sub-image data; based on the multiple target sub-image data, determining the classification category corresponding to the image data through the pre-trained classification model. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of this specification or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0012] FIG1 is a schematic diagram of a data processing system of the present specification;

[0013] FIG2A is a flow chart of an embodiment of a data processing method of this specification;

[0014] FIG2B is a schematic diagram of a processing process of a data processing method of this specification;

[0015] FIG3 is a schematic diagram of image data of this specification;

[0016] FIG4 is a schematic diagram of the processing process of another data processing method of this specification;

[0017] FIG5 is a schematic diagram of a process for determining the number of second blocks in this specification;

[0018] FIG6 is a schematic structural diagram of a data processing device according to an embodiment of the present specification;

[0019] FIG7 is a schematic structural diagram of a data processing device in this specification. DETAILED DESCRIPTION

[0020] The embodiments of this specification provide a data processing method, apparatus, and device.

[0021] To help those skilled in the art better understand the technical solutions in this specification, the following will provide a clear and complete description of the technical solutions in the embodiments of this specification, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments in this specification without creative work should fall within the scope of protection of this specification.

[0022] The technical solution of this specification can be applied to a data processing system. As shown in FIG1 , the data processing system may include terminal devices and servers, wherein the server may be an independent server or a server cluster composed of multiple servers, and the terminal device may be a device such as a personal computer or a mobile terminal device such as a mobile phone or a tablet computer.

[0023] The data processing system may include n terminal devices and m servers, where n and m are positive integers greater than or equal to 1. The server may be a background server of an application, and the terminal device may be a client device of the application. The application may be an application that can provide users with services such as resource transfer services, video viewing services, and instant messaging services.

[0024] The terminal device can collect image data corresponding to a user-triggered service and send the collected image data to the server. The server can then obtain the image data corresponding to the service triggered by the user and, based on the image data and using a pre-trained block prediction model, determine the target number of blocks corresponding to the image data. The server can then block the image data based on the target number of blocks to obtain multiple target sub-image data. Finally, based on the multiple target sub-image data, the server can determine the classification category corresponding to the image data using a pre-trained classification model.

[0025] In addition, the server can also store the image data collected by the terminal device so that when the model training cycle is reached, the block prediction model and / or classification model can be trained using the stored image data to obtain a trained model.

[0026] In addition, the data processing system may also be provided with a central server (such as server 1), which may receive historical image data stored by the terminal device and / or server to determine the image sample data based on the historical image data. The central server may then determine the prediction accuracy score corresponding to the second number of blocks of the image sample data using a pre-trained classification model and sub-image data obtained by block processing the image sample data according to the second number of blocks. Furthermore, based on the prediction accuracy score corresponding to the second number of blocks of the image sample data, a first number of blocks of the image sample data is selected from a plurality of second numbers of blocks. Finally, the central server may train a block prediction model constructed by a preset machine learning algorithm based on the image sample data and the first number of blocks of the image sample data to obtain a trained block prediction model. In this way, the central server may send the model parameters of the trained block prediction model to other servers in the data processing system. The other servers may then update their local block prediction models based on the received model parameters to obtain a trained block prediction model. Furthermore, the trained block prediction model may be used to determine the target number of blocks corresponding to the image data, thereby determining the classification category corresponding to the image data based on the target number of blocks and the pre-trained classification model. Avoid business interruptions caused by the need to train the block prediction model and meet users' business needs.

[0027] The central server can also train the classification model and send the model parameters of the trained classification model to other servers in the data processing system. Other servers can then update the local classification model based on the received model parameters to obtain the trained classification model.

[0028] In addition, since the model level of the classification model may be larger than the model level of the block prediction model, the classification model can be trained by the central server, and the model parameters of the trained classification model can be sent to other servers in the data processing system, and the block prediction model can be trained locally by other servers in the data processing system to improve resource utilization.

[0029] The data processing method in the following embodiments can be implemented based on the above data processing system structure.

[0030] Example 1

[0031] As shown in Figures 2A and 2B, an embodiment of this specification provides a data processing method, wherein the execution subject of the method can be a server, wherein the server can be an independent server or a server cluster composed of multiple servers. The method can specifically include the following steps S202 to S208.

[0032] In S202, image data corresponding to the target service triggered by the target user is obtained.

[0033] Among them, the target business can be any business that may involve data risk issues such as user privacy data leakage. For example, the target business can be a resource transfer business, an identity authentication business, an account registration business, a resource accounting business, etc. The image data can be data that can indicate the image type related to the target user triggering the execution of the target business. For example, the image data can include image type data used to indicate the user identity of the target user, image type data used to indicate the processing process or processing results of the target business, etc. Specifically, taking the target business as the resource transfer business as an example, the image data can include the target user's identity authentication image (that is, the image input by the target user that can be used for identity verification, or it can also be an image determined based on the video data input by the target user containing the target user's biometric information, etc.) data, the resource transfer page corresponding to the target user triggering the execution of the resource transfer business (that is, the page containing information such as the resource transfer quantity, resource transfer method, and resource transfer correspondence) data, etc. Taking the target business as the account registration business as an example, the image data can include the target user's identity authentication image data, the account registration page corresponding to the target user triggering the execution of the account registration business (that is, the page containing information such as the registered account, registered user name, registered password, verification method, etc.) data, etc.

[0034] In practice, with the rapid development of the Internet industry, the types and quantities of business services provided by network operators to users are increasing, and how to classify the image data of users when using business services, so as to better provide business services to users according to the classification categories corresponding to the image data (such as quickly and accurately performing identity authentication based on the classification categories corresponding to the image data to protect the user's privacy data from being leaked, etc.), has become the focus of network operators. The image data can be divided into multiple sub-image data, and the sub-image data is input into the classification model to determine the classification category corresponding to the image data. The more the number of image block divisions, the better the classification effect of the classification model. Similarly, as the number of image block divisions increases, the resource consumption of the classification model is also greater. Therefore, a solution is needed that can reduce the resource consumption of the classification model while ensuring the model performance of the classification model. To this end, the embodiment of this specification provides a technical solution that can solve the above-mentioned problem. For details, please refer to the following content.

[0035] Taking the target business as the resource transfer business as an example, the target user can trigger the start of the resource transfer business through the resource transfer application installed in the terminal device. That is, when the terminal device detects that the target user triggers the start of the resource transfer business through a resource transfer application, the terminal device can collect the target user's identity authentication picture data (such as the user login page when the target user logs in to the resource transfer application, the certificate picture data entered by the target user that can be used for identity authentication, etc.), as well as the resource transfer page corresponding to the target user triggering the execution of the resource transfer business, as shown in Figure 3.

[0036] The terminal device may determine the collected image data as image data corresponding to the target service triggered by the target user, and send the image data to the server, that is, the server may receive the image data corresponding to the target service triggered by the target user.

[0037] In addition, the above-mentioned method for obtaining image data is an optional and feasible determination method. In actual application scenarios, there may be a variety of different acquisition methods, which may vary according to the actual application scenarios. The embodiments of this specification do not make specific limitations on this.

[0038] In S204, based on the image data, the target number of blocks corresponding to the image data is determined by a pre-trained block prediction model.

[0039] Among them, the pre-trained block prediction model can be obtained by training a prediction model constructed by a preset machine learning algorithm based on the picture sample data and the first number of blocks of the picture sample data. The first number of blocks of the picture sample data can be a prediction accuracy score corresponding to the second number of blocks of the picture sample data. The number of blocks selected from multiple second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the picture sample data can be determined based on the sub-picture data obtained by block processing of the picture sample data according to the second number of blocks, and a pre-trained classification model. The classification model can be a picture classification model constructed based on a preset deep learning algorithm. The second number of blocks can be any number of blocks. For example, the picture data can be blocked based on a ratio of 2*2, that is, the second number of blocks can be 4, or the picture data can be blocked based on a ratio of 3*3, that is, the second number of blocks can be 9, etc.

[0040] In implementation, the server can obtain a pre-trained image classification model, wherein the pre-trained image classification model can be obtained by the server locally training the image classification model based on historical image data, or the pre-trained image classification model can also be obtained by the central server in the data processing system training the image classification model based on historical image data, that is, the server can train the image classification model locally, or the server can also receive model parameters of the trained image classification model sent by the central server.

[0041] Then, the server can select a predetermined number of image data from the historical image data as image sample data, and determine the prediction accuracy score corresponding to each second number of blocks of the image sample data based on the sub-image data obtained by block processing the image sample data according to the second number of blocks, and the pre-trained classification model.

[0042] The server may select a number of blocks from the plurality of second numbers of blocks as the first number of blocks for the image sample data based on the prediction accuracy scores corresponding to the second number of blocks for the image sample data. Furthermore, the server may train a block prediction model based on the image sample data and the first number of blocks for the image sample data to obtain a trained block prediction model.

[0043] Furthermore, to conserve server data processing resources and avoid service interruptions caused by model training, a central server in the data processing system can train the block prediction model based on the above process and send the model parameters of the trained block prediction model to other servers. Specifically, the server can update the model parameters of its local block prediction model based on the trained block prediction model parameters sent by the central server, thereby obtaining a trained block prediction model.

[0044] The server can input the image data into a pre-trained block prediction model to obtain the target number of blocks corresponding to the image data.

[0045] In S206 , the image data is divided into blocks based on the target number of blocks to obtain a plurality of target sub-image data.

[0046] In implementation, taking the resource transfer page as shown in Figure 3 as an example, the server can perform average block processing on the image data based on the target number of blocks to obtain multiple target sub-image data. For example, when the target number of blocks is 4, the server can divide the image data into 4 target sub-image data of the same size based on a ratio of 2*2.

[0047] The above-mentioned block processing method is an optional and feasible processing method. In actual application scenarios, there can be a variety of different processing methods. Different processing methods can be selected according to different actual application scenarios. The embodiments of this specification do not make specific limitations on this.

[0048] In S208 , based on the plurality of target sub-image data, a classification category corresponding to the image data is determined using a pre-trained classification model.

[0049] Among them, the classification category corresponding to the image data can be determined based on the detection requirements corresponding to the target business. For example, taking the target business as resource transfer business as an example, the detection requirements corresponding to the resource transfer business can be risk detection requirements. Correspondingly, the classification category corresponding to the image data can include high-risk category, medium-risk category, low-risk category, etc., or, taking the target business as account registration business as an example, the detection requirements corresponding to the account registration business can be information classification detection requirements. Correspondingly, the classification category corresponding to the image data can include ID category, facial information category, iris information category, fingerprint information category, etc.

[0050] In implementation, the server can input multiple target sub-image data into a pre-trained classification model to obtain the classification category corresponding to the image data. The server can determine whether to execute the target business based on the classification category corresponding to the image data, or execute the target business based on the classification category corresponding to the image data.

[0051] For example, taking the target business as resource transfer business, the server can determine whether to execute the target business based on the classification category corresponding to the image data (i.e. high-risk category, medium-risk category, low-risk category, etc.). That is, if it is determined that the image data is risky based on the classification category corresponding to the image data, then the server can suspend the execution of the target business.

[0052] Alternatively, taking the target business as account registration business as an example, the server can determine the corresponding business processing strategy based on the classification category corresponding to the image data, that is, the business processing strategies corresponding to classification categories such as ID card category, facial information category, iris information category, and fingerprint information category are different. The server can execute the target business based on the determined business processing logic and image data.

[0053] An embodiment of the present specification provides a data processing method, which obtains image data corresponding to a target service triggered by a target user, determines the target number of blocks corresponding to the image data based on the image data through a pre-trained block prediction model, the pre-trained block prediction model is based on image sample data and a first number of blocks of the image sample data, and is obtained by training a prediction model constructed by a preset machine learning algorithm, the first number of blocks of the image sample data is a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks selected from multiple second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data is sub-image data obtained by block processing of the image sample data by the second number of blocks, and is determined by a pre-trained classification model, the classification model is a picture classification model constructed based on a preset deep learning algorithm, the image data is block processed based on the target number of blocks to obtain multiple target sub-image data, based on multiple target sub-images Slice data, through a pre-trained classification model, the classification category corresponding to the image data is determined, and by introducing a pre-trained block prediction model, the number of blocks corresponding to the image data can be accurately predicted, and the first number of blocks corresponding to the image sample data used to train the block prediction model is based on the prediction accuracy score corresponding to the second number of blocks of the image sample data, and the number of blocks selected from the second number of blocks. In this way, the trained block prediction model can output the corresponding number of blocks for the image data, that is, when the task difficulty of the image data is low, the trained block prediction model can output a smaller number of block schemes, and when the task difficulty of the image data is high, the trained block prediction model can output a larger number of block schemes, thereby ensuring the overall performance of the classification model, and effectively reducing the average number of image blocks required by the classification model in the category prediction process, so as to reduce the computational overhead of the network and reduce the resource consumption of the classification model.

[0054] Example 2

[0055] As shown in Figure 4, an embodiment of this specification provides a data processing method, the execution subject of which can be a server, wherein the server can be an independent server or a server cluster composed of multiple servers. The method can specifically include the following steps S202 to S424.

[0056] In S202, image data corresponding to the target service triggered by the target user is obtained.

[0057] In S402, first image sample data and a first type corresponding to the first image sample data are obtained.

[0058] In S404 , the first image sample data is divided into blocks based on a preset number of blocks to obtain a plurality of first sub-image data.

[0059] In S406 , based on the plurality of first sub-picture data, a prediction type corresponding to the first picture sample data is determined by a classification model, and based on the prediction type and the first type, it is determined whether the classification model converges.

[0060] The classification model may be a neural network model for computer vision recognition tasks. For example, the classification model may be a VIT model based on the Vision Transformer algorithm, a SwinT model based on the Swin Transformer algorithm, or the like.

[0061] In implementation, taking the VIT model as an example, the input image is divided into several blocks and sent to the VIT model network as input. The more blocks the image is divided into, the better the classification model will be. However, since the time and computing resource consumption of the Vision Transformer network increases quadratically with the increase in the number of input blocks, more image blocks will bring better performance but also greater resource consumption.

[0062] In addition, to ensure the generalization of the classification model, image data that is divided into blocks based on different numbers of blocks can be input when training the classification model, that is, there can be multiple preset numbers of blocks. The processing method of the above S422 can refer to the following steps 1 to 2.

[0063] Step 1: Based on a plurality of first sub-picture data corresponding to each preset number of blocks, a classification model is used to determine a prediction type corresponding to the first picture sample data that is processed by the preset number of blocks.

[0064] Step 2: Determine whether the classification model converges based on the predicted type corresponding to the first image sample data that is block-processed by a preset number of blocks and the first type corresponding to the first image sample data.

[0065] In S408 , when it is determined that the classification model has not converged, the classification model is continuously trained based on the first image sample data and the first type corresponding to the first image sample data until the classification model converges, thereby obtaining a trained classification model.

[0066] In S410 , image sample data and a second type corresponding to the image sample data are obtained.

[0067] The picture sample data may be picture data corresponding to the target service, and the data volume of the picture sample data may be smaller than the data volume of the first picture sample data.

[0068] During implementation, in order to improve the accuracy of the classification model and the block prediction model, as well as the efficiency of model training, the server can select image data corresponding to the target business from the historical image data as image sample data for training the block prediction model. At the same time, the server can also determine the historical image data as the first image sample data for training the classification prediction model, that is, the data volume of the image sample data can be smaller than the data volume of the first image sample data.

[0069] In S412, based on the multiple second block numbers, the image sample data is divided into blocks to obtain multiple sub-image data corresponding to each second block number.

[0070] In S414, based on the plurality of sub-picture data corresponding to each second number of blocks, a prediction type corresponding to each second number of blocks is determined by a pre-trained classification model.

[0071] In S416 , based on the prediction type corresponding to the second number of blocks and the second type corresponding to the picture sample data, a prediction accuracy score corresponding to each second number of blocks of the picture sample data is determined.

[0072] In S418 , based on the prediction accuracy score corresponding to the second number of blocks of the picture sample data and a preset accuracy threshold, a first number of blocks of the picture sample data is selected from a plurality of second numbers of blocks of the picture sample data.

[0073] In implementation, as shown in Figure 5, assuming that the second number of blocks includes the second number of blocks 1 (i.e., 2*2), the second number of blocks 2 (i.e., 3*3), and the second number of blocks 2 (i.e., 4*4), the picture sample data can be separately block-processed based on these three second number of blocks to obtain multiple sub-picture data corresponding to each second number of blocks.

[0074] Then, the server can input multiple sub-image data corresponding to each second number of blocks into a pre-trained classification model to obtain the prediction type corresponding to each second number of blocks, and determine the prediction accuracy score corresponding to each second number of blocks of the image sample data based on the second type corresponding to the image sample data.

[0075] In this way, the server can select the first number of blocks of the image sample data from multiple second numbers of blocks of the image sample data according to the preset accuracy threshold and the determined prediction accuracy score. For example, assuming that the preset accuracy threshold is 0.8, then the second number of blocks 3 can be determined as the first number of blocks, that is, the first number of blocks of the image sample data can be 4*4.

[0076] Among them, the preset accuracy threshold can be determined based on the business processing requirements of the target business. For example, when the business processing accuracy requirements of the target business are high and the business processing efficiency requirements are low, the preset accuracy threshold can be threshold 1. When the business processing accuracy requirements of the target business are low and the business processing efficiency requirements are high, the preset accuracy threshold can be threshold 2, and threshold 1 is greater than threshold 2.

[0077] In S420, based on the picture sample data, the number of second blocks corresponding to the picture sample data is determined by a block prediction model.

[0078] Among them, in order to reduce resource consumption, a lightweight network structure can be selected to build a block prediction model, that is, the model level of the block prediction model can be smaller than the preset level threshold. For example, the block prediction model can be built based on efficient-net.

[0079] In S422, based on the first number of blocks and the second number of blocks, it is determined whether the block prediction model has converged. If it is determined that the block prediction model has not converged, the block prediction model is continued to be trained based on the image sample data and the first number of blocks of the image sample data until the block prediction model converges to obtain a trained block prediction model.

[0080] In S204, based on the image data, the target number of blocks corresponding to the image data is determined by a pre-trained block prediction model.

[0081] In S206 , the image data is divided into blocks based on the target number of blocks to obtain a plurality of target sub-image data.

[0082] In S208 , based on the plurality of target sub-image data, a classification category corresponding to the image data is determined using a pre-trained classification model.

[0083] In S424 , based on the classification category corresponding to the image data, it is determined whether there is a risk in triggering the target user to execute the target service.

[0084] During implementation, the server can determine whether there is a risk when the target user triggers the execution of the target business based on the level of risk detection requirements corresponding to the target business and the classification category corresponding to the image data. For example, taking the target business as a resource transfer business, the level of risk detection requirements corresponding to the resource transfer business is higher. If the classification category corresponding to the image data is high risk or medium risk, then it can be determined that there is a risk when the target user triggers the execution of the target business.

[0085] When the server determines that there is a risk in triggering the target user to execute the target service, the server may output a preset alarm message to the target user and stop executing the target service.

[0086] An embodiment of the present specification provides a data processing method, which obtains image data corresponding to a target service triggered by a target user, determines the target number of blocks corresponding to the image data based on the image data through a pre-trained block prediction model, the pre-trained block prediction model is based on image sample data and a first number of blocks of the image sample data, and is obtained by training a prediction model constructed by a preset machine learning algorithm, the first number of blocks of the image sample data is a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks selected from multiple second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data is sub-image data obtained by block processing of the image sample data by the second number of blocks, and is determined by a pre-trained classification model, the classification model is a picture classification model constructed based on a preset deep learning algorithm, the image data is block processed based on the target number of blocks to obtain multiple target sub-image data, based on multiple target sub-images Slice data, through a pre-trained classification model, the classification category corresponding to the image data is determined, and by introducing a pre-trained block prediction model, the number of blocks corresponding to the image data can be accurately predicted, and the first number of blocks corresponding to the image sample data used to train the block prediction model is based on the prediction accuracy score corresponding to the second number of blocks of the image sample data, and the number of blocks selected from the second number of blocks. In this way, the trained block prediction model can output the corresponding number of blocks for the image data, that is, when the task difficulty of the image data is low, the trained block prediction model can output a smaller number of block schemes, and when the task difficulty of the image data is high, the trained block prediction model can output a larger number of block schemes, thereby ensuring the overall performance of the classification model, and effectively reducing the average number of image blocks required by the classification model in the category prediction process, so as to reduce the computational overhead of the network and reduce the resource consumption of the classification model.

[0087] Example 3

[0088] The above is the data processing method provided in the embodiment of this specification. Based on the same idea, the embodiment of this specification also provides a data processing device, as shown in FIG6 .

[0089] The data processing device includes: a first acquisition module 601 , a first determination module 602 , a first blocking module 603 and a category determination module 604 .

[0090] The first acquisition module 601 is used to acquire image data corresponding to the target service triggered by the target user;

[0091] A first determination module 602 is configured to determine, based on the image data, a target number of blocks corresponding to the image data using a pre-trained block prediction model, wherein the pre-trained block prediction model is obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, wherein the first number of blocks of the image sample data is a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks selected from a plurality of second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data is determined based on sub-image data obtained by block processing the image sample data according to the second number of blocks, and a pre-trained classification model, wherein the classification model is an image classification model constructed based on a preset deep learning algorithm;

[0092] A first block division module 603 is configured to perform block processing on the image data based on the target number of blocks to obtain a plurality of target sub-image data;

[0093] The category determination module 604 is configured to determine the classification category corresponding to the image data based on the plurality of target sub-image data using the pre-trained classification model.

[0094] In an embodiment of the present specification, the device also includes: a second acquisition module, used to obtain first image sample data, and a first type corresponding to the first image sample data; a second blocking module, used to block the first image sample data based on a preset number of blocks to obtain multiple first sub-image data; a first judgment module, used to determine the prediction type corresponding to the first image sample data through the classification model based on the multiple first sub-image data, and determine whether the classification model converges based on the prediction type and the first type; a first training module, used to continue training the classification model based on the first image sample data and the first type corresponding to the first image sample data until the classification model converges to obtain a trained classification model when it is determined that the classification model has not converged.

[0095] In an embodiment of the present specification, there are multiple preset numbers of blocks, and the first judgment module is used to: determine, based on multiple first sub-image data corresponding to each of the preset numbers of blocks, through the classification model, the prediction type corresponding to the first image sample data that is block-processed with the preset number of blocks; based on the prediction type corresponding to the first image sample data that is block-processed with the preset number of blocks, and the first type corresponding to the first image sample data, determine whether the classification model converges.

[0096] In an embodiment of the present specification, the device also includes: a third acquisition module, used to obtain the image sample data, and the second type corresponding to the image sample data; a third blocking module, used to block the image sample data based on multiple second blocking numbers, and obtain multiple sub-image data corresponding to each second blocking number; a type determination module, used to determine the prediction type corresponding to each second blocking number through the pre-trained classification model based on the multiple sub-image data corresponding to each second blocking number; a score determination module, used to determine the prediction accuracy score corresponding to each second blocking number of the image sample data based on the prediction type corresponding to the second blocking number and the second type corresponding to the image sample data.

[0097] In an embodiment of the present specification, the device also includes: a number selection module, which is used to select the first number of blocks of the picture sample data from multiple second numbers of blocks of the picture sample data based on the prediction accuracy score corresponding to the second number of blocks of the picture sample data and a preset accuracy threshold; a block determination module, which is used to determine the second number of blocks corresponding to the picture sample data through the block prediction model based on the picture sample data; a second training module, which is used to determine whether the block prediction model converges based on the first number of blocks and the second number of blocks, and if it is determined that the block prediction model has not converged, continue to train the block prediction model based on the picture sample data and the first number of blocks of the picture sample data until the block prediction model converges to obtain a trained block prediction model.

[0098] In the embodiment of this specification, the model level of the block prediction model is less than a preset level threshold, and the classification model is a neural network model for computer vision recognition tasks.

[0099] In the embodiment of this specification, the picture sample data is picture data corresponding to the target service, and the data volume of the picture sample data is smaller than the data volume of the first picture sample data.

[0100] In the embodiment of the present specification, the device further includes: a risk judgment module, which is used to determine whether there is a risk when the target user triggers the execution of the target business based on the classification category corresponding to the image data.

[0101] An embodiment of the present specification provides a data processing device, which obtains image data corresponding to the target service triggered by the target user, and determines the target number of blocks corresponding to the image data based on the image data through a pre-trained block prediction model, the pre-trained block prediction model is based on the image sample data and the first number of blocks of the image sample data, and is obtained by training a prediction model constructed by a preset machine learning algorithm, the first number of blocks of the image sample data is a prediction accuracy score corresponding to the second number of blocks of the image sample data, the number of blocks selected from multiple second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data is based on the sub-image data obtained by block processing of the image sample data by the second number of blocks, and is determined by a pre-trained classification model, the classification model is a picture classification model constructed based on a preset deep learning algorithm, the image data is block processed based on the target number of blocks to obtain multiple target sub-image data, based on multiple target sub-images Slice data, through a pre-trained classification model, the classification category corresponding to the image data is determined, and by introducing a pre-trained block prediction model, the number of blocks corresponding to the image data can be accurately predicted, and the first number of blocks corresponding to the image sample data used to train the block prediction model is based on the prediction accuracy score corresponding to the second number of blocks of the image sample data, and the number of blocks selected from the second number of blocks. In this way, the trained block prediction model can output the corresponding number of blocks for the image data, that is, when the task difficulty of the image data is low, the trained block prediction model can output a smaller number of block schemes, and when the task difficulty of the image data is high, the trained block prediction model can output a larger number of block schemes, thereby ensuring the overall performance of the classification model, and effectively reducing the average number of image blocks required by the classification model in the category prediction process, so as to reduce the computational overhead of the network and reduce the resource consumption of the classification model.

[0102] Example 4

[0103] Based on the same idea, the embodiment of this specification also provides a data processing device, as shown in FIG7 .

[0104] The data processing device may vary significantly due to different configurations or performance, and may include one or more processors 701 and memory 702. The memory 702 may store one or more applications or data. The memory 702 may be a temporary storage or a persistent storage. The application stored in the memory 702 may include one or more modules (not shown), each of which may include a series of computer-executable instructions for the data processing device. Furthermore, the processor 701 may be configured to communicate with the memory 702 to execute the series of computer-executable instructions in the memory 702 on the data processing device. The data processing device may also include one or more power supplies 703, one or more wired or wireless network interfaces 704, one or more input / output interfaces 705, and one or more keyboards 706.

[0105] Specifically, in this embodiment, the data processing device includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer-executable instructions for the data processing device, and the one or more programs are configured to be executed by one or more processors, including computer-executable instructions for performing the following:

[0106] Obtain the image data corresponding to the target user triggering the execution of the target business;

[0107] Based on the image data, determining a target number of blocks corresponding to the image data through a pre-trained block prediction model, the pre-trained block prediction model being obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, the first number of blocks of the image sample data being a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks being selected from a plurality of second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data being determined based on sub-image data obtained by block processing the image sample data according to the second number of blocks, and a pre-trained classification model, the classification model being an image classification model constructed based on a preset deep learning algorithm;

[0108] Performing block processing on the image data based on the target number of blocks to obtain a plurality of target sub-image data;

[0109] Based on the multiple target sub-image data, the classification category corresponding to the image data is determined through the pre-trained classification model.

[0110] The various embodiments in this specification are described in a progressive manner. Similar portions between the various embodiments can be referenced to each other. Each embodiment focuses on the differences between the other embodiments. In particular, the data processing device embodiment is generally similar to the method embodiment, so its description is relatively simple. For relevant portions, refer to the description of the method embodiment.

[0111] An embodiment of the present specification provides a data processing device, which obtains image data corresponding to the target service triggered by the target user, and determines the target number of blocks corresponding to the image data based on the image data through a pre-trained block prediction model, the pre-trained block prediction model is based on the image sample data and the first number of blocks of the image sample data, and is obtained by training a prediction model constructed by a preset machine learning algorithm, the first number of blocks of the image sample data is a prediction accuracy score corresponding to the second number of blocks of the image sample data, the number of blocks selected from multiple second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data is based on the sub-image data obtained by block processing of the image sample data by the second number of blocks, and is determined by a pre-trained classification model, the classification model is a picture classification model constructed based on a preset deep learning algorithm, the image data is block processed based on the target number of blocks to obtain multiple target sub-image data, based on multiple target sub-images Slice data, through a pre-trained classification model, the classification category corresponding to the image data is determined, and by introducing a pre-trained block prediction model, the number of blocks corresponding to the image data can be accurately predicted, and the first number of blocks corresponding to the image sample data used to train the block prediction model is based on the prediction accuracy score corresponding to the second number of blocks of the image sample data, and the number of blocks selected from the second number of blocks. In this way, the trained block prediction model can output the corresponding number of blocks for the image data, that is, when the task difficulty of the image data is low, the trained block prediction model can output a smaller number of block schemes, and when the task difficulty of the image data is high, the trained block prediction model can output a larger number of block schemes, thereby ensuring the overall performance of the classification model, and effectively reducing the average number of image blocks required by the classification model in the category prediction process, so as to reduce the computational overhead of the network and reduce the resource consumption of the classification model.

[0112] Example 5

[0113] The embodiments of this specification also provide a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the various processes of the above-mentioned data processing method embodiments are implemented and can achieve the same technical effects. To avoid repetition, the details are not described here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0114] An embodiment of the present specification provides a computer-readable storage medium, which obtains image data corresponding to a target service triggered by a target user, determines the target number of blocks corresponding to the image data based on the image data through a pre-trained block prediction model, the pre-trained block prediction model is based on image sample data and a first number of blocks of the image sample data, and is obtained by training a prediction model constructed by a preset machine learning algorithm, the first number of blocks of the image sample data is a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks selected from multiple second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data is sub-image data obtained by block processing of the image sample data by the second number of blocks, and is determined by a pre-trained classification model, the classification model is an image classification model constructed based on a preset deep learning algorithm, the image data is block processed based on the target number of blocks to obtain multiple target sub-image data, based on multiple target Sub-image data, through a pre-trained classification model, determines the classification category corresponding to the image data, and by introducing a pre-trained block prediction model, can accurately predict the number of blocks corresponding to the image data, and the first number of blocks corresponding to the image sample data used to train the block prediction model is based on the prediction accuracy score corresponding to the second number of blocks of the image sample data, and the number of blocks selected from the second number of blocks. In this way, the trained block prediction model can output the corresponding number of blocks for the image data, that is, when the task difficulty of the image data is low, the trained block prediction model can output a smaller number of block schemes, and when the task difficulty of the image data is high, the trained block prediction model can output a larger number of block schemes, thereby ensuring the overall performance of the classification model, and effectively reducing the average number of image blocks required by the classification model in the category prediction process, so as to reduce the computational overhead of the network and reduce the resource consumption of the classification model.

[0115] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0116] In the 1990s, technological improvements could be clearly distinguished as either hardware improvements (for example, improvements to circuit structures such as diodes, transistors, and switches) or software improvements (improvements to process flows). However, with the advancement of technology, many process flow improvements today can now be considered direct improvements to hardware circuit structures. Designers almost always create the corresponding hardware circuit structure by programming the improved process flow into the hardware circuit. Therefore, it cannot be said that a process flow improvement cannot be implemented using hardware modules. For example, a programmable logic device (PLD), such as a field programmable gate array (FPGA), is an integrated circuit whose logical function is determined by user programming. Designers can "integrate" a digital system on a PLD by programming it themselves, without having to hire a chip manufacturer to design and manufacture a dedicated integrated circuit chip. Moreover, nowadays, instead of manually fabricating integrated circuit chips, this programming is mostly done using "logic compiler" software. This is similar to the software compiler used when developing programs. Before compilation, the original code must also be written in a specific programming language, called a hardware description language (HDL). There is not just one HDL, but many, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art will also understand that by simply programming the method flow in one of these hardware description languages ​​and then programming it into an integrated circuit, a hardware circuit that implements the logic method flow can be easily obtained.

[0117] The controller can be implemented in any suitable manner. For example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also know that in addition to implementing the controller in a purely computer-readable program code format, the controller can be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered a hardware component, and the devices included therein for implementing various functions can also be considered as structures within the hardware component. Or even, the devices for implementing various functions can be considered as both software modules that implement the method and structures within the hardware component.

[0118] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0119] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing one or more embodiments of this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0120] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, one or more embodiments of this specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, one or more embodiments of this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0121] The embodiments of this specification are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0122] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0123] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0124] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0125] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0126] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0127] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0128] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Thus, one or more embodiments of this specification may take the form of a fully hardware embodiment, a fully software embodiment, or an embodiment combining software and hardware. Furthermore, one or more embodiments of this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0129] One or more embodiments of this specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. One or more embodiments of this specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communications network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0130] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0131] The foregoing is merely an example of the present invention and is not intended to limit the present invention. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be included within the scope of the claims of the present invention.

Claims

1. A data processing method, comprising: Obtain the image data corresponding to the target user triggering the execution of the target business; Based on the image data, determining a target number of blocks corresponding to the image data through a pre-trained block prediction model, the pre-trained block prediction model being obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, the first number of blocks of the image sample data being a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks being selected from a plurality of second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data being determined based on sub-image data obtained by block processing the image sample data according to the second number of blocks, and a pre-trained classification model, the classification model being an image classification model constructed based on a preset deep learning algorithm; Performing block processing on the image data based on the target number of blocks to obtain a plurality of target sub-image data; Based on the multiple target sub-image data, the classification category corresponding to the image data is determined through the pre-trained classification model.

2. The method according to claim 1, before determining the target number of blocks corresponding to the image data using a pre-trained block prediction model based on the image data, further comprising: Acquire first image sample data and a first type corresponding to the first image sample data; Based on a preset number of blocks, the first image sample data is processed into blocks to obtain a plurality of first sub-image data; Determining, based on the plurality of first sub-picture data, a prediction type corresponding to the first picture sample data using the classification model, and determining whether the classification model has converged based on the prediction type and the first type; When it is determined that the classification model has not converged, the classification model is continuously trained based on the first image sample data and the first type corresponding to the first image sample data until the classification model converges to obtain a trained classification model.

3. The method according to claim 2, wherein the number of preset blocks is multiple, and determining the prediction type corresponding to the first image sample data using the classification model based on the multiple first sub-image data, and determining whether the classification model has converged based on the prediction type and the first type, comprising: Based on a plurality of first sub-picture data corresponding to each of the preset number of blocks, determining, by the classification model, a prediction type corresponding to the first picture sample data subjected to block processing by the preset number of blocks; Based on the prediction type corresponding to the first picture sample data that is block-processed by the preset number of blocks and the first type corresponding to the first picture sample data, it is determined whether the classification model converges.

4. The method according to claim 3, further comprising: Acquire the image sample data and the second type corresponding to the image sample data; Based on the plurality of second block numbers, the image sample data is respectively subjected to block processing to obtain a plurality of sub-image data corresponding to each second block number; Based on the plurality of sub-picture data corresponding to each second number of blocks, determining the prediction type corresponding to each second number of blocks by using the pre-trained classification model; Based on the prediction type corresponding to the second number of blocks and the second type corresponding to the picture sample data, a prediction accuracy score corresponding to each second number of blocks of the picture sample data is determined.

5. The method according to claim 4, further comprising: Selecting the first number of blocks of the picture sample data from a plurality of second numbers of blocks of the picture sample data based on the prediction accuracy score corresponding to the second number of blocks of the picture sample data and a preset accuracy threshold; Based on the picture sample data, determining the number of second blocks corresponding to the picture sample data through the block prediction model; Based on the first number of blocks and the second number of blocks, determine whether the block prediction model converges. If it is determined that the block prediction model has not converged, continue to train the block prediction model based on the image sample data and the first number of blocks of the image sample data until the block prediction model converges, thereby obtaining a trained block prediction model.

6. According to the method of claim 5, the model level of the block prediction model is less than a preset level threshold, and the classification model is a neural network model for computer vision recognition tasks. 7 . The method according to claim 6 , wherein the picture sample data is picture data corresponding to the target service, and the data volume of the picture sample data is smaller than the data volume of the first picture sample data.

8. The method according to claim 7, further comprising: Based on the classification category corresponding to the image data, it is determined whether there is a risk in triggering the target user to execute the target service.

9. A data processing device comprising: The first acquisition module is used to acquire image data corresponding to the target service triggered by the target user; a first determination module, configured to determine, based on the image data, a target number of blocks corresponding to the image data using a pre-trained block prediction model, the pre-trained block prediction model being obtained by training a prediction model constructed by a preset machine learning algorithm based on image sample data and a first number of blocks of the image sample data, the first number of blocks of the image sample data being a prediction accuracy score corresponding to a second number of blocks of the image sample data, the number of blocks being selected from a plurality of second numbers of blocks, the prediction accuracy score corresponding to the second number of blocks of the image sample data being determined based on sub-image data obtained by block processing the image sample data according to the second number of blocks, and a pre-trained classification model, the classification model being an image classification model constructed based on a preset deep learning algorithm; A first block division module is used to perform block processing on the image data based on the target number of blocks to obtain a plurality of target sub-image data; The category determination module is used to determine the classification category corresponding to the image data based on the multiple target sub-image data through the pre-trained classification model.

10. A data processing device, comprising: processor; as well as a memory arranged to store computer-executable instructions which, when executed, cause the processor to: Obtain the image data corresponding to the target user triggering the execution of the target business; Based on the image data, a target number of blocks corresponding to the image data is determined by a pre-trained block prediction model, wherein the pre-trained block prediction model is obtained by training a prediction model constructed by a preset machine learning algorithm based on the image sample data and the first number of blocks of the image sample data, wherein the first number of blocks of the image sample data is a prediction accuracy score corresponding to the second number of blocks of the image sample data, and The number of blocks selected from the plurality of second numbers of blocks, and the prediction accuracy score corresponding to the second number of blocks of the image sample data, are determined based on sub-image data obtained by block processing the image sample data according to the second number of blocks, and a pre-trained classification model, where the classification model is an image classification model constructed based on a preset deep learning algorithm; Performing block processing on the image data based on the target number of blocks to obtain a plurality of target sub-image data; Based on the multiple target sub-image data, the classification category corresponding to the image data is determined through the pre-trained classification model.

Images