TW202145040A - Security protection method for model parameters, security protection device, computer device - Google Patents

Abstract

The invention provides a security protection method for model parameters, a security protection device and a computer device. The method is applied to the computer device, and the computer device is connected with at least one security protection device, the method includes: training a data model and obtaining parameter information in the data model; encrypting the parameter information and generating a configuration file including encrypted parameter information; sending the configuration file to the security protection device. The method is also applied to the security protection device, the method includes: receiving the configuration file sent by the computer device, finding the encrypted parameter information in the configuration file, acquiring the decryption key, decrypting the parameter information, and configuring the parameter information in a data model.

Description

Model parameter safety protection method, safety protection device and computer device

The invention relates to the technical field of big data analysis, in particular to a security protection method for model parameters, a security protection device and a computer device.

In the era of big data, people will use different data models to analyze various data in life and production, and use the results of data analysis to provide convenience for production and life. Therefore, the construction of the data model is a key part of data analysis, and the parameter setting of the data model directly affects the analysis efficiency and accuracy of the data model. The existing data model does not protect the parameters of the data model during the application process, and there are security problems such as leakage of data model parameters.

In view of the above content, it is necessary to propose an image identification method, an image identification device, a computer device and a storage medium, so that the image identification can be carried out in a more intelligent and efficient manner.

A first aspect of the present application provides a security protection method for model parameters, the method is applied to a computer device, and the computer device is communicatively connected to at least one security protection device, and the method includes: training a data model, and obtaining parameter information in the data model; encrypting the parameter information, and generating a configuration file including the encrypted parameter information; Send the configuration file to the security device.

Preferably, the parameter information includes: one or more of neuron weights and neuron offset values in the artificial neural network road algorithm.

Preferably, the step of encrypting the parameter information and generating a configuration file including the encrypted parameter information includes: performing binarization processing on the parameter information; An encryption algorithm in the compiler is called to encrypt the binarized parameter information, wherein the encryption algorithm includes a symmetric encryption algorithm and an asymmetric encryption algorithm.

Preferably, the method includes: Receive the request information sent by the security device to send the configuration file; According to the account information of the security protection device, determine whether the security protection device has the permission to acquire the configuration file; If the security protection device has the permission, the configuration file is sent to the security protection device.

A second aspect of the present application provides a computer device, the computer device includes a processor and a memory, and the processor is configured to implement the aforementioned model parameter security protection method when executing a computer program stored in the memory.

A third aspect of the present application provides a method for safety protection of model parameters, the method is applied in a safety protection device, the safety protection device communicates with a computer device, and the method includes: receiving a configuration file sent by the computer device; Find encrypted parameter information in the configuration file; Obtain the decryption key, and decrypt the parameter information; The parameter information is configured in the data model.

Preferably, the step of obtaining the decryption key and decrypting the parameter information includes: Receive the decryption key sent by the computer device, and use the decryption key to decrypt the parameter information.

Preferably, the step of obtaining the decryption key and decrypting the parameter information further comprises: Search for a preset decryption program to decrypt the parameter information.

Preferably, the step of configuring the parameter information in the data model includes: Obtain the location information of the parameters in the parameter and parameter location comparison table of the data model; Determine the parameter name corresponding to the location information from the data model, and search for the corresponding decryption parameter in the decrypted parameter information according to the parameter name; The searched decryption parameters are loaded into the data model at the location corresponding to the location information.

A fourth aspect of the present application provides a safety protection device, the device comprising: Memory: used to store computer programs; Processor: implements the aforementioned security protection method for model parameters when executing the computer program in the memory.

The security protection method, security protection device and computer device for model parameters of the present invention, the computer device trains a data model, extracts parameter information of the data model, encrypts the parameter information to generate a configuration file and sends it to the security protection device, so The security protection device receives the configuration file, decrypts the parameter information in the configuration file, and loads it into the data model. The security of the parameters of the model is guaranteed by the security protection method of the model parameters.

In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments may be combined with each other in the case of no conflict.

In the following description, many specific details are set forth in order to facilitate a full understanding of the present invention, and the described embodiments are only some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Unless otherwise defined, 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 invention belongs. The terms used herein in the description of the present invention are for the purpose of describing specific embodiments only, and are not intended to limit the present invention.

Example 1

Referring to FIG. 1 , it is a schematic diagram of an application environment architecture of the security protection method for model parameters provided by Embodiment 1 of the present invention.

The security protection method for model parameters in the present invention is applied in a computer device 1, and the computer device 1 and at least one security protection device 2 establish a communication connection through a network. The network may be a wired network or a wireless network, such as radio, Wireless Fidelity (WIFI), cellular, satellite, broadcast, and the like.

The computer device 1 is used for training a data model, obtaining parameter information in the data model; encrypting the parameter information, generating a configuration file including the encrypted parameter information; and sending the configuration file to the security protection device 2 . In yet another embodiment of the present invention, the computer device 1 is further configured to store sample information of the data model used for training. The sample information is used to train parameters of the data model.

The security protection device 2 is used for receiving the configuration file, searching for encrypted parameter information in the configuration file, obtaining a decryption key, and decrypting the parameter information; and configuring the parameter information in a data model.

The computer device 1 may be an electronic device installed with security protection software for model parameters, such as a personal computer, a server, etc., wherein the server may be a single server, a server cluster, or a cloud server.

The safety protection device 2 can be an electronic device equipped with a safety protection software of model parameters, including but not limited to a single chip microcomputer, a field programmable logic gate array, a smart phone, a smart wearable device, a tablet computer, a laptop convenient computer, desktop computer, etc.

Embodiment 2

Please refer to FIG. 2 , which is a flowchart of a method for security protection of model parameters applied to a computer device according to Embodiment 2 of the present invention. According to different requirements, the order of the steps in the flowchart can be changed, and some steps can be omitted.

Step S11 , training a data model, and acquiring parameter information in the data model.

In an embodiment of the present invention, the data model is a data model based on an artificial neural network. The parameter information includes one or more of the weights of the neurons of the artificial neural network and the offset values of the neurons.

In one embodiment, the data model is applied in the field of image detection to obtain feature information in the image to be detected, and the training process of the data model includes: Obtain multiple images with known feature information; dividing the image into a training set and a validation set; establishing an artificial neural network-based data model, and using the training set to train the artificial neural network-based data model; Use the verification set to verify the data model based on the artificial neural network after training, and statistically obtain a model prediction accuracy rate according to the verification result; Judging whether the prediction accuracy of the model is less than a preset threshold; If the prediction accuracy of the model is not less than the preset threshold, the trained data model based on the artificial neural network is used as the data model for acquiring feature information in the image to be detected.

If the prediction accuracy of the model is less than the preset threshold, adjust the neuron weights and neuron offsets of the artificial neural network-based data model, and use the training set to recalculate the artificial neural network-based data model. The data model of the road is trained; Use the verification set to verify the retrained artificial neural network-based data model, re-statistically obtain a model prediction accuracy rate according to each verification result, and determine whether the re-statistically obtained model prediction accuracy rate is smaller than the prediction accuracy rate of the model. set threshold; If the model prediction accuracy obtained from the re-statistics is not less than the preset threshold, the retrained data model based on the artificial neural network is used as the data model for acquiring feature information in the image to be detected; If the model prediction accuracy obtained from the re-statistics is less than the preset threshold, the above steps are repeated until the model prediction accuracy obtained by the verification set is not less than the preset threshold.

Acquire the data model based on the artificial neural network that has been trained, and extract the weight of the neuron and the offset value of the neuron in the data model according to the position information of the neuron.

Step S12, encrypt the parameter information, and generate a configuration file including the encrypted parameter information.

In an embodiment of the present invention, the step further includes performing a binarization process on the parameter information. The parameter information is compiled into binary code by a compiler.

The method for encrypting the parameter information includes any one of a symmetric encryption algorithm and an asymmetric encryption algorithm. The symmetric encryption algorithms include but are not limited to DES (Data Encryption Standard), IDEA (International Data Encryption Algorithm), and AES (Advanced Encryption Standard). The asymmetric encryption algorithm includes but is not limited to RSA (Rivest Shamir Adleman encryption algorithm), Elgamal encryption algorithm, and knapsack algorithm.

Generate a configuration file from the encrypted parameter information according to the format requirements of the compiler. The compiler includes but is not limited to C language, Python speech, and JAVA language.

Step S13, sending the configuration file to the security protection device.

In one embodiment, the method of outputting the configuration file may further include: receiving request information from a security protection device for sending a configuration file; and judging whether the security protection device has the ability to obtain the information according to the account information of the security protection device. the permission of the configuration file; if the security protection device has the permission, send the configuration file to the security protection device.

Please refer to FIG. 3 , which is a flowchart of a security protection method applied to a model parameter of a security protection device according to Embodiment 2 of the present invention. According to different requirements, the order of the steps in the flowchart can be changed, and some steps can be omitted.

Step S21, receiving the configuration file sent by the computer device.

Step S22, searching for encrypted parameter information in the configuration file.

In an embodiment of the present invention, the parameter information in the configuration file is searched by a keyword search method using a programming language preset in the security protection device.

Step S23: Obtain a decryption key, and decrypt the parameter information.

In an embodiment of the present invention, the step of obtaining a decryption key and decrypting the parameter information may include: receiving a decryption key sent by a computer device, and decrypting the parameter information by using the decryption key. The method is adapted to be implemented in the manner of asymmetric encryption as the encryption method of the acquired parameter information.

In yet another embodiment of the present invention, the step of obtaining the decryption key and decrypting the parameter information may further include: Search for a preset decryption program to decrypt the parameter information. The method is adapted to the encryption method of the parameter information by means of symmetric encryption.

Step S24, configure the parameter information in the data model.

In an embodiment of the present invention, the implementation steps of step S24 may include: Obtain the location information of the parameter in the parameter and parameter location comparison table of the data model. The safety protection device 2 pre-stores the type and purpose of the data model, the parameter name required by the data model and the location information of the parameter, and stores the parameter name and the location information of the parameter in the parameter correspondingly. and the parameter position comparison table.

The parameter name corresponding to the location information is determined from the data model, and the corresponding decryption parameter is searched in the decrypted parameter information according to the parameter name.

The searched decryption parameters are loaded into the data model at the location corresponding to the location information.

The above-mentioned FIG. 2 and FIG. 3 describe the security protection method of the model parameters of the present invention in detail. The following describes the hardware device architecture for realizing the security protection method of the model parameters with reference to FIG. 4 and FIG. 5 .

It should be understood that the embodiments are only used for illustration, and are not limited by this structure in the scope of the patent application.

Embodiment 3

FIG. 4 is a schematic diagram of a computer device according to Embodiment 3 of the present invention.

The computer device 1 includes a first memory 20, a first processor 30, and a first computer program 40 stored in the first memory 20 and executable on the first processor 30, such as model parameters security protection method program. When the first processor 30 executes the first computer program 40 , the steps in the above-mentioned embodiment of the security protection method for model parameters are implemented, for example, steps S11 to S14 shown in FIG. 2 .

The computer device 1 may be a computing device such as a desktop computer, a notebook, a palmtop computer, and a cloud server. Those skilled in the art can understand that the schematic diagram is only an example of the computer device 1, and does not constitute a limitation on the computer device 1. It may include more or less components than the one shown, or combine some components, or different Components, such as the computer device 1, may also include input and output devices, network access devices, bus bars, and the like.

The so-called first processor 30 may be a central processing unit (Central Processing Unit, CPU), and may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the first processor 30 can also be any conventional processor, etc. The first processor 30 is the control center of the computer device 1, and uses various interfaces and lines to connect the entire system. Various parts of the computer device 1 .

The first memory 20 can be used to store the first computer program 40 and/or modules/units, and the first processor 30 runs or executes the computer program stored in the first memory 20 and/or modules/units, and call the data stored in the first memory 20 to realize various functions of the computer device 1 . The first memory 20 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; The data storage area can store data (such as audio data, phone book, etc.) created according to the use of the computer device 1 and the like. In addition, the first memory 20 may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, at least one disk memory device, flash memory device, or other volatile solid state memory device.

If the modules/units integrated in the computer device 1 are implemented in the form of software functional units and sold or used as independent products, they may be stored in a computer-readable storage medium. Based on this understanding, the present invention realizes all or part of the processes in the methods of the above embodiments, and can also be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, When the computer program is executed by the processor, the steps of the above method embodiments can be implemented. Wherein, the computer program includes computer program code, and the computer program code may be in the form of original code, object code, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, removable hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-only memory) Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium Electric carrier signals and telecommunication signals are not included.

FIG. 5 is a schematic diagram of a safety protection device provided in Embodiment 3 of the present invention.

The safety protection device 2 includes a second memory 50, a second processor 60, and a second computer program 70, such as a model, stored in the second memory 50 and executable on the second processor 60 Parameter security method program. When the second processor 60 executes the second computer program 70 , the steps in the above-mentioned embodiment of the security protection method for model parameters are implemented, for example, steps S21 to S24 shown in FIG. 3 .

The security protection device 2 may be a computing device such as a smart phone or a smart wearable device. Those skilled in the art can understand that the schematic diagram is only an example of the safety protection device 2, and does not constitute a limitation on the safety protection device 2, and may include more or less components than the one shown, or combine certain components, or Different components, for example, the safety protection device 2 may also include input and output devices, network access devices, bus bars, and the like.

The so-called second processor 60 may be a central processing unit (Central Processing Unit, CPU), and may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the second processor 60 can also be any conventional processor, etc. The second processor 60 is the control center of the computer device 1, and uses various interfaces and lines to connect the entire Various parts of the computer device 1 .

The second memory 50 can be used to store the second computer program 70 and/or modules/units, and the second processor 60 executes or executes the computer program stored in the second memory 50 by running or executing the second computer program 70 . and/or modules/units, and call the data stored in the second memory 50 to realize various functions of the computer device 1 . The second memory 50 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; The storage data area can store data (such as audio data, phone book, etc.) created according to the use of the security device 2 . In addition, the second memory 50 may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, at least one disk memory device, flash memory device, or other volatile solid state memory device.

If the modules/units integrated in the security protection device 2 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the present invention realizes all or part of the processes in the methods of the above embodiments, and can also be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, When the computer program is executed by the processor, the steps of the above method embodiments can be implemented. Wherein, the computer program includes computer program code, and the computer program code may be in the form of original code, object code, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, removable hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-only memory) Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium Electric carrier signals and telecommunication signals are not included.

It will be apparent to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, but that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments are to be regarded in all respects as illustrative and not restrictive, and the scope of the present invention is defined by the appended claims rather than the foregoing description, and is therefore intended to fall within the scope of the claims. All changes within the meaning and range of the equivalents of , are included in the present invention. Any reference sign in a claim should not be construed as limiting the claim to which it relates. Furthermore, it is clear that the word "comprising" does not exclude other units or steps and the singular does not exclude the plural. A plurality of units or computer devices stated in the computer device claim may also be implemented by the same unit or computer device by software or hardware. The terms first, second, etc. are used to denote names and do not denote any particular order.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be Modifications or equivalent substitutions can be made without departing from the spirit and scope of the technical solutions of the present invention.

1: Computer device 2: Safety protection device 20: first memory 30: The first processor 40: The first computer program 50: Second memory 60: Second processor 70: Second computer program

FIG. 1 is a schematic diagram of an application environment architecture of a security protection method for model parameters provided by Embodiment 1 of the present invention.

FIG. 2 is a flowchart of a method for security protection of model parameters applied to a computer device according to Embodiment 2 of the present invention.

FIG. 3 is a flowchart of a safety protection method applied to a model parameter of a safety protection device according to Embodiment 2 of the present invention.

FIG. 4 is a schematic diagram of a computer device according to Embodiment 3 of the present invention.

FIG. 5 is a schematic diagram of a safety protection device provided in Embodiment 3 of the present invention.

Claims (10)

A security protection method for model parameters, the method is applied to a computer device, and the computer device is connected in communication with at least one security protection device. The improvement is that the method includes: training a data model, and obtaining parameter information in the data model; encrypting the parameter information, and generating a configuration file including the encrypted parameter information; Send the configuration file to the security device. The security protection method for model parameters according to claim 1, wherein the parameter information includes: one or more of neuron weights and neuron offset values in the artificial neural network road algorithm. The security protection method for model parameters according to claim 1, wherein the step of encrypting the parameter information and generating a configuration file including the encrypted parameter information includes: performing binarization processing on the parameter information; An encryption algorithm in the compiler is called to encrypt the binarized parameter information, wherein the encryption algorithm includes a symmetric encryption algorithm and an asymmetric encryption algorithm. The security protection method for model parameters according to claim 1, wherein the method includes: Receive the request information sent by the security device to send the configuration file; According to the account information of the security protection device, determine whether the security protection device has the permission to acquire the configuration file; If the security protection device has the permission, the configuration file is sent to the security protection device. A computer device, the improvement is that: the computer device includes a processor and a memory, and the processor is configured to implement any one of claim 1 to claim 4 when executing a computer program stored in the memory The security method of the model parameters. A safety protection method for model parameters, the method is applied in a safety protection device, the safety protection device communicates with a computer device, and is improved in that the method comprises: receiving a configuration file sent by the computer device; Find encrypted parameter information in the configuration file; Obtain the decryption key, and decrypt the parameter information; The parameter information is configured in the data model. The security protection method for model parameters according to claim 6, wherein the step of obtaining a decryption key and decrypting the parameter information comprises: Receive the decryption key sent by the computer device, and use the decryption key to decrypt the parameter information. The security protection method for model parameters according to claim 6, wherein the step of obtaining a decryption key and decrypting the parameter information further comprises: Search for a preset decryption program to decrypt the parameter information. The security protection method for model parameters according to claim 6, wherein the step of configuring the parameter information in the data model comprises: Obtain the location information of the parameters in the parameter and parameter location comparison table of the data model; Determine the parameter name corresponding to the location information from the data model, and search for the corresponding decryption parameter in the decrypted parameter information according to the parameter name; The searched decryption parameters are loaded into the data model at the location corresponding to the location information. A safety protection device, which is improved in that the device comprises: Memory: used to store computer programs; Processor: when executing the computer program in the memory, the security protection method for the model parameters described in any one of claim 6 to claim 9 is implemented.

Images