Standard file management method and device, electronic equipment and storage medium

Abstract

The invention relates to a standard file management method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining an input standard file, extracting text data of the standard file, and performing semantic analysis on the text data to obtain intention data of the standard file; establishing a two-stage training network model, and performing text authentication learning on the input standard file by using the two-stage training network model, specifically, inputting the intention data into a first-stage training network model, and performing training through the first-stage training network model to obtain feature data; performing down-sampling processing on the text data to obtain sampling data, and inputting the sampling data and the feature data into a second-stage training network model for training to obtain training data; performing up-sampling processing on the training data to obtain training text data; and comparing the training text data with a pre-stored standard file, and determining whether the input standard file is a legal file or not. According to the invention, the text management security is improved.

Description

Technical Field

[0001] The present disclosure relates to the verification and management technology of text depth information, and in particular to a management method and device for deep standard files, an electronic device, and a storage medium. Background Art

[0002] In recent years, with the continuous expansion of enterprises, the unified management of subsidiaries and employees of each enterprise has brought huge challenges. In some large enterprises, there are many standardized texts, which need to be used by subsidiaries and their employees as guidance manuals, so the uniqueness and security of the texts are extremely important attributes. However, the current standard texts are generally only used by publishing them on the server and calling them by different enterprises. The security management of texts through the server is easy to be exploited by management rights, and there is a risk of tampering with files. Moreover, since they are within the enterprise, these preventive measures are very limited, resulting in large loopholes in text management and the security of texts cannot be effectively guaranteed. Summary of the invention

[0003] The present disclosure provides a standard file management method and device, electronic device, and storage medium, which can at least solve the above-mentioned technical problems.

[0004] According to a first aspect of an embodiment of the present disclosure, a method for managing a standard file is provided, comprising:

[0005] Obtaining an input standard file, extracting text data of the standard file, and performing semantic analysis on the text data to obtain intent data of the standard file;

[0006] A two-level training network model is constructed, and the input standard file is subjected to text authentication learning using the two-level training network model. Specifically, the intent data is input into the first-level training network model, and the first-level training network model is used for training to obtain feature data; the text data is subjected to downsampling processing to obtain sampling data, and the sampling data and the feature data are input into the second-level training network model for training to obtain training data;

[0007] Performing upsampling processing on the training data to obtain training text data;

[0008] Comparing the training text data with a pre-stored standard file to determine whether the input standard file is a legal file;

[0009] The determined illegal standard files are removed from the standard file library.

[0010] As an optional method, the two-level training network model is constructed, and the two-level training network model is used to perform text authentication learning on the input standard file, including:

[0011] Setting the first-level training network model and the second-level training network model;

[0012] Inputting the intention data into the input end of the first-level training network model and outputting feature data;

[0013] wherein, the texture features in the intent data are obtained through the first-level training network model to determine the first depth information of the intent data in the first scene; the feature data is obtained based on the fusion of the first depth information and the downsampled text data processed by the first-level training network model;

[0014] Inputting the intention data and the feature data into a second-level training network model, and outputting training data;

[0015] Among them, the intention data and the feature data are fused through the second-level training network model to obtain the texture features of the fused first fused data to determine the second depth information of the first fused data in the first scene; the training data is obtained based on the fusion of the second depth information and the downsampled text data processed by the second-level training network model.

[0016] As an optional manner, the method further includes:

[0017] A dilated convolution training module is embedded in each level of the training network model; wherein dilated convolution parameters with different dilation parameters are inserted into the dilated convolution training module; and the dilation parameters of the dilated convolution of the encoder part of the dilated convolution training module are different from the dilation parameters of the dilated convolution of the decoder part of the dilated convolution training module.

[0018] As an optional manner, the method further includes:

[0019] Simulated data and real data in a first scenario for training are obtained, and a conversion model is trained based on the simulated data and real data for training to obtain a realistic conversion model.

[0020] According to a second aspect of an embodiment of the present disclosure, there is provided a standard file management device, including:

[0021] An acquisition unit, used for acquiring an input standard file, extracting text data of the standard file, and performing semantic analysis on the text data to obtain intention data of the standard file;

[0022] The model building unit is used to build a two-level training network model, and use the two-level training network model to perform text authentication learning on the input standard file. Specifically, the intention data is input into the first-level training network model, and the first-level training network model is used for training to obtain feature data; the text data is downsampled to obtain sampled data, and the sampled data and the feature data are input into the second-level training network model for training to obtain training data;

[0023] An upsampling unit, used for performing upsampling processing on the training data to obtain training text data;

[0024] A comparison unit, used to compare the training text data with a pre-stored standard file to determine whether the input standard file is a legal file;

[0025] The removal unit is used to remove the determined illegal standard files from the standard file library.

[0026] As an optional method, the model building unit is further used for:

[0027] Setting the first-level training network model and the second-level training network model;

[0028] Inputting the intention data into the input end of the first-level training network model and outputting feature data;

[0029] wherein, the texture features in the intent data are obtained through the first-level training network model to determine the first depth information of the intent data in the first scene; the feature data is obtained based on the fusion of the first depth information and the downsampled text data processed by the first-level training network model;

[0030] Inputting the intention data and the feature data into a second-level training network model, and outputting training data;

[0031] Among them, the intention data and the feature data are fused through the second-level training network model to obtain the texture features of the fused first fused data to determine the second depth information of the first fused data in the first scene; the training data is obtained based on the fusion of the second depth information and the downsampled text data processed by the second-level training network model.

[0032] As an optional method, the device further includes:

[0033] An embedding unit is used to embed a dilated convolution training module in each level of the training network model; wherein dilated convolution parameters with different dilation parameters are inserted into the dilated convolution training module; and the dilation parameters of the dilated convolution of the encoder part of the dilated convolution training module are different from the dilation parameters of the dilated convolution of the decoder part of the dilated convolution training module.

[0034] As an optional manner, the acquisition unit is further used to:

[0035] Simulated data and real data in a first scenario for training are obtained, and a conversion model is trained based on the simulated data and real data for training to obtain a realistic conversion model.

[0036] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device comprising a processor and a memory for storing processor executable instructions, wherein the processor is configured to execute the steps of the standard file management method when calling the executable instructions in the memory.

[0037] According to a fourth aspect of an embodiment of the present disclosure, a non-temporary computer-readable storage medium is provided, which, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute the steps of the standard file management method.

[0038] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:

[0039] The technical solution of the embodiment of the present disclosure performs data conversion on the input standard text data, and uses the converted intention data as the training data of the depth estimation model. When setting the depth estimation model, a two-level training network model is set, and each level of the training network model uses the output image data of the previous level as the input of the training network model of this level, so that the training data can be mutually verified, thereby improving the accuracy of text data training. By verifying the security of text data through training, standard files can be prevented from being tampered with, ensuring the security of standard texts.

[0040] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0042] Figure 1 A flowchart of a method for managing standard files shown in an embodiment of the present disclosure;

[0043] Figure 2A schematic diagram of the structure of a standard file management device shown in an embodiment of the present disclosure;

[0044] Figure 3 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0045] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0046] Figure 1 FIG. 1 is a flow chart of a method for managing standard files shown in an embodiment of the present disclosure. Figure 1 As shown, the standard file management method of the embodiment of the present disclosure may include the following processing steps:

[0047] Step 101, obtain an input standard file, extract text data of the standard file, perform semantic analysis on the text data, and obtain the intention data of the standard file.

[0048] In the disclosed embodiment, an input standard file is obtained, and the standard file is a file to be published. The text to be published is a file input by a manager with management authority through a specific management identity or a specific excuse. In theory, the standard file should be safe, but during the transmission process, due to transmission path problems or malicious interception, the standard file still has various possibilities of being tampered with. Therefore, it is necessary to perform semantic analysis on the text data of the input standard text to obtain its intended data.

[0049] Step 102, build a two-level training network model, and use the two-level training network model to perform text authentication learning on the input standard file. Specifically, the intent data is input into the first-level training network model, and the first-level training network model is used for training to obtain feature data; the text data is downsampled to obtain sampling data, and the sampling data and the feature data are input into the second-level training network model for training to obtain training data.

[0050] In the disclosed embodiment, the intention data is data obtained by processing the simulation data with a conversion algorithm. The simulation data and the intention data in the first scenario for training are obtained, and the conversion model is trained based on the simulation data and the intention data for training to obtain a realistic conversion model.

[0051] In the disclosed embodiment, the constructed two-level training network model is a neural network training algorithm, and the at least two-level training network model is provided with initial relevant parameter assignments. By continuously inputting the image data to be trained into the at least two-level training network model, the relevant parameters of the at least two-level training network model can be continuously adjusted to make it closer to a more optimal depth information determination method, and the image depth information determined by the at least two-level training network model is more accurate.

[0052] The building of a two-level training network model for supervised learning may include: setting the first-level training network model and the second-level training network model;

[0053] Inputting the intention data into the input end of the first-level training network model and outputting feature data;

[0054] wherein, the texture features in the intent data are obtained through the first-level training network model to determine the first depth information of the intent data in the first scene; the feature data is obtained based on the fusion of the first depth information and the downsampled text data processed by the first-level training network model;

[0055] Inputting the intention data and the feature data into a second-level training network model, and outputting training data;

[0056] Among them, the intention data and the feature data are fused through the second-level training network model to obtain the texture features of the fused first fused data to determine the second depth information of the first fused data in the first scene; the training data is obtained based on the fusion of the second depth information and the downsampled text data processed by the second-level training network model.

[0057] In the disclosed embodiment, in order to improve the estimation accuracy of the depth estimation model, it is necessary to embed a dilated convolution training module in each level of the training network model; wherein, dilated convolution parameters with different dilation parameters are inserted into the dilated convolution training module; the dilation parameters of the dilated convolution of the encoder part of the dilated convolution training module are different from the dilation parameters of the dilated convolution of the decoder part of the dilated convolution training module. As an implementation method, the dilation parameters of the four dilated convolutions of the encoder part are set to 1, 1, 2, 4, and the dilation parameters of the four dilated convolutions of the decoder part are set to 8, 4, 2, 1.

[0058] Step 103: up-sample the training data to obtain training text data.

[0059] In the disclosed embodiment, after the training data is obtained through the two-stage training network model, the training data is then subjected to upsampling recovery processing to obtain comparison data, and it is determined whether the recovered data meets the anti-tampering comparison parameters.

[0060] Step 104, comparing the training text data with a pre-stored standard file to determine whether the input standard file is a legal file.

[0061] In the disclosed embodiment, by comparing the training text data, when the comparison result exceeds the set threshold, it is determined that the input standard file has not been tampered with; otherwise, the input standard file is deemed to be an illegal file, and a reminder message of the standard file abnormality is output to the management personnel, and the next stage of processing of the input standard file, such as uploading to the server, is rejected.

[0062] Step 105 is used to remove the determined illegal standard files from the standard file library.

[0063] In the embodiment of the present application, the identified illegal standard files need to be removed to prevent the standard files from being eroded. In addition, for illegal standard files, it is also necessary to trace the source according to the identity of the transmitter, the text editor identifier, the transmission address, etc. to prevent malicious tampering, and locate and identify the tamperer to determine his identity, so as to completely prevent similar tampering.

[0064] The technical solution of the embodiment of the present disclosure performs data conversion on the input standard text data, and uses the converted intention data as the training data of the depth estimation model. When setting the depth estimation model, a two-level training network model is set, and each level of the training network model uses the output image data of the previous level as the input of the training network model of this level, so that the training data can be mutually verified, thereby improving the accuracy of text data training. By verifying the security of text data through training, standard files can be prevented from being tampered with, ensuring the security of standard texts.

[0065] Figure 2 The schematic diagram of the structure of the standard file management device shown in the embodiment of the present disclosure is as follows: Figure 2 As shown, the management device of the standard file of the embodiment of the present disclosure includes:

[0066] The acquisition unit 20 is used to acquire an input standard file, extract text data of the standard file, perform semantic analysis on the text data, and obtain the intention data of the standard file;

[0067] The model building unit 21 is used to build a two-level training network model, and use the two-level training network model to perform text authentication learning on the input standard file. Specifically, the intention data is input into the first-level training network model, and the first-level training network model is used for training to obtain feature data; the text data is downsampled to obtain sampled data, and the sampled data and the feature data are input into the second-level training network model for training to obtain training data;

[0068] An upsampling unit 22, configured to perform upsampling processing on the training data to obtain training text data;

[0069] A comparison unit 23 is used to compare the training text data with a pre-stored standard file to determine whether the input standard file is a legal file;

[0070] The removing unit 24 is used to remove the determined illegal standard files from the standard file library.

[0071] As an optional manner, the model building unit 21 is further used for:

[0072] Setting the first-level training network model and the second-level training network model;

[0073] Inputting the intention data into the input end of the first-level training network model and outputting feature data;

[0074] wherein, the texture features in the intent data are obtained through the first-level training network model to determine the first depth information of the intent data in the first scene; the feature data is obtained based on the fusion of the first depth information and the downsampled text data processed by the first-level training network model;

[0075] Inputting the intention data and the feature data into a second-level training network model, and outputting training data;

[0076] Among them, the intention data and the feature data are fused through the second-level training network model to obtain the texture features of the fused first fused data to determine the second depth information of the first fused data in the first scene; the training data is obtained based on the fusion of the second depth information and the downsampled text data processed by the second-level training network model.

[0077] As an optional method, the device further includes:

[0078] Embedded Unit ( Figure 2(not shown) is used to embed a dilated convolution training module in each level of the training network model; wherein dilated convolution parameters with different dilation parameters are inserted into the dilated convolution training module; and the dilation parameters of the dilated convolution of the encoder part of the dilated convolution training module are different from the dilation parameters of the dilated convolution of the decoder part of the dilated convolution training module.

[0079] As an optional manner, the acquisition unit 20 is further configured to:

[0080] Simulated data and real data in a first scenario for training are obtained, and a conversion model is trained based on the simulated data and real data for training to obtain a realistic conversion model.

[0081] In an exemplary embodiment, the above-mentioned processing units and the like may be implemented by one or more central processing units (CPU), graphics processing units (GPU), baseband processors (BP), application specific integrated circuits (ASIC), digital signal processors (DSP), programmable logic devices (PLD), complex programmable logic devices (CPLD), field programmable gate arrays (FPGA), general-purpose processors, controllers, microcontrollers (MCU), microprocessors, or other electronic components.

[0082] In the disclosed embodiment, Figure 2 The specific manner in which each unit in the management device for the standard file shown performs operations has been described in detail in the embodiment of the method, and will not be elaborated on here.

[0083] Figure 3 is a block diagram of an electronic device according to an exemplary embodiment. Figure 3As shown, electronic devices are intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0084] like Figure 3 As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the electronic device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0085] Multiple components in the electronic device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0086] The computing unit 801 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 801 performs the various methods and processes described above, such as the management method of standard files. For example, in some embodiments, the management method of standard files may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the management method of the standard file described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured as a management method of standard files by any other appropriate means (e.g., by means of firmware).

[0087] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0088] The program code for implementing the method of the present application can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, implements the functions / operations specified in the flow chart and / or block diagram. The program code can be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0089] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0090] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0091] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0092] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0093] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0094] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0095] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for managing standard documents, characterized in that: The method comprises: Obtaining an input standard file, extracting text data of the standard file, and performing semantic analysis on the text data to obtain intent data of the standard file; A two-level training network model is constructed, and the input standard file is subjected to text authentication learning using the two-level training network model. Specifically, the intent data is input into the first-level training network model, and the first-level training network model is used for training to obtain feature data; the text data is subjected to downsampling processing to obtain sampling data, and the sampling data and the feature data are input into the second-level training network model for training to obtain training data; Performing upsampling processing on the training data to obtain training text data; Comparing the training text data with a pre-stored standard file to determine whether the input standard file is a legal file; The determined illegal standard files are removed from the standard file library.

2. The method according to claim 1, characterized in that The two-level training network model is constructed, and the two-level training network model is used to perform text authentication learning on the input standard file, including: Setting the first-level training network model and the second-level training network model; Inputting the intention data into the input end of the first-level training network model and outputting feature data; wherein, the texture features in the intent data are obtained through the first-level training network model to determine the first depth information of the intent data in the first scene; the feature data is obtained based on the fusion of the first depth information and the downsampled text data processed by the first-level training network model; Inputting the intention data and the feature data into a second-level training network model, and outputting training data; Among them, the intention data and the feature data are fused through the second-level training network model to obtain the texture features of the fused first fused data to determine the second depth information of the first fused data in the first scene; the training data is obtained based on the fusion of the second depth information and the downsampled text data processed by the second-level training network model.

3. The method according to claim 1 or 2, characterized in that: The method further comprises: A dilated convolution training module is embedded in each level of the training network model; wherein dilated convolution parameters with different dilation parameters are inserted into the dilated convolution training module; and the dilation parameters of the dilated convolution of the encoder part of the dilated convolution training module are different from the dilation parameters of the dilated convolution of the decoder part of the dilated convolution training module.

4. The method according to claim 3, characterized in that The method further comprises: Simulated data and real data in a first scenario for training are obtained, and a conversion model is trained based on the simulated data and real data for training to obtain a realistic conversion model.

5. A standard file management device, characterized in that: The device comprises: An acquisition unit, used for acquiring an input standard file, extracting text data of the standard file, and performing semantic analysis on the text data to obtain intention data of the standard file; The model building unit is used to build a two-level training network model, and use the two-level training network model to perform text authentication learning on the input standard file. Specifically, the intention data is input into the first-level training network model, and the first-level training network model is used for training to obtain feature data; the text data is downsampled to obtain sampled data, and the sampled data and the feature data are input into the second-level training network model for training to obtain training data; An upsampling unit, used for performing upsampling processing on the training data to obtain training text data; A comparison unit, used to compare the training text data with a pre-stored standard file to determine whether the input standard file is a legal file; The removal unit is used to remove the determined illegal standard files from the standard file library.

6. The device according to claim 5, characterized in that The model building unit is further used for: Setting the first-level training network model and the second-level training network model; Inputting the intention data into the input end of the first-level training network model and outputting feature data; wherein, the texture features in the intent data are obtained through the first-level training network model to determine the first depth information of the intent data in the first scene; the feature data is obtained based on the fusion of the first depth information and the downsampled text data processed by the first-level training network model; Inputting the intention data and the feature data into a second-level training network model, and outputting training data; Among them, the intention data and the feature data are fused through the second-level training network model to obtain the texture features of the fused first fused data to determine the second depth information of the first fused data in the first scene; the training data is obtained based on the fusion of the second depth information and the downsampled text data processed by the second-level training network model.

7. The device according to claim 5 or 6, characterized in that The device also includes: An embedding unit is used to embed a dilated convolution training module in each level of the training network model; wherein dilated convolution parameters with different dilation parameters are inserted into the dilated convolution training module; and the dilation parameters of the dilated convolution of the encoder part of the dilated convolution training module are different from the dilation parameters of the dilated convolution of the decoder part of the dilated convolution training module.

8. The device according to claim 7, characterized in that The acquisition unit is further used for: Simulated data and real data in a first scenario for training are obtained, and a conversion model is trained based on the simulated data and real data for training to obtain a realistic conversion model.

9. An electronic device comprising a processor and a memory for storing instructions executable by the processor, characterized in that: The processor is configured to execute the steps of the standard file management method according to any one of claims 1 to 4 when calling the executable instructions in the memory.

10. A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the steps of the standard file management method according to any one of claims 1 to 4.

Images