Credential processing method, system, device, and readable storage medium

By using the diffusion model to generate credential images on the blockchain, the problems of overdisclosure and privacy leakage of traditional credential information are solved, and the intuitive expression and privacy protection of credential content are realized, which improves data security and interactive traceability.

WO2024230354A1PCT designated stage expired Publication Date: 2024-11-14GUANGDONG INSPUR BIG DATA RES CO LTD

Patent Information

Application Number
PCT/CN2024/084653
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-06
Filing Date
2024-03-29
Publication Date
2024-11-14

AI Technical Summary

Technical Problem

Traditional physical credentials and their electronic methods have the risk of excessive disclosure of information or privacy data leakage, and the statement content is not intuitive, and the credential inspection operations lack identity authentication, rights confirmation and authorization, and data interaction cannot be traced.

Method used

Using a blockchain-based verifiable credential system, the diffusion model is used to generate credential images, convert text description information into images, store and map it into a unified resource locator, provide intuitive two-dimensional image information, hide specific disclosure information, and realize credentials Protection of content and privacy data.

Benefits of technology

It improves the credential usage experience, protects private data, realizes traceability and immutability of data interaction, and reduces the risk of information leakage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2024084653_14112024_PF_FP_ABST
    Figure CN2024084653_14112024_PF_FP_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of data security, and discloses a credential processing method, a system, a device, and a non-volatile readable storage medium. The method comprises: acquiring character description information in a digital trusted credential of a target work; generating, by using a diffusion model, a credential image corresponding to the character description information; storing the credential image, and mapping a storage path to a universal resource locator; and providing the credential image to the outside by using the universal resource locator. In the present application, the credential image for describing the character description information of the digital trusted credential is generated for the digital trusted credential by using the diffusion model, so that visual and summary two-dimensional image information for credential content is provided, and the credential use experience of the user is improved; in addition, specific disclosure information of the credential is hidden, that is, the character description information is hidden, and in the credential interaction process, especially in scenarios where it is not necessary to disclose specific data, the content of a corresponding work and privacy data can be effectively protected.
Need to check novelty before this filing date? Find Prior Art

Description

A credential processing method, system, device and readable storage medium

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese patent application filed with the China Patent Office on May 6, 2023, with application number 202310505780.4 and application name “A voucher processing method, system, device and readable storage medium”, all contents of which are incorporated by reference into this application. Technical Field

[0003] The present application relates to the field of data security technology, and in particular to a credential processing method, system, device, and non-volatile readable storage medium. Background Art

[0004] Traditional physical credentials are often issued by an authoritative institution, such as ID cards, driver's licenses, and educational certificates. Electronic versions of traditional physical credentials represent the electronic and information-based nature of traditional paper credentials. They feature centralized authentication, with data authenticity backed by an authoritative institution and relying on a centralized platform for verification. Once a traditional paper credential is created, its form and content cannot be altered unless physically damaged. Reading electronic credentials may require access to storage using specialized equipment. However, when presented to a third party for verification, they cannot mask unnecessary information, posing the risk of over-disclosure or privacy breaches.

[0005] In other words, traditional physical credentials and their electronic methods have the risk of excessive information disclosure or privacy data leakage, and the declaration content is not intuitive; in addition, the credential verification operation does not include identity authentication, title confirmation, authorization and other operations, and data interaction cannot be traced.

[0006] In summary, how to effectively solve problems such as voucher processing is a technical problem that technical personnel in this field urgently need to solve.

[0007] Summary of the Invention

[0008] The purpose of this application is to provide a credential processing method, system, device and non-volatile readable storage medium. In this application, a diffusion model is used to generate a credential image that describes the credential summary information (that is, verifiable text description information, such as statement content) for a digital trusted credential, providing intuitive and summary two-dimensional image information of the credential content, thereby improving the user's credential usage experience; at the same time, the specific disclosure information of the credential is also hidden, and during the credential interaction process (in scenarios where it is not necessary to disclose specific data), the content and privacy data of the corresponding work are effectively protected.

[0009] To solve the above technical problems, this application provides the following technical solutions:

[0010] A credential processing method, applied to a blockchain-based verifiable credential system, comprising:

[0011] Obtaining text description information from the digital trusted certificate of the target work;

[0012] Using the diffusion model, a voucher image corresponding to the text description information is generated;

[0013] Storing the credential image and mapping the storage path to a uniform resource locator;

[0014] Use the uniform resource locator to provide the credential image to the outside world.

[0015] In some embodiments of the present application, a diffusion model is used to generate a voucher image corresponding to text description information, including:

[0016] Generate candidate images corresponding to text description information using the diffusion model;

[0017] From the candidate images, a voucher image is selected.

[0018] In some embodiments of the present application, a credential image is provided externally using a uniform resource locator, including:

[0019] Receiving a request for access to a uniform resource locator;

[0020] Display the credential image in the visual interface.

[0021] In some embodiments of the present application, further comprising:

[0022] receiving a request for viewing detailed description information of a credential for a credential image;

[0023] After authorization by the owner of the target work, text description information will be fed back.

[0024] In some embodiments of the present application, after the text description information is fed back, the following is also included:

[0025] receiving a request for viewing a target work;

[0026] After the owner of the target work has authorized it, the target work will be fed back.

[0027] In some embodiments of the present application, further comprising:

[0028] Store authorization records in the blockchain.

[0029] In some embodiments of the present application, storing a credential image includes:

[0030] Store credential images on the blockchain.

[0031] In some embodiments of the present application, after storing the credential image and mapping the storage path to a uniform resource locator, the method further includes:

[0032] Generate verification data for the credential image;

[0033] Store verification data and URLs on the blockchain.

[0034] In some embodiments of the present application, obtaining text description information in the digital trusted certificate of the target work includes:

[0035] Use the smart contract interface to obtain text description information from the blockchain.

[0036] In some embodiments of the present application, before obtaining text description information from the blockchain using the smart contract interface, the following steps are included:

[0037] Using the smart contract interface, the target work is written into the blockchain ledger;

[0038] Generate digital trusted credentials for the target work.

[0039] In some embodiments of the present application, a diffusion model is used to generate a voucher image corresponding to text description information, including:

[0040] Convert text description information into text embedding vector representation;

[0041] Based on the text embedding vector representation, a conditional embedding representation containing image details is obtained;

[0042] The conditional embedding representation is decoded into the image pixel space to generate the voucher image.

[0043] In some embodiments of the present application, obtaining a conditional embedding representation containing image details based on a text embedding vector representation includes:

[0044] Based on the cross-attention mechanism, the text embedding vector representation is transformed to obtain a conditional embedding representation containing image details.

[0045] In some embodiments of the present application, obtaining a conditional embedding representation containing image details based on a text embedding vector representation includes:

[0046] Convert the latent space representation to the input representation;

[0047] Apply frequency-based positional embeddings to the input representation;

[0048] The input representation is converted using the text embedding vector representation to obtain the conditional embedding representation.

[0049] In some embodiments of the present application, decoding the conditional embedded representation into an image pixel space to generate a voucher image includes:

[0050] Mapping data from image pixel space to latent space;

[0051] The data in the latent space is mapped to the image pixel space, resulting in a voucher image with high-frequency details removed and compressed to an abstract representation.

[0052] A blockchain-based verifiable credential system, comprising:

[0053] A voucher image generation module is used to obtain the text description information in the digital credential of the target work; and generate a voucher image corresponding to the text description information using a diffusion model;

[0054] A data storage module, configured to store the credential image and map the storage path to a uniform resource locator;

[0055] The credential verification module is used to provide a credential image to the outside world using a uniform resource locator.

[0056] In some embodiments of the present application, further comprising:

[0057] Smart contract interface, used to implement data interaction with the blockchain.

[0058] In some embodiments of the present application, including:

[0059] The work chain module is used to write digital work data into the blockchain ledger through a pre-defined smart contract interface;

[0060] The on-chain data query module is used to read the corresponding data from the blockchain through a pre-defined smart contract interface.

[0061] In some embodiments of the present application, including:

[0062] The voucher image generation module is specifically used to call the remote text description generation AI (Artificial Intelligence) image service module to generate a voucher image.

[0063] In some embodiments of the present application, including:

[0064] Service interface, used to provide Web (World Wide Web) services;

[0065] Identity authentication module, used to complete user identity login and authority control;

[0066] Work registration module, used to register and manage digital works;

[0067] The work chain module is used to write digital work data into the blockchain ledger;

[0068] On-chain data query module, used to read corresponding data from the blockchain;

[0069] The certificate template management module is used to add, delete, modify and query certificate templates;

[0070] The work copyright certificate management module is used to provide copyright certificate application, viewing and deletion services for work registration users.

[0071] An electronic device, comprising:

[0072] memory for storing computer programs;

[0073] A processor is used to implement the steps of the above-mentioned credential processing method when executing a computer program.

[0074] A non-volatile readable storage medium stores a computer program, which implements the steps of the above-mentioned credential processing method when executed by a processor.

[0075] In a verifiable credential system based on blockchain, the method provided by some embodiments of the present application is applied, including: obtaining text description information in the digital trusted credential of the target work; using a diffusion model to generate a credential image corresponding to the text description information; storing the credential image and mapping the storage path to a uniform resource locator; and using the uniform resource locator to provide the credential image to the outside.

[0076] In the present application, the text description information in the digital trusted certificate of the target work is obtained. Then, the diffusion model is used to generate a certificate image corresponding to the text description information. The certificate image is stored, and the storage path is mapped to a uniform resource locator. Then, the certificate image can be provided to the outside with the help of the uniform resource locator. That is to say, in the present application, the diffusion model is used to generate a certificate image that describes the certificate summary information (i.e., text description information, such as declaration content) for the digital trusted certificate, providing intuitive and summary two-dimensional image information of the certificate content, thereby improving the user's certificate usage experience; at the same time, the specific disclosure information of the digital trusted certificate is also hidden, that is, the text description information is hidden. During the certificate interaction process, especially in the scenario where it is not necessary to disclose specific data, the content and privacy data of the corresponding work can be effectively protected.

[0077] Accordingly, some embodiments of the present application also provide a blockchain-based verifiable credential system, device and non-volatile readable storage medium corresponding to the above-mentioned credential processing method, which has the above-mentioned technical effects and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate some embodiments of the present application or the technical solutions in related technologies, the following briefly introduces the drawings required for use in some embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0079] FIG1 is a flowchart of an implementation method of a credential processing method in some embodiments of the present application;

[0080] FIG2 is a schematic diagram of image generation based on a diffusion model in some embodiments of the present application;

[0081] FIG3 is a schematic diagram of the structure of a blockchain-based verifiable credential system in some embodiments of the present application;

[0082] FIG4 is a schematic diagram of the structure of a blockchain-based verifiable credential system in some embodiments of the present application;

[0083] FIG5 is a schematic structural diagram of an electronic device in some embodiments of the present application;

[0084] FIG6 is a schematic diagram of a specific structure of an electronic device in some embodiments of the present application;

[0085] FIG7 is a schematic diagram of an image generation process in some embodiments of the present application. DETAILED DESCRIPTION

[0086] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific embodiments. Obviously, some of the embodiments described are only a portion of the embodiments of the present application, rather than all of the embodiments. Based on some of the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making any creative efforts are within the scope of protection of this application.

[0087] Please refer to Figure 1, which is a flowchart of a credential processing method in some embodiments of the present application. This method can be applied to a verifiable credential system based on blockchain. The detailed information such as the architecture of this system can be shown in Figure 3 and will not be repeated here. The method includes the following steps:

[0088] S101. Obtain text description information in the digital trusted certificate of the target work.

[0089] In this application, the target work can be any digital work, such as self-media articles, poems, novels, papers and other text digital works, and of course it can also be digital works such as audio and video.

[0090] Verifiable Credentials (VCs) differ from physical credentials in their electronic form, fundamentally different at the identity, application, and trust levels. They rely on a distributed digital identity (DID) system, offering distributed verifiability. DIDs differ from traditional identifiers in that data is autonomously stored rather than centrally managed; identifiers are self-generated, self-assigned, and self-managed rather than centrally assigned; and data authenticity is backed by cryptographic algorithms rather than by an authoritative body, thus reducing the risk of data breaches. VC credentials support distributed verification, meaning the authenticity of the credential data can be verified on any distributed node. Data sharing for VC credentials requires authorization from the holder, and selective disclosure policies based on data attributes can be implemented to minimize information disclosure. In a blockchain-based distributed digital identity system, entity identity authentication, ownership confirmation, and authorization operations can be performed on-chain, ensuring that data interactions are traceable, verifiable, and tamper-proof.

[0091] The declaration content of the VC certificate is information in text form, that is, text description information.

[0092] In a specific embodiment of the present application, obtaining the text description information in the digital trusted certificate of the target work includes: using the smart contract interface to obtain the text description information from the blockchain. Of course, before obtaining the text description information from the blockchain, it includes:

[0093] Using the smart contract interface, the target work is written into the blockchain ledger;

[0094] Generate digital trusted credentials for the target work.

[0095] In other words, the target work can be a work written into a blockchain ledger, and data interaction with the blockchain is achieved through a smart contract interface. The digital trusted certificate can be a digital trusted certificate generated after the target work is written into the blockchain.

[0096] In this application, the text description information can specifically be the certificate summary description information in the digital trusted certificate of the target work, such as the declaration content, which can be used to describe the descriptive text content of the certificate of the target work.

[0097] S102: Generate a voucher image corresponding to the text description information using a diffusion model.

[0098] Diffusion models are also generative models. The intuition behind these models comes from physics. In physics, gas molecules diffuse from areas of high concentration to areas of low concentration, similar to the information loss caused by noise interference. Therefore, by introducing noise and then attempting to remove the noise, the model generates an image. Over multiple iterations over time, the model learns to generate new images given some noise input.

[0099] Specifically, in this application, text description information is input into the diffusion model, and based on the characteristics of the diffusion model, a voucher image corresponding to the text description information is generated. If the text description information is different, the generated voucher image will be a new image, thus enabling the generation of different voucher images for different digital works.

[0100] In a specific embodiment of the present application, during the process of generating a voucher image using a diffusion model, the images generated by the diffusion model can also be screened / selected so that the final voucher image is consistent with the text description information or meets the expectations of the owner of the target work. That is, step S102, generating a voucher image corresponding to the text description information using a diffusion model, includes:

[0101] Step 1: Generate candidate images corresponding to text description information using a diffusion model;

[0102] Step 2: Select a credential image from the candidate images.

[0103] For ease of description, the above two steps are combined for explanation below.

[0104] When using the diffusion model to generate images corresponding to text descriptions, different candidate images can be generated multiple times, or multiple sequences of candidate images can be generated at once. These candidate images can then be displayed to the user who registered the target work, allowing them to select one or more voucher images from the candidate images.

[0105] For example, let's assume the target article is an article covering the Shenzhou manned spacecraft. Its summary describes the text: "Shenzhou XII manned spacecraft successfully docked with the Chinese space station." The candidate images generated using the diffusion model could include images of the manned spacecraft docking with the space station, images of the manned spacecraft, images of the space station, and so on. Users can select a supporting image for the article from these images.

[0106] In a specific embodiment of the present application, step S101, using a diffusion model to generate a voucher image corresponding to the text description information, includes:

[0107] Step 1: Convert text description information into text embedding vector representation;

[0108] Step 2: Based on the text embedding vector representation, obtain the conditional embedding representation containing image details;

[0109] Step 3: Decode the conditional embedding representation into the image pixel space to generate the voucher image.

[0110] For ease of description, the above three steps are combined and explained below.

[0111] In this application, in the diffusion model, the text description information can be converted into a text embedding vector representation, and then based on the text embedding vector representation, a conditional embedding representation containing image details is obtained, and finally the conditional embedding representation is decoded into the image pixel space to obtain the voucher image.

[0112] In a specific implementation of the present application, the above step 2 may specifically include but is not limited to the following two specific implementation methods:

[0113] Method 1: Based on the cross-attention mechanism, the text embedding vector representation is transformed to obtain a conditional embedding representation containing image details.

[0114] In this application, the text embedding vector can be transformed based on the cross-attention mechanism to obtain the conditional embedding representation.

[0115] In a specific implementation of the present application, the above step three may specifically include:

[0116] Step 1: Map the data in the image pixel space to the latent space;

[0117] Step 2: Map the data in the latent space to the image pixel space to obtain a voucher image that removes high-frequency details and compresses it to an abstract representation.

[0118] Specifically, a variational autoencoder (VAE) can be used in the perceptual compression stage to map high-dimensional data in the pixel space to a latent space. The latent vector in the latent space is a compressed representation of the original image. This compression improves computational efficiency and introduces complex operations. The model consists of two parts: an encoder and a decoder. The encoder maps the high-dimensional data in the pixel space to the latent space, while the decoder restores the image from the latent space to the pixel space. The model compresses the data into an abstract representation by removing high-frequency details, which helps to build a stable representation.

[0119] For example, if the diffusion model includes a text input component, a text encoder, a U-Net model (a deep learning network for image semantic segmentation), a variational autoencoder, and an image output component, then step S102 uses the diffusion model to generate a voucher image, including:

[0120] Step 1: Use the text input component to receive text description information;

[0121] Step 2: Use the text encoder to generate a text embedding vector representation corresponding to the text description information;

[0122] Step 3: Using the U-Net model, we can obtain a conditional embedding representation containing image details based on the text embedding vector representation.

[0123] Specifically, the cross-attention layer in the U-Net model can be used to map the text embedding vector representation to the middle layer of the U-Net model to obtain the conditional embedding representation.

[0124] Step 4: Use the variational autoencoder to decode the conditional embedding representation into the image pixel space to generate the voucher image;

[0125] Specifically, the text input component receives text description information sent from the outside.

[0126] The text encoder uses the Transformer encoder architecture, a technique often used in natural language processing. The Transformer model demonstrates generalizability and adaptability not only in text processing but also in visual applications, making it suitable for multimodal learning tasks. The Transformer model excels at capturing semantic structure in text and images, combining its generalization capabilities with the detail-preserving capabilities of diffusion models.

[0127] The U-Net model is used in the semantic compression stage. Image generation methods need to capture the semantic structure present in the conditional input. The Transformer text encoding encodes the conditional text and image into latent embeddings and forms a cross-attention with the U-Net model. The attention calculation formula is as follows:

[0128] Where Q, K and V are learnable projection matrices corresponding to query, key and value matrices respectively, i.e., the input consists of a matrix of dimension d k The query, key, and dimension is d v The value of .

[0129] The cross-attention layer maps the fused representation to the middle layer of U-Net. The concepts and semantic structures expressed in the text provide the context and mutual relationships of various objects in the image. The U-Net model provides conditional embedding representations for generating image details and is able to preserve the semantic structure in the text.

[0130] The variational autoencoder (VAE) is used in the perceptual compression stage to map high-dimensional data from the pixel space to a latent space. The latent vector in the latent space is a compressed representation of the original image. This compression improves computational efficiency but introduces complex operations. It consists of two parts: an encoder and a decoder. The encoder maps the high-dimensional data from the pixel space to the latent space, while the decoder restores the image from the latent space to the pixel space. The model compresses the data into an abstract representation by removing high-frequency details, which helps to build a stable representation.

[0131] The image output component outputs the generated image (which may contain a set of ordered images). The output image can be a candidate image or a credential image.

[0132] For ease of understanding, please refer to FIG. 2 below, and combine the data change process of converting text to image to explain the image generation process in detail.

[0133] The latent diffusion model includes a cross-attention mechanism that fuses the conditional input and the image representation present in the current latent space.

[0134] First, the text description module receives text input and uses a text encoder to generate a text embedding vector representation that preserves key information such as concepts and semantic structure. A similar encoding process can be used for other types of conditional input (such as image data) as a guide.

[0135] Secondly, the U-Net model can obtain conditional latent representations containing image details. The U-Net model can adjust its output on text embedding through the cross-attention layer. The operations after U-Net can be regarded as an upsampling process.

[0136] Again, a variational autoencoder (VAE) consists of an encoder and a decoder. The encoder is used to convert an image from pixel space to a low-dimensional potential representation, and the decoder can be seen as the reverse process of the encoder, using which the image can be reconstructed in pixel space.

[0137] Finally, the model uses the transformed and denoised input in the latent space to generate the final high-resolution image.

[0138] Method 2: Convert the latent space representation to the input representation;

[0139] Apply frequency-based positional embeddings to the input representation;

[0140] The input representation is converted using the text embedding vector representation to obtain the conditional embedding representation.

[0141] In this application, the latent space representation can be converted into an input representation, and then the frequency-based position embedding is applied to all input representations, and finally the input representation is converted using the text embedding vector representation to obtain a conditional embedding representation. Specifically, the conversion processing here may include but is not limited to the in-context conditioning processing block (such as noise time step, class label, natural language, etc.), the cross-attention processing block, the adaptive layer norm (adaLN) processing block, and the adaLN-zero processing block for conversion processing.

[0142] The conditional information processing block adds two additional conditional representations (tokens) to the input sequence by embedding the vectors of the noise time step t and the class label c, treating them differently from the image representation. The standard ViT block is used for processing (unmodified), and after the last block, the conditional representations are removed from the sequence. This approach introduces minimal (essentially negligible) computational overhead to the model.

[0143] Cross-Attention Block: This is an attention mechanism in the Transformer architecture that mixes two different embedding sequences. These two sequences must have the same dimensions and can be in different modalities (e.g., text, sound, image). One sequence serves as the input Q, which defines the output sequence length, while the other sequence provides the input K&V. The attention calculation formula is as follows:

[0144] Adaptive layer norm (adaLN) processing block: Adaptive layer norm (adaLN) replaces the standard layer norm layer (LayerNorm, a common normalization method) in the Transformer block. Instead of directly learning the dimension scale and shift parameters γ and β, adaLN regresses them from the sum of the embedding vectors of the noisy time step t and the class label c. adaLN adds little computational overhead and is computationally efficient.

[0145] adaLN-zero processing block: Initializing each residual block to the identity function is beneficial. The Diffused U-Net model uses a similar initialization strategy, zero-initializing the final convolutional layer in each block before any residual connections. The adaLN-zero block has the same functionality, initializing the Multi-Layer Perception (MLP) to output a zero vector for all alphas. adaLN-zero is based on the identity function (initializing the MLP to output a zero vector for all alphas), and experimental results show that it can significantly reduce computational overhead.

[0146] For example, in this approach, the U-Net model in the model structure shown in Figure 2 can be replaced with a conversion model. The conversion model then first passes the VAE latent space representation input through the first layer of the network and linearly embeds each data slice into the input, thereby converting the spatial input into a sparse input representation (token).

[0147] Among them, data slicing processing, that is, given an input representation (size: I x I x C), after data slicing processing by specifying a slicing vector (size: pxp), a length of T = (I / p) can be obtained. 2 A data shard sequence, and a hidden dimension d.

[0148] The model then applies standard Vision Transformers (ViTs) frequency-based positional embeddings to all input representations.

[0149] Finally, the input representation is processed by a series of transformer blocks to obtain a denoised latent representation. These transformer blocks include in-context conditioning processing blocks (such as noise time steps, class labels, natural language, etc.), cross-attention processing blocks, adaptive layer norm (adaLN) processing blocks, and adaLN-zero processing blocks.

[0150] Among them, the Adaptive layer norm (adaLN) module: adaLN can be used to replace the native LayerNorm to reduce the risk of overfitting;

[0151] adaLN-zero module: It is beneficial to use the same initialization function for each residual module in ResNets (Residual Networks). Diffusion Transformers (DiT) attenuate the parameters γ, β, and α of the residual module to achieve a similar purpose.

[0152] During model training, the diffusion transformation model processes data as shown in Figure 7. The input is a spatial representation Z = I x I x C (for example, for a 256 × 256 × 3 image, the matrix shape of Z is 32 × 32 × 4). After data slicing, a data slicing sequence T = (I / p) is obtained. 2 , T can be processed in four different ways (including conditional information processing, cross-attention processing, adaLN processing, and adaLN-zero processing) to generate latent representations. Taking conditional information processing as an example, the input data includes the data segment sequence T and the attached conditional representations (such as noise time step t, class label c, natural language), which are mainly processed by normalization (LayerNorm) and multi-head self-attention mechanism.

[0153] When the input is a text sequence, the processing block will segment the input text sequence and vectorize the text through the word embedding matrix. The self-attention mechanism compares each word with all other words in the input text to calculate the attention score, which is used to evaluate the contribution of other words in the text to the current word, as shown in formula (1). Through the multi-head self-attention mechanism, each attention can handle different aspects and learn from different perspectives, as shown in formula (2). Then, the attention score is used as the weight of the word representation and enters the feedforward neural network through weighted aggregation to generate a new representation for the word.

[0154] Formula (1) describes the scaled dot product attention, where Q, K, V are query, key, and value matrices. The input consists of a matrix of dimension d k The query and key, and the dimension is d v The value of .

[0155] Where W 0 is the parameter matrix.

[0156] After the image is output, step S103 is executed to save the voucher image to the local file system, and the file storage path is mapped to a URL (Uniform Resource Locator) for external access. External programs can download the image file through the URL.

[0157] S103: Store the credential image and map the storage path to a uniform resource locator.

[0158] The credential image can be stored directly in the blockchain, or the verification data of the credential image can be stored in the blockchain to ensure that the credential image is not tampered with.

[0159] Among them, the Universal Resource Locator (URL) can be mapped through the storage path of the credential image. For the specific mapping method, please refer to the generation principle and function of the URL, which will not be described in detail here.

[0160] Specifically, the generated credential image (which may contain a set of ordered pictures) can be saved to the local file system, and the file storage path is mapped to a URL for external access, through which external programs can download the image file.

[0161] In a specific embodiment of the present application, after storing the credential image and mapping the storage path to a uniform resource locator, the following steps may be performed:

[0162] Step 1: Generate verification data for the credential image;

[0163] Step 2: Store the verification data and uniform resource locator on the blockchain.

[0164] First, verification data for the credential image is generated. This data can be calculated using a hash algorithm or other verification algorithm. This verification data and a uniform resource locator (URL) are then stored on the blockchain, allowing access to the reliable credential image based on the URL and verification data.

[0165] When saving a credential image, it can be saved as a credential image file. For example, the credential image file can be stored on the blockchain; or the credential image file URL and its verification data (such as MD5 (MD5 Message-Digest Algorithm)) can be stored on the blockchain; the credential image file is stored in the server-side file system, an access URL is provided, and the image file path is associated with the corresponding credential in the database of the system.

[0166] S104: Use the uniform resource locator to provide the credential image to the outside.

[0167] After step S103 is completed, the credential image can be accessed via the URL. That is, the configuration image can be provided externally based on the URL, for example, a service of accessing the configuration image via the URL is provided to the user, or an interface for accessing the configuration image via the URL is provided to an upper layer application.

[0168] Since the credential image is presented in the form of a two-dimensional image, it can provide verification to the user more intuitively. In addition, since the credential image is not a descriptive text information, the image can also protect the privacy of the credential content.

[0169] In a specific embodiment of the present application, step S104 uses a uniform resource locator to provide a credential image externally, including:

[0170] Step 1: Receive a Uniform Resource Locator access request;

[0171] Step 2: Display the voucher image on the visual interface.

[0172] For ease of description, the above two steps are combined for explanation below.

[0173] When a user or an application needs to display the credential image, an access request for the uniform resource locator may be issued, and then the access request is responded to, thereby displaying the configuration image on the visual interface.

[0174] For example, when the user selects the subject corresponding to the credential image, it can be regarded as receiving the access request, and then the credential image can be directly displayed on the visual interface for the user to check.

[0175] In a specific implementation of the present application, after executing step S104, the following steps may be further performed:

[0176] Step 1: Receive a request for viewing detailed description information of a credential for a credential image;

[0177] Step 2: After the owner of the target work authorizes you, provide text description information.

[0178] For ease of description, the above two steps are combined for explanation below.

[0179] That is, in this application, when a user views a voucher image and is interested in the specific voucher content represented by the voucher image, a request for viewing the voucher detailed description information for the voucher image may be sent.

[0180] After receiving the viewing request, the system can remind the owner of the target work to authorize the operation, and after the owner of the target work authorizes the operation, feedback text description information. In other words, the specific content of the voucher is displayed to the user who requested the viewing.

[0181] Furthermore, after providing the text description information, you can also perform the following steps:

[0182] Step 1: Receive a request to view the target work;

[0183] Step 2: After the owner of the target work has authorized you, provide feedback on the target work.

[0184] For ease of description, the above two steps are combined for explanation below.

[0185] That is, when a user views the text description information and becomes interested in the target work itself, a viewing request for the target work may be initiated.

[0186] After the system receives a request to view the target work, it can also feedback the target work to the requester after obtaining authorization from the owner of the target work.

[0187] In order to make the authorization record searchable and to ensure that the authorization record cannot be tampered with, the authorization record can also be stored in the blockchain. Due to the tamper-proof nature of the blockchain, the authorization record is safe, reliable and searchable.

[0188] In a verifiable credential system based on blockchain, the method provided by some embodiments of the present application is applied, including: obtaining text description information in the digital trusted credential of the target work; using a diffusion model to generate a credential image corresponding to the text description information; storing the credential image and mapping the storage path to a uniform resource locator; and using the uniform resource locator to provide the credential image to the outside.

[0189] In the present application, the text description information in the digital trusted certificate of the target work is obtained. Then, the diffusion model is used to generate a certificate image corresponding to the text description information. The certificate image is stored, and the storage path is mapped to a uniform resource locator. Then, the certificate image can be provided to the outside with the help of the uniform resource locator. That is to say, in the present application, the diffusion model is used to generate a certificate image that describes the certificate summary information (i.e., text description information, such as declaration content) for the digital trusted certificate, providing intuitive and summary two-dimensional image information of the certificate content, thereby improving the user's certificate usage experience; at the same time, the specific disclosure information of the digital trusted certificate is also hidden, that is, the text description information is hidden. During the certificate interaction process, especially in the scenario where it is not necessary to disclose specific data, the content and privacy data of the corresponding work can be effectively protected.

[0190] Corresponding to some of the above method embodiments, some embodiments of the present application also provide a blockchain-based verifiable credential system. The blockchain-based verifiable credential system described below and the credential processing method described above can be referenced to each other.

[0191] As shown in Figure 3, the blockchain-based verifiable credential system includes:

[0192] A voucher image generation module is used to obtain the text description information in the digital credential of the target work; and generate a voucher image corresponding to the text description information using a diffusion model;

[0193] A data storage module, configured to store the credential image and map the storage path to a uniform resource locator;

[0194] The credential verification module is used to provide a credential image to the outside world using a uniform resource locator.

[0195] In a verifiable credential system based on blockchain, the text description information in the digital trusted credential of the target work is obtained; a credential image corresponding to the text description information is generated using a diffusion model; the credential image is stored, and the storage path is mapped to a uniform resource locator; and the credential image is provided to the outside using the uniform resource locator.

[0196] In the present application, the text description information in the digital trusted certificate of the target work is obtained. Then, the diffusion model is used to generate a certificate image corresponding to the text description information. The certificate image is stored, and the storage path is mapped to a uniform resource locator. Then, the certificate image can be provided to the outside with the help of the uniform resource locator. That is to say, in the present application, the diffusion model is used to generate a certificate image that describes the certificate summary information (i.e., text description information, such as declaration content) for the digital trusted certificate, providing intuitive and summary two-dimensional image information of the certificate content, thereby improving the user's certificate usage experience; at the same time, the specific disclosure information of the digital trusted certificate is also hidden, that is, the text description information is hidden. During the certificate interaction process, especially in the scenario where it is not necessary to disclose specific data, the content and privacy data of the corresponding work can be effectively protected.

[0197] Furthermore, the system also includes:

[0198] Smart contract interface, used to implement data interaction with the blockchain.

[0199] Correspondingly, the work on-chain module is specifically used to write the digital work data into the blockchain ledger through a pre-defined smart contract interface;

[0200] The on-chain data query module is specifically used to read the corresponding data from the blockchain through a pre-defined smart contract interface.

[0201] It should be noted that in this system, the modules implementing the aforementioned method embodiments and the described credential processing methods can be located within the system or outside the system and serve the system through invocation. That is, in one specific embodiment of this application, the credential image generation module is specifically configured to invoke a remote text description generation AI image service module to generate a credential image.

[0202] In a specific embodiment of the present application, the system specifically includes:

[0203] Service interface, used to provide Web services;

[0204] Identity authentication module, used to complete user identity login and authority control;

[0205] Work registration module, used to register and manage digital works;

[0206] The work chain module is used to write digital work data into the blockchain ledger;

[0207] On-chain data query module, used to read corresponding data from the blockchain;

[0208] The certificate template management module is used to add, delete, modify and query certificate templates;

[0209] The work copyright certificate management module is used to provide copyright certificate application, viewing and deletion services for work registration users.

[0210] For example, Figure 2 shows an implementation system for blockchain-based verifiable credential image generation. It consists of two components: a digital work verifiable credential system and a credential image generation module (i.e., a text description generation AI image service module). The text description generation AI image service module can be a functional module within the digital work verifiable credential system, or it can be independently developed as a service entity providing an AI image generation engine. As shown in Figure 2, the AI ​​image generation service can be provided to external applications via RESTful APIs. RESTful refers to a design and development approach for web applications based on HTTP (Hypertext Transfer Protocol) and can be defined in XML (Extensible Markup Language) or JSON (JavaScript Object Notation, a lightweight data exchange format). A RESTful API is a REST-style API (Application Programming Interface). REST (Representational State Transfer) describes an architectural style for web systems that is programming language-independent and platform-independent, using HTTP as the transport protocol.

[0211] To facilitate understanding of the blockchain-based verifiable credential system provided in this application, its architecture, and its functional roles, the credential processing system is explained below in combination with relevant technologies and specific application scenarios.

[0212] Blockchain has been widely used in the field of digital work certification, including related applications such as NFTs (Non-Fungible Tokens) and digital collectibles. These applications leverage blockchain's characteristics (such as cryptographic credit endorsement, immutability, data transparency, and traceability) to record the copyright of specific works and artworks, generating unique digital certificates for the works. This protects the digital copyrights while enabling the exchange of digital works' certificates, better supporting the distribution, trading, collection, and use of the works.

[0213] Decentralized Identifier (DID) and Verifiable Credential (VC) are W3C (World Wide Web Consortium) standards and specifications. A distributed identifier is a globally unique digital identifier that is distributed, resolvable, and verifiable. A verifiable credential is typically a descriptive statement issued by an authoritative entity to endorse certain attributes of another entity, with its own digital signature attached. In addition to the DID and VC standards, W3C also defines three related specifications: DID Document, DID Resolver, and Identity Hub. The above five technical specifications are widely used in the construction of distributed digital identity and verifiable digital credential systems to ensure that digital identities and digital credentials can interoperate across different networks and platforms.

[0214] Given the vulnerability of digital intellectual property to copying, photographing, and pirating, this poses significant challenges to both copyright protection and content privacy. While there are numerous intellectual property protection measures for text-based intellectual property works such as self-media articles, poems, novels, papers, and patents, the dissemination process inevitably leads to information leakage.

[0215] A blockchain-based digital work storage system can use digital credentials to create copyright certificates for a large number of digital works. It also utilizes selective disclosure technology for digitally verifiable credentials to shield unnecessary data from disclosure, enhancing privacy and content protection. A digital certificate contains a certificate identification code (DID), the issuer's signature, a declaration, and credential metadata. The declaration is key to the certificate and contains information that must be disclosed (such as gender, age, author, and work summary) for verification by the credential verifier. The credential verifier verifies that the certificate is issued by the issuer, verifies the identity of the user and the credential issuer, and verifies the disclosed data (partial or complete declaration). If the declaration is too simple or lacks necessary information, the verifiable certificate (such as a copyright certificate for a work) will be unclear, making it difficult to understand its value and verify key data. If the declaration is too detailed, it may over-disclose information, exposing data content and privacy risks.

[0216] Therefore, this application provides a method for disclosing declaration content, which can not only protect privacy and work data content, but also provide intuitive and verifiable key data to the verifier.

[0217] As can be seen from the above, this application applies the AI ​​model of generating images from text descriptions to a verifiable credential system based on blockchain, and proposes an image representation method for verifiable credentials, thereby realizing the protection of credential declaration content and privacy data; proposes an image generation framework for verifiable credentials based on blockchain, and provides a flexible Web service for the text-generated image model in the form of a RESTful interface to realize the generation function of verifiable credential images; and constructs an AI image generation function module to realize the process of generating pictures from text.

[0218] Specifically, the blockchain-based verifiable credential system can build a distributed digital identity system that complies with international standards based on the W3C's Decentralized Identifier (DID) and Verifiable Credential (VC) protocol specifications, and implement a distributed decentralized infrastructure based on the blockchain, while storing key data (such as credential templates, credential description files, etc.) on the blockchain.

[0219] During the actual system development and implementation process, open source projects that comply with standards and specifications can be leveraged, and open source components can be used to quickly build the system. In this application, based on the open source consortium chain platform (FISCO BCOS, Blockchain Open Consortium Operating System), the open source blockchain management component (WeBASE, WeBank Blockchain Application Software Extension), and the open source distributed digital identity component (WeIdentity), a blockchain-based digital work verifiable credential system that complies with W3C DID and VC specifications is constructed.

[0220] As shown in Figure 4, the digital work verifiable credential system consists of two parts: front-end and back-end, which exchange data through a RESTful interface. The back-end uses the open source blockchain management component (WeBASE) to manage blockchain data and the open source distributed digital identity component (WeIdentity) to implement distributed digital identity (DID) and verifiable credential interaction functions. The open source consortium chain platform (FISCO BCOS) (the consortium chain shown in the figure) serves as the blockchain node network foundation, and MySQL (a relational database management system) is used to store key data (including work data, credential templates, credential files, user data, etc.). Users of the digital work verifiable credential system include ordinary anonymous users, work registration users, credential issuing organizations, and administrators.

[0221] The application program interface for connecting to the database may specifically adopt JDBC (Java Database Connectivity, an application program interface in the Java language used to standardize how a client program accesses a database, and provides methods such as querying and updating data in the database).

[0222] The database may specifically be MySQL (an open source database).

[0223] The blockchain management platform shown in the figure can be specifically WeBASE, WeBASE (WeBank Blockchain Application Software Extension), an open source blockchain management component, a set of common components built between blockchain applications and FISCO BCOS nodes.

[0224] The identity system platform can be specifically WeIdentity, which is an open source distributed digital identity system component and a distributed multi-center technical solution.

[0225] In other words, the credential processing method (a method using AI image generation) and blockchain-based verifiable credential system provided by this application not only achieve an intuitive image representation of the verifiable credential content (such as the statement content), but also provide an image summary of the verifiable credential content (such as the statement content), hiding the specific content of the text information, and protecting content and privacy when disclosure is not necessary. Specific technical key points include:

[0226] (1) This application applies an AI model that generates images from text descriptions to a verifiable credential system based on blockchain, and proposes an image representation method for verifiable credentials to achieve protection of the content of verifiable credentials (such as statement content) and privacy data.

[0227] (2) This application proposes a blockchain-based verifiable credential image generation framework, which provides image generation services to external applications through a web service, thereby realizing the generation function of verifiable credential images.

[0228] (3) This application constructs an AI image generation function module to realize the process of generating pictures from text.

[0229] (4) The generated credential image is automatically generated based on the AI ​​algorithm model, so the generated credential image is original.

[0230] (5) Adding corresponding summary images to the voucher improves the user's experience satisfaction when viewing the voucher.

[0231] (6) Utilize AI algorithm models to automatically generate credential images without the need for manual image production, thus avoiding the additional cost of manual image production.

[0232] (7) Introducing AI models to generate credential images can make the system or product clearly different from similar systems or products, making the system or product unique.

[0233] Those skilled in the art will appreciate that the method and system proposed in this application can extend the generation of credential images and the expression of credential declaration content through images to other application fields such as medical credentials, academic credentials, civil affairs credentials (including social welfare security certificates, bravery, retired soldiers, and distribution of disaster relief funds and materials), etc.

[0234] Corresponding to the above method embodiments, some embodiments of the present application further provide an electronic device. The electronic device described below and the credential processing method described above can refer to each other.

[0235] As shown in FIG5 , the electronic device includes:

[0236] Memory 332, for storing computer programs;

[0237] The processor 322 is configured to implement the steps of the credential processing method of some of the above method embodiments when executing a computer program.

[0238] Specifically, please refer to Figure 6, which is a schematic diagram of the specific structure of an electronic device provided in some embodiments of the present application. The electronic device may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 322 (for example, one or more processors) and a memory 332, and the memory 332 stores one or more computer programs 342 or data 344. Among them, the memory 332 can be a temporary storage or a permanent storage. The program stored in the memory 332 may include one or more modules (not marked in the figure), and each module may include a series of instruction operations in the data processing device. Furthermore, the processor 322 can be configured to communicate with the memory 332 to execute a series of instruction operations in the memory 332 on the electronic device 301.

[0239] The electronic device 301 may further include one or more power supplies 326 , one or more wired or wireless network interfaces 350 , one or more input / output interfaces 358 , and / or one or more operating systems 341 .

[0240] The steps in the credential processing method described above can be implemented by the structure of an electronic device.

[0241] Corresponding to some of the above method embodiments, some embodiments of the present application further provide a non-volatile readable storage medium. The non-volatile readable storage medium described below and the credential processing method described above can refer to each other.

[0242] A non-volatile readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the credential processing method of the above method embodiment.

[0243] The readable storage medium may specifically be any readable storage medium that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0244] Some embodiments of this application are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the same or similar parts between some embodiments. For the systems disclosed in some embodiments of this application, since they correspond to the methods disclosed in some embodiments of this application, the description is relatively simple, and the relevant parts can be referred to the method description.

[0245] Those skilled in the art may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0246] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM (compact disc read-only memory), or any other form of non-volatile readable storage medium known in the art.

[0247] Finally, it should be noted that, in this document, relationships such as first and second, etc., are used solely to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms include, comprise, or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0248] This document uses some specific embodiments to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core ideas of the present application. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A voucher processing method, characterized in that: Applications in blockchain-based verifiable credential systems include: Obtaining text description information in the digital trusted certificate of the target work; Generate a voucher image corresponding to the text description information using a diffusion model; storing the credential image and mapping the storage path to a uniform resource locator; The credential image is provided externally using the uniform resource locator.

2. The voucher processing method according to claim 1, characterized in that: Using a diffusion model, generating a voucher image corresponding to the text description information includes: Using the diffusion model, generating a candidate image corresponding to the text description information; The voucher image is selected from the candidate images.

3. The voucher processing method according to claim 1, characterized in that: Providing the credential image externally by using the uniform resource locator includes: receiving an access request for the uniform resource locator; The credential image is displayed on a visual interface.

4. The voucher processing method according to claim 1, characterized in that: Also includes: receiving a request for viewing detailed credential information of the credential image; After the owner of the target work authorizes, the text description information is fed back.

5. The voucher processing method according to claim 4, characterized in that: After the text description information is fed back, it also includes: Receiving a request for viewing the target work; After the owner of the target work authorizes, the target work is fed back.

6. The voucher processing method according to claim 4, characterized in that: Also includes: An authorization record is stored in the blockchain.

7. The voucher processing method according to claim 1, characterized in that: Storing the credential image, comprising: The credential image is stored on the blockchain.

8. The voucher processing method according to claim 1, characterized in that: After storing the credential image and mapping the storage path to a uniform resource locator, the method further includes: generating verification data for the credential image; The verification data and the uniform resource locator are stored on the blockchain.

9. The voucher processing method according to claim 1, characterized in that: Obtain the text description information in the digital trusted certificate of the target work, including: The text description information is obtained from the blockchain using a smart contract interface.

10. The voucher processing method according to claim 9, characterized in that: Before obtaining the text description information from the blockchain using the smart contract interface, the following steps are included: Using a smart contract interface, writing the target work into the ledger of the blockchain; Generate a digital trusted credential for the target work.

11. The voucher processing method according to any one of claims 1 to 10, characterized in that: Using a diffusion model, generating a voucher image corresponding to the text description information includes: Convert the text description information into a text embedding vector representation; Based on the text embedding vector representation, obtaining a conditional embedding representation containing image details; The conditional embedded representation is decoded into an image pixel space to generate the voucher image.

12. The voucher processing method according to claim 11, characterized in that: Based on the text embedding vector representation, a conditional embedding representation containing image details is obtained, including: Based on the cross-attention mechanism, the text embedding vector representation is transformed to obtain a conditional embedding representation containing image details.

13. The voucher processing method according to claim 11, characterized in that: Based on the text embedding vector representation, a conditional embedding representation containing image details is obtained, including: Convert the latent space representation to the input representation; applying a frequency-based positional embedding to the input representation; The input representation is converted using a text embedding vector representation to obtain the conditional embedding representation.

14. The voucher processing method according to claim 11, characterized in that: Decoding the conditional embedded representation into an image pixel space to generate the voucher image includes: Mapping data from image pixel space to latent space; The data in the latent space is mapped to the image pixel space, thereby obtaining the voucher image with high-frequency details removed and compressed to an abstract representation.

15. The voucher processing method according to claim 1, characterized in that: The step of using a diffusion model to generate a voucher image corresponding to the text description information includes: Obtaining the text description information; Using a text encoder, generating a text embedding vector representation corresponding to the text description information; Using a cross attention layer in a U-Net model, mapping the text embedding vector representation to an intermediate layer of the U-Net model to obtain a conditional embedding representation; The conditional embedding representation is decoded into an image pixel space using a variational autoencoder to generate the voucher image.

16. The voucher processing method according to claim 15, characterized in that: The method of using a variational autoencoder to decode the conditional embedding representation into an image pixel space to generate the voucher image includes: The variational autoencoder includes an encoder and a decoder; converting the voucher image from the image pixel space into a low-dimensional latent representation using the encoder; The voucher image is reconstructed in the image pixel space using the decoder.

17. The voucher processing method according to claim 13, characterized in that: Also includes: Passing the latent space representation input of the variational autoencoder through the first layer network of the conversion model so that each data slice is linearly embedded into the latent space representation input, and converting the latent space representation input into a sparse input representation; applying a standard visual transformer frequency-based positional embedding to all said input representations; The input representation is transformed to obtain the denoised conditional embedding representation.

18. A verifiable credential system based on blockchain, characterized in that: include: A voucher image generation module is used to obtain text description information in the digital trusted voucher of the target work; Generate a voucher image corresponding to the text description information using a diffusion model; A data storage module, used for storing the credential image and mapping the storage path to a uniform resource locator; The credential verification module is used to provide the credential image to the outside by using the uniform resource locator.

19. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the credential processing method as claimed in any one of claims 1 to 17 when executing the computer program.

20. A non-volatile readable storage medium, characterized in that: The non-volatile readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the credential processing method according to any one of claims 1 to 17 are implemented.

Citation Information

Patent Citations

  • Image data transmission device, control method and program

    CN102172034A

  • Multi-dimensional digital identity authentication system based on block chain

    CN112580102A

  • Information security interaction method and device

    CN113704580A

  • Data verification method and device, computer, storage medium and program product

    CN115174260A

  • Voucher processing method, system and device and readable storage medium

    CN116522293A

Cited By

  • Accounting data management method, system and equipment based on edge computing and medium

    CN121304358A