Digital human image generation method and device, electronic equipment and storage medium

Abstract

This application provides a method, apparatus, electronic device, and storage medium for generating digital human images, applicable to the field of blockchain technology. The method includes: acquiring parameter information of a pre-generated digital human; distributing the parameter information to multiple clothing designers based on a pre-configured smart contract, allowing the designers to design and create various clothing and accessories; receiving feedback from the multiple clothing designers on the clothing and accessories, and writing the clothing and accessories into the blockchain; and generating multiple digital human images based on the clothing and accessories and the digital human. This method distributes the parameter information of the digital human to different clothing designers to design suitable clothing and accessories, thereby forming different digital human images through a rich combination of clothing and accessories. This reduces the generation cycle of digital human images and improves the generation efficiency and quality.

Description

Technical Field

[0001] This application relates to the field of blockchain technology, and in particular to a method, apparatus, electronic device, and storage medium for generating digital human images. Background Technology

[0002] Digital humans, a product of the deep integration of artificial intelligence and computer graphics, are digital human figures that closely resemble human appearances, constructed using advanced algorithms. They are created through technologies such as computer graphics, 3D modeling, motion capture, and speech synthesis to achieve realistic human-like appearances, and utilize AI capabilities such as natural language processing and deep learning to achieve real-time interaction with humans.

[0003] Under the current technological system, the creation of digital human images mainly relies on the combination of different clothing and accessories. However, this traditional method of digital human image generation is characterized by long creation cycles and low efficiency. It is difficult for digital human creators to launch a large number of images with different styles in a short period of time, and they cannot meet the dynamic demands of the market in a timely manner. This seriously restricts the innovation and development of digital human images.

[0004] Therefore, how to break through the bottlenecks in existing digital human image design and improve the efficiency and quality of image generation has become an important issue that urgently needs to be addressed in the field of digital human technology. Summary of the Invention

[0005] This application provides a method, apparatus, electronic device, and storage medium for generating digital human images, in order to solve the technical problem of low efficiency in the generation of digital human images in existing digital human image generation technologies.

[0006] According to the first aspect disclosed in this application, this application provides a method for generating a digital human image, including:

[0007] Obtain parameter information of a pre-generated digital human;

[0008] Based on a pre-configured smart contract, the parameter information is distributed to multiple clothing designers, who then design and create various clothing accessories according to the parameter information.

[0009] Receive clothing and accessories feedback from multiple clothing designers and write the clothing and accessories into the blockchain;

[0010] Based on the clothing and accessories and the digital human, multiple digital human images are generated.

[0011] In one feasible implementation, based on the clothing and accessories and the digital human, multiple digital human images are generated, including:

[0012] Read clothing and accessories from the blockchain to create a clothing and accessories resource library;

[0013] The clothing and accessories in the clothing material library are combined to create multiple outfit schemes;

[0014] Based on the clothing and accessories in the aforementioned outfit scheme and the digital human, a digital human image is formed.

[0015] In one feasible implementation, clothing and accessories from the outfit material library are combined to form multiple outfit schemes, including:

[0016] The clothing and accessories in the clothing material library are input into a pre-built clothing combination model to obtain multiple clothing schemes output by the clothing combination model; wherein, the clothing combination model is obtained based on a deep learning model training.

[0017] In one feasible implementation, clothing and accessories from the outfit material library are combined to form multiple outfit schemes, including:

[0018] The clothing and accessories in the clothing material library are input into a pre-built tag extraction model to obtain the feature extraction results of each clothing and accessory; wherein, the tag extraction model is trained based on a deep learning model, and the feature extraction results include feature tags under multiple preset classification dimensions;

[0019] Based on preset matching rules, clothing and accessories in the outfit material library are combined using feature tags to obtain multiple outfit schemes.

[0020] In one feasible implementation, the method further includes:

[0021] The digital human image is sent to the reviewer based on a pre-configured smart contract;

[0022] Obtain the first review result from the reviewer and write the first review result into the blockchain;

[0023] Read the first review result on the blockchain. If the first review result indicates that the review is passed, write the digital human image into the blockchain.

[0024] In one feasible implementation, the method further includes:

[0025] Each piece of clothing and accessories read from the blockchain is input into a pre-built element decomposition model to obtain multiple clothing elements obtained from the decomposition of each piece of clothing and accessories; wherein, the element decomposition model is obtained by training a deep learning model;

[0026] The clothing elements are input into a pre-built clothing recombination model to obtain reconstructed clothing; wherein the clothing recombination model is obtained based on a deep learning model.

[0027] In one feasible implementation, the method further includes:

[0028] The reconstructed clothing is sent to the reviewer based on a pre-configured smart contract;

[0029] Obtain the second review result from the reviewer and write the second review result into the blockchain;

[0030] Read the second audit result on the blockchain. If the second audit result indicates that the audit is passed, write the reconstructed clothing into the blockchain.

[0031] According to a second aspect disclosed in this application, this application provides a digital human image generation apparatus, comprising:

[0032] The data acquisition module is used to acquire parameter information of the pre-generated digital human;

[0033] The data distribution module is used to distribute the parameter information to multiple clothing designers based on a pre-configured smart contract, so that the clothing designers can design and create a number of clothing accessories based on the parameter information.

[0034] The data writing module is used to receive clothing and accessories feedback from multiple clothing designers and write the clothing and accessories into the blockchain.

[0035] The image generation module is used to generate multiple digital human images based on the clothing and accessories and the digital human.

[0036] According to a third aspect disclosed in this application, this application provides an electronic device, including a processor and a memory communicatively connected to the processor;

[0037] The memory stores computer-executed instructions;

[0038] The processor executes computer execution instructions stored in the memory to implement the method described in any one of the first aspects.

[0039] According to the fourth aspect disclosed in this application, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed, are used to implement the method described in any one of the first aspects.

[0040] According to the fifth aspect disclosed in this application, this application provides a computer program product, including a computer program, which, when executed, is used to implement the method described in any one of the first aspects.

[0041] Compared with the prior art, this application has the following advantages:

[0042] This application provides a method, apparatus, electronic device, and storage medium for generating digital human images. By distributing the parameter information of the digital human to different clothing designers, the clothing designers can design clothing and accessories suitable for the digital human based on the parameter information. This can quickly increase the number of clothing and accessories and form different outfit combinations. Then, the clothing and accessories in the outfit combinations are combined with the digital human to form different digital human images. This reduces the generation cycle of digital human images and improves the generation efficiency and quality of digital human images. Attached Figure Description

[0043] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0044] Figure 1 A flowchart illustrating a digital human image generation method provided in an embodiment of this application;

[0045] Figure 2 A flowchart illustrating another method for generating a digital human image provided in an embodiment of this application;

[0046] Figure 3 This is a schematic diagram of the structure of a digital human image generation device provided in an embodiment of this application;

[0047] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0048] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0049] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0050] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0051] Furthermore, the technical solution involved in this application, which involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.) and the use of artificial intelligence technology for automated decision-making, and makes decisions that have a significant impact on personal rights based on the results of automated decision-making, provides users with corresponding operation entry points for users to choose to agree to or reject the results of automated decision-making; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0052] Digital humans, a product of the deep integration of artificial intelligence and computer graphics, are digital avatars that closely resemble human figures, constructed using advanced algorithms. They are created through technologies such as computer graphics, 3D modeling, motion capture, and speech synthesis to achieve lifelike human appearances and enable real-time interaction with humans using AI capabilities such as natural language processing and deep learning. Digital humans can mimic human facial expressions, body movements, and voice intonation to become virtual anchors and brand ambassadors active on social media, or provide professional services as intelligent customer service representatives and educational mentors, and even simulate human emotional expression through affective computing. Their core advantage lies in overcoming physical limitations, enabling 24 / 7 uninterrupted operation, and rapidly switching identities and skills according to scenario requirements. They demonstrate broad application prospects in entertainment, finance, and healthcare, gradually reshaping a new paradigm of human-machine collaboration.

[0053] Under the current technological system, the image creation of digital humans mainly relies on the combination of different clothing and accessories. Specifically, designers carefully select clothing of various styles and match them with a wide variety of accessories. By cleverly combining these different elements, they create a unique external image for digital humans, enabling them to better suit different application scenarios and user needs.

[0054] However, in this traditional method of digital human character generation, the design of clothing and accessories is usually undertaken by the digital human generator. The design team often needs to start from scratch, undertaking a series of complex workflows including conceptualization, drawing, and modeling. Each step requires a significant investment of time and energy. Due to the long creation cycle and low efficiency, the digital human generator struggles to produce a large number of diverse character designs in a short period, failing to meet the dynamic demands of the market in a timely manner. This severely restricts the innovation and development of digital human characters. Therefore, how to break through the current bottlenecks in digital human character design and improve the efficiency and quality of character generation has become a crucial issue that urgently needs to be addressed in the field of digital human technology.

[0055] To address the aforementioned technical issues, this application proposes a method, apparatus, electronic device, and storage medium for generating digital human images. By distributing the parameter information of the digital human to different clothing designers to design suitable clothing and accessories, different digital human images are formed through a rich combination of clothing and accessories. This reduces the generation cycle of digital human images and improves the generation efficiency and quality of digital human images.

[0056] It should be noted that the digital human image generation method, device, electronic device and storage medium provided in this application can be used in the field of blockchain technology, or in any field other than blockchain. The application field of the digital human image generation method, device, electronic device and storage medium in this application is not limited.

[0057] The technical solution of the digital human image generation method provided in this application will be described in detail below through specific embodiments. It should be noted that the following embodiments may exist alone or in combination with each other, and the same or similar content may not be described again in different embodiments.

[0058] It should be noted that the execution entity of the digital human image generation method provided in this application embodiment is the server of the digital human platform, and correspondingly, the digital human image generation device is also set in the server of the digital human platform.

[0059] Figure 1 A flowchart illustrating a digital human image generation method provided in this application embodiment is shown below. Figure 1 In some embodiments, the digital human image generation method includes the following steps:

[0060] S101, Obtain the parameter information of the pre-generated digital human.

[0061] Specifically, the parameters of a digital human include height, weight, chest circumference, waist circumference, and other parameters.

[0062] S102 distributes parameter information to multiple clothing designers based on pre-configured smart contracts, allowing the designers to create various clothing accessories according to the parameter information.

[0063] By sending the generated digital human's parameter information to clothing designers, a design standard can be provided to multiple clothing designers. Based on this standard, clothing and accessories created by different designers, such as hats, glasses, and clothes, can be adapted to this digital human. Since all clothing and accessories are based on the same design standard, compatibility and consistency with the digital human can be ensured. For clothing designers, having the specific parameter information of the digital human eliminates the need to design the adaptability of each garment and accessory from scratch, greatly simplifying the design process and improving efficiency.

[0064] Specifically, using smart contracts to transmit parameter information fully leverages the distributed architecture of blockchain, eliminating reliance on a single centralized server. This effectively avoids the risk of data loss or leakage due to central node failure or attacks, greatly enhancing the reliability and security of data transmission. Furthermore, the transparent and immutable nature of its code ensures clear, unchanging data transmission rules, allowing all parties to interact according to established rules. This guarantees that data is not maliciously tampered with or forged during transmission, ensuring data integrity and authenticity. Moreover, smart contracts can automatically execute preset logic, quickly completing data transmission and processing once conditions are met, eliminating the need for tedious manual operations and audits. This not only improves transmission efficiency but also reduces human error and costs, providing strong support for secure, efficient, and reliable data transmission.

[0065] S103 receives clothing and accessories feedback from multiple clothing designers and writes the clothing and accessories into the blockchain.

[0066] Among these methods, writing the designed clothing and accessories into the blockchain can leverage the immutable and decentralized nature of the blockchain to create a unique and trustworthy digital identity for the clothing and accessories, thereby eliminating counterfeits from the design source and protecting the rights of original designs.

[0067] S104 generates multiple digital human figures based on clothing, accessories, and digital humans.

[0068] In this process, clothing and accessories from the outfit scheme are combined with digital humans, and the clothing and accessories are worn on the digital humans to form a digital human image.

[0069] In this embodiment, by distributing the parameter information of the digital human to different clothing designers, the clothing designers can design clothing and accessories that are suitable for the digital human based on the parameter information. This can quickly increase the number of clothing and accessories, and form different outfit combinations with the rich clothing and accessories. Then, the clothing and accessories in the outfit combinations are combined with the digital human to form different digital human images. This reduces the generation cycle of digital human images and improves the generation efficiency and quality of digital human images.

[0070] exist Figure 1 Based on the embodiments shown, the following is combined with Figure 2 The technical solution of the above-mentioned digital human image generation method will be further introduced.

[0071] Figure 2 A flowchart illustrating another digital human image generation method provided in this application embodiment is shown below. Figure 2 In some embodiments, the digital human image generation method includes the following steps:

[0072] S201, Obtain parameter information of the pre-generated digital human.

[0073] S202 distributes parameter information to multiple clothing designers based on pre-configured smart contracts, allowing the designers to create various clothing and accessories according to the parameters.

[0074] S203 receives clothing and accessories feedback from multiple clothing designers and writes the clothing and accessories into the blockchain.

[0075] S204: Input each piece of clothing and accessories read from the blockchain into a pre-built element decomposition model to obtain multiple clothing elements obtained from the decomposition of each piece of clothing and accessories; wherein, the element decomposition model is obtained by training a deep learning model.

[0076] In addition to clothing and accessories created by designers, these items can also be deconstructed and recombined to generate new ones, further enriching the variety of clothing and accessories available. The element deconstruction model breaks down clothing and accessories into clothing elements, making it easier to combine these elements to create new garments and accessories.

[0077] Specifically, the construction of the element decomposition model first requires collecting a large dataset containing clothing and accessories. Each garment and accessory is then meticulously labeled using manual annotation or semi-automatic algorithms to clearly distinguish its different elements, such as neckline style, button shape, patterns, and material texture, thus building a rich and accurately labeled training dataset. Next, a suitable deep learning architecture, such as a convolutional neural network (CNN), is selected, and the model is trained using the labeled dataset. By continuously adjusting the network parameters, the model learns the mapping relationship between clothing and accessories and their constituent elements. During training, an appropriate loss function is used to measure the difference between the model's predictions and the actual labels, and optimization algorithms are used to continuously minimize the loss and improve model performance. Simultaneously, a validation set is used to evaluate the trained model, and hyperparameters are adjusted based on the evaluation results to prevent overfitting. Finally, a test set is used to comprehensively test the trained model, ensuring that it can accurately and stably decompose various clothing and accessories into their corresponding clothing elements.

[0078] S205, input the clothing elements into the pre-built clothing recombination model to obtain the recombined clothing; wherein, the clothing recombination model is obtained based on the training of a deep learning model.

[0079] Among them, the clothing recombination model can quickly and automatically combine clothing elements disassembled into recombined clothing to enrich the quantity of clothing and accessories.

[0080] Specifically, the construction of the clothing recombination model first involves collecting a large number of different styles of clothing and accessories, along with their various disassembled clothing elements, and simultaneously collecting other new clothing samples that are reasonably combined from these clothing elements as the target output. Next, a suitable generative model architecture is selected, such as a Generative Adversarial Network (GAN) or a Variational Autoencoder (VAE). For example, taking a Generative Adversarial Network (GAN) as an example, a GAN includes a generator and a discriminator. The model is trained using prepared data, allowing the generator to learn how to cleverly fuse and arrange the elements based on the input features to create new clothing and accessories that conform to aesthetic and design logic. Simultaneously, the discriminator continuously distinguishes the generated new clothing and accessories from real clothing samples, prompting the generator to improve its generation quality. During training, model parameters are continuously adjusted, the loss function is optimized, and the model performance is monitored using a validation set to prevent overfitting. Finally, the model is evaluated using a test set to ensure that it can stably and effectively recombine the disassembled clothing elements into novel and reasonable recombined clothing.

[0081] Optionally, to ensure the generated reconstructed clothing has practical value, the reconstructed clothing can be reviewed, specifically including:

[0082] Step 1: Send the reconstructed clothing to the reviewer based on a pre-configured smart contract.

[0083] Among these measures, the reconstituted clothing was reviewed through manual verification.

[0084] Step 2: Obtain the second review result from the reviewer and write the second review result into the blockchain.

[0085] In this process, writing the audit results to the blockchain leverages the blockchain's immutability to ensure that once the results are on the chain, they cannot be arbitrarily modified or deleted, greatly enhancing the credibility and authority of the audit results. This effectively prevents data from being maliciously tampered with for illicit gain, ensuring the fairness and seriousness of the audit. Furthermore, the blockchain's distributed ledger technology allows audit results to be recorded and stored by multiple nodes, enabling all participants to view and verify them in real time. This improves information transparency and sharing, avoids information silos, and helps all parties conduct subsequent work based on accurate and reliable audit results. In addition, the blockchain's encryption mechanism provides strong security for the audit results, ensuring the confidentiality of data during transmission and storage, preventing the leakage of sensitive information, and protecting the privacy and rights of relevant parties.

[0086] Step 3: Read the second review result on the blockchain. If the second review result indicates that the review is passed, write the reconstructed clothing into the blockchain.

[0087] If the second review result indicates that the reconstructed clothing has passed the review, it means that the reconstructed clothing has practical value and is written into the blockchain for subsequent digital human image generation.

[0088] Correspondingly, if the second review result indicates that the review has failed, it means that the reconstructed clothing has no practical value and can be discontinued.

[0089] S206 reads clothing and accessories from the blockchain to create a clothing material library.

[0090] S207, input the clothing and accessories in the clothing material library into the pre-built clothing combination model, and obtain multiple clothing schemes output by the clothing combination model; wherein, the clothing combination model is obtained based on the training of a deep learning model.

[0091] By inputting clothing and accessories into the outfit combination model, outfit schemes can be automatically generated to meet the needs of digital human avatar generation.

[0092] Specifically, the construction of an outfit combination model first requires collecting outfit data. This data includes combinations of various clothing and accessories (such as clothes, shoes, and accessories) for different genders, ages, occasions, and seasons, with detailed annotations of the features of each item (such as style, color, and material) and the overall style. Next, a suitable model architecture is selected, such as a deep learning-based Generative Adversarial Network (GAN) or Graph Neural Network (GNN), and the model is trained using the labeled outfit data to learn the matching rules, style coordination, and overall aesthetics of clothing and accessories. During training, the loss function is continuously optimized to ensure that the outfit schemes generated by the model are as close as possible to real and high-quality outfit examples. Simultaneously, the model parameters are adjusted using a validation set to prevent overfitting. Finally, the model is evaluated using a test set to ensure that it can stably and accurately generate reasonable outfit schemes based on multiple input clothing and accessories.

[0093] S208: Input the clothing and accessories in the clothing material library into the pre-built label extraction model to obtain the feature extraction results of each clothing and accessory; wherein, the label extraction model is trained based on a deep learning model, and the feature extraction results include feature labels under multiple preset classification dimensions.

[0094] Among them, the label extraction model can obtain the feature labels of clothing and accessories, which can be used to combine clothing and accessories according to the feature labels to form a wearing scheme.

[0095] Specifically, the construction of the label extraction model first requires building a sample dataset covering various types of clothing and accessories, and accurately labeling each item with feature labels across multiple classification dimensions (such as material, color, style, design, and applicable occasions). Next, a suitable deep learning model architecture is selected, such as a convolutional neural network (CNN) model, to capture the visual features of the clothing and accessories, while natural language processing techniques are used to process the textual attribute information. By inputting wearing samples into the model, it learns the mapping relationship between clothing and accessories and feature labels across various dimensions. The loss function is calculated using the labeled data, and backpropagation is used to adjust the model parameters. During training, data augmentation techniques are employed to expand the diversity of the dataset and prevent overfitting. Simultaneously, a validation set is used to monitor model performance and continuously optimize hyperparameters. Finally, a test set is used to evaluate the model, ensuring that it can accurately and comprehensively extract feature labels for clothing and accessories across multiple classification dimensions.

[0096] S209: Based on preset matching rules, combine clothing and accessories in the outfit material library according to feature tags to obtain multiple outfit schemes.

[0097] This process involves creating new outfit schemes based on pre-set matching rules and the characteristic tags of clothing and accessories. First, a matching rule library needs to be built, covering multiple dimensions of matching rules such as color coordination (e.g., complementary colors, analogous colors), style unity (e.g., pairing retro styles with retro elements), and material compatibility (e.g., combining soft fabrics with crisp fabrics). Then, for each piece of clothing and accessory, including color, style, material, and design, the characteristic tags of the clothing and accessories are matched and analyzed against the matching rule library to filter out combinations of clothing and accessories that meet the matching rules. At the same time, a weighting mechanism can be introduced to assign corresponding weights according to the importance of different rules to optimize the combination results, and finally, an outfit scheme is formed.

[0098] S210, based on clothing and accessories in the outfit scheme and digital human, forms a digital human image.

[0099] Optionally, to ensure that the generated digital avatar meets user needs, the generated digital avatar can be reviewed:

[0100] Step 1: Send the digital human image to the reviewer based on a pre-configured smart contract.

[0101] Among these measures, the digital human image is reviewed through manual verification.

[0102] Step 2: Obtain the first review result from the reviewer and write the first review result into the blockchain.

[0103] Step 3: Read the first review result on the blockchain. If the first review result indicates that the review is passed, then write the digital human image into the blockchain.

[0104] If the first review result indicates that the digital human image has passed the review, it means that the digital human image meets the user's needs and is written into the blockchain.

[0105] Correspondingly, if the first review result indicates that the review has failed, it means that the digital human image does not meet the user's needs and can be left unsaved.

[0106] In this embodiment, by automatically generating new clothing and accessories and outfit combinations, the creation cycle of digital human images is shortened, the issuance efficiency of digital human images is improved, and labor costs are reduced.

[0107] Figure 3 This is a schematic diagram of the structure of a digital human image generation device provided in an embodiment of this application. (See attached diagram.) Figure 3 The digital human image generation device includes various functional modules for implementing the aforementioned digital human image generation method, and any functional module can be implemented by software and / or hardware.

[0108] In some embodiments, the digital human image generation device 300 includes a data acquisition module 301, a data distribution module 302, a data writing module 303, and an image generation module 304. Wherein:

[0109] The data acquisition module 301 is used to acquire parameter information of the pre-generated digital human;

[0110] The data distribution module 302 is used to distribute parameter information to multiple clothing designers based on a pre-configured smart contract, so that the clothing designers can design and create a number of clothing accessories according to the parameter information.

[0111] Data writing module 303 is used to receive clothing and accessories feedback from multiple clothing designers and write the clothing and accessories into the blockchain;

[0112] The image generation module 304 is used to generate multiple digital human images based on clothing, accessories, and digital humans.

[0113] In some embodiments, the image generation module 304 is specifically used for:

[0114] Read clothing and accessories from the blockchain to create a clothing and accessories resource library;

[0115] Combine and match clothing and accessories from the outfit resource library to create multiple outfit options;

[0116] A digital human image is created based on clothing, accessories, and the digital human in the outfit scheme.

[0117] In some embodiments, the image generation module 304 is further configured to:

[0118] The clothing and accessories in the clothing material library are input into a pre-built clothing combination model to obtain multiple clothing schemes output by the clothing combination model; the clothing combination model is obtained by training a deep learning model.

[0119] In some embodiments, the image generation module 304 is further configured to:

[0120] The clothing and accessories in the outfit material library are input into a pre-built label extraction model to obtain the feature extraction results of each clothing and accessory; the label extraction model is trained based on a deep learning model, and the feature extraction results include feature labels under multiple preset classification dimensions;

[0121] Based on preset matching rules, clothing and accessories in the outfit material library are combined using feature tags to obtain multiple outfit schemes.

[0122] In some embodiments, the device further includes a data verification module 305, which is specifically used for:

[0123] The digital human image is sent to the reviewer based on a pre-configured smart contract;

[0124] Obtain the first review result from the reviewer and write the first review result into the blockchain;

[0125] Read the first review result on the blockchain. If the first review result indicates that the review is passed, write the digital human image into the blockchain.

[0126] In some embodiments, the device further includes a garment recombination module 306, which is specifically used for:

[0127] Each piece of clothing and accessories read from the blockchain is input into a pre-built element decomposition model to obtain multiple clothing elements obtained from the decomposition of each piece of clothing and accessories; wherein, the element decomposition model is obtained by training a deep learning model;

[0128] Clothing elements are input into a pre-built clothing recombination model to obtain reconstructed clothing; the clothing recombination model is obtained by training a deep learning model.

[0129] In some embodiments, the data verification module 305 is further configured to:

[0130] The reconstructed clothing is sent to the reviewer based on a pre-configured smart contract;

[0131] Obtain the second review result from the reviewer and write the second review result into the blockchain;

[0132] Read the second audit result on the blockchain. If the second audit result indicates that the audit is passed, write the reconstructed clothing to the blockchain.

[0133] The digital human image generation device 300 provided in this application embodiment is used to execute the technical solution provided in the aforementioned digital human image generation method embodiment. Its implementation principle and technical effect are similar to those in the aforementioned method embodiment, and will not be repeated here.

[0134] It should be noted that the division of the various modules in the above device is merely a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, these modules can be implemented entirely in software via processing elements, entirely in hardware, or partially in software via processing elements and partially in hardware. For example, the data acquisition module 301 can be a separate processing element, or it can be integrated into a chip in the above device. Alternatively, it can be stored as program code in the memory of the above device, and its functions can be called and executed by a processing element. The implementation of other modules is similar. Moreover, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed through integrated logic circuits in the hardware of the processor element or through software instructions.

[0135] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. (See attached diagram.) Figure 4 The electronic device 400 includes a processor 401 and a memory 402 communicatively connected to the processor 401;

[0136] Memory 402 stores instructions executed by the computer;

[0137] The processor 401 executes the computer execution instructions stored in the memory 402 to implement the technical solution of the aforementioned digital human image generation method.

[0138] In the aforementioned electronic device 400, the memory 402 and the processor 401 are electrically connected directly or indirectly to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines, such as bus connections. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be classified as address buses, data buses, control buses, etc., but this does not mean that there is only one bus or one type of bus. The memory 402 stores computer execution instructions for implementing the aforementioned digital human image generation method, including at least one software functional module that can be stored in the memory 402 in the form of software or firmware. The processor 401 executes various functional applications and data processing by running the software programs and modules stored in the memory 402.

[0139] The memory 402 includes at least one type of readable storage medium, not limited to Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The memory 402 stores programs, which are executed by the processor 401 upon receiving execution instructions. Furthermore, the software programs and modules within the memory 402 may also include an operating system, which may include various software components and / or drivers for managing system tasks (e.g., memory management, storage device control, power management, etc.) and can communicate with various hardware or software components to provide an operating environment for other software components.

[0140] Processor 401 can be an integrated circuit chip with signal processing capabilities. The aforementioned processor 401 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), etc. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor, or processor 401 can be any conventional processor.

[0141] The electronic device 400 is used to execute the technical solution provided in the aforementioned embodiment of the digital human image generation method. Its implementation principle and technical effect are similar to those in the aforementioned method embodiment, and will not be repeated here.

[0142] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed, are used to implement the technical solution of the aforementioned digital human image generation method.

[0143] The aforementioned computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The computer-readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0144] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Alternatively, the readable storage medium can be an integral part of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components within the control device of the digital human image generation apparatus.

[0145] This application also provides a computer program product, including a computer program, which, when executed, is used to implement the technical solution of the aforementioned digital human image generation method.

[0146] In the above embodiments, those skilled in the art will understand that the above method embodiments can be implemented entirely or partially by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented entirely or partially in the form of a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless network, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

[0147] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0148] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the appended claims.

[0149] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for generating a digital human image, characterized in that, include: Obtain parameter information of a pre-generated digital human; Based on a pre-configured smart contract, the parameter information is distributed to multiple clothing designers, who then design and create various clothing accessories according to the parameter information. Receive clothing and accessories feedback from multiple clothing designers and write the clothing and accessories into the blockchain; Based on the clothing and accessories and the digital human, multiple digital human images are generated.

2. The method according to claim 1, characterized in that, Based on the clothing and accessories and the digital human, multiple digital human images are generated, including: Read clothing and accessories from the blockchain to create a clothing and accessories resource library; The clothing and accessories in the clothing material library are combined to create multiple outfit schemes; Based on the clothing and accessories in the aforementioned outfit scheme and the digital human, a digital human image is formed.

3. The method according to claim 2, characterized in that, The clothing and accessories in the clothing and accessories library are combined to create multiple outfit schemes, including: The clothing and accessories in the clothing material library are input into a pre-built clothing combination model to obtain multiple clothing schemes output by the clothing combination model; wherein, the clothing combination model is obtained based on a deep learning model training.

4. The method according to claim 2, characterized in that, The clothing and accessories in the clothing and accessories library are combined to create multiple outfit schemes, including: The clothing and accessories in the clothing material library are input into a pre-built tag extraction model to obtain the feature extraction results of each clothing and accessory; wherein, the tag extraction model is trained based on a deep learning model, and the feature extraction results include feature tags under multiple preset classification dimensions; Based on preset matching rules, clothing and accessories in the outfit material library are combined using feature tags to obtain multiple outfit schemes.

5. The method according to any one of claims 2-4, characterized in that, The method further includes: The digital human image is sent to the reviewer based on a pre-configured smart contract; Obtain the first review result from the reviewer and write the first review result into the blockchain; Read the first review result on the blockchain. If the first review result indicates that the review is passed, write the digital human image into the blockchain.

6. The method according to any one of claims 2-4, characterized in that, The method further includes: Each piece of clothing and accessories read from the blockchain is input into a pre-built element decomposition model to obtain multiple clothing elements obtained from the decomposition of each piece of clothing and accessories; wherein, the element decomposition model is obtained by training a deep learning model; The clothing elements are input into a pre-built clothing recombination model to obtain reconstructed clothing; wherein the clothing recombination model is obtained based on a deep learning model.

7. The method according to claim 6, characterized in that, The method further includes: The reconstructed clothing is sent to the reviewer based on a pre-configured smart contract; Obtain the second review result from the reviewer and write the second review result into the blockchain; Read the second audit result on the blockchain. If the second audit result indicates that the audit is passed, write the reconstructed clothing into the blockchain.

8. A digital human image generation device, characterized in that, include: The data acquisition module is used to acquire parameter information of the pre-generated digital human; The data distribution module is used to distribute the parameter information to multiple clothing designers based on a pre-configured smart contract, so that the clothing designers can design and create a number of clothing accessories based on the parameter information. The data writing module is used to receive clothing and accessories feedback from multiple clothing designers and write the clothing and accessories into the blockchain. The image generation module is used to generate multiple digital human images based on the clothing and accessories and the digital human.

9. An electronic device, characterized in that, Includes a processor and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed, are used to implement the method as described in any one of claims 1 to 7.

Images