A method, system and device for managing digital human identity.

Abstract

The application provides a digital human identity management method, system and device, relates to the technical field of digital human management, and comprises the following steps: obtaining the three-dimensional characteristic value of a digital human according to a three-dimensional model; obtaining the voice characteristic value of the digital human according to a basic voice segment; obtaining the appearance characteristic value of the digital human according to appearance data; performing hash operation on the three-dimensional characteristic value, the voice characteristic value and the appearance characteristic value of the digital human to obtain a hash value, and generating the identity of the digital human on a current platform according to the hash value; and when the digital human performs cross-platform operation from the current platform, updating the identity of the current platform according to the three-dimensional model, the basic voice segment and the appearance data of the digital human on a target platform to obtain the identity of the target platform. The identity is generated by comprehensively using the three-dimensional model, the basic voice segment and the appearance data of the digital human, and the association between the digital human and the identity and the management effect are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of digital human management technology, and more specifically, to a digital human identity management method, system, and device. Background Technology

[0002] With the rise of live-streaming e-commerce and virtual social networking, digital humans need to seamlessly migrate between different platforms. Therefore, precise and efficient management of the digital human's identity across different platforms is necessary to ensure the uniqueness of the digital human's identity on each platform.

[0003] In related technologies, in order to ensure the uniqueness of digital humans across different platforms, the method of randomly generated identity identifiers is usually relied upon to ensure that digital humans are unique. However, due to the randomness of the identity identifiers, a large amount of other data support is required when verifying across platforms. At the same time, due to the randomness, the correlation between the identity identifiers of digital humans and the digital humans themselves is weak, which increases the complexity of digital human management. Summary of the Invention

[0004] The problem addressed by this invention is how to improve the correlation and management effectiveness between digital humans and identity identifiers.

[0005] To address the above problems, this invention provides a digital human identity management method, system, and device.

[0006] In a first aspect, the digital human identity management method of the present invention includes:

[0007] Acquire the 3D model, basic voice clips, and appearance data of the digital human on the current platform;

[0008] Geometric features are extracted from the three-dimensional model to obtain the three-dimensional feature values ​​of the digital human.

[0009] Speech encoding is performed on the basic speech segments to obtain the speech feature values ​​of the digital human;

[0010] Based on the appearance data, an appearance analysis is performed to obtain the appearance feature values ​​of the digital human.

[0011] A hash value is obtained by performing a hash operation on the three-dimensional feature value, the voice feature value, and the appearance feature value of the digital human, and an identity identifier of the digital human on the current platform is generated based on the hash value.

[0012] When the digital human crosses platforms from the current platform, the identity identifier of the digital human on the target platform is updated based on the digital human's 3D model, basic voice segments, and appearance data on the target platform to obtain the identity identifier of the target platform.

[0013] Optionally, the step of extracting geometric features from the three-dimensional model to obtain the three-dimensional features of the digital human includes:

[0014] The three-dimensional model is analyzed to obtain the vertex coordinates of all vertices of the three-dimensional model;

[0015] The vertex coordinates are standardized and sorted, and then the sorted vertex coordinates are concatenated according to a preset format to obtain the coordinate string of the three-dimensional model.

[0016] Convert the coordinate string into a byte stream;

[0017] A hash operation is performed on the byte stream to generate the three-dimensional feature values ​​of the digital human.

[0018] Optionally, the step of performing speech encoding based on the basic speech segment to obtain the speech feature values ​​of the digital human includes:

[0019] The basic speech segment is denoised to obtain the effective speech signal in the basic speech segment.

[0020] The effective speech signal is extracted to obtain the MFCC coefficients of the basic speech segment, and the MFCC coefficients are normalized by mean-variance and time-series aggregation to obtain a static feature vector.

[0021] Principal component analysis (PCA) is used to reduce the dimensionality of the static feature vectors to obtain real number vectors.

[0022] The real number vector is uniformly quantized to obtain the binary code corresponding to the real number vector, and the binary code is used as the speech feature value of the digital human.

[0023] Optionally, the step of performing appearance analysis based on the appearance data to obtain the appearance feature values ​​of the digital human includes:

[0024] The appearance data is analyzed to obtain multiple appearance attributes of the data person;

[0025] Enumerate and encode all the appearance attributes to obtain the binary code corresponding to each appearance attribute;

[0026] The binary codes of all the appearance attributes are concatenated in a first preset order to obtain a binary string of all the appearance attributes, and the binary string is used as the appearance feature value of the digital human.

[0027] Optionally, the step of performing a hash operation on the three-dimensional feature values, the voice feature values, and the appearance feature values ​​of the digital human to obtain a hash value, and generating an identity identifier for the digital human on the current platform based on the hash value, includes:

[0028] The three-dimensional feature value, the voice feature value, and the appearance feature value are concatenated in a second preset order to obtain a feature string containing the three-dimensional feature value, the voice feature value, and the appearance feature value;

[0029] Perform a hash operation on the feature string to generate the hash value corresponding to the digital human;

[0030] The hash value is formatted using a preset format to obtain the identity identifier of the digital human on the current platform.

[0031] Optionally, the step of combining the hash value with a preset format to obtain the identity identifier of the digital human on the current platform includes:

[0032] Obtain the timestamp that formats the hash value and the current version number of the digital human;

[0033] Convert the timestamp into a timestamp string;

[0034] The hash value, the timestamp string, and the current version number are concatenated with preset characters to obtain a structured string, which is then used as the identity identifier of the digital human on the current platform.

[0035] Optionally, when the digital human crosses platforms from the current platform, the identity identifier of the digital human on the target platform is updated based on the digital human's 3D model, basic voice segments, and appearance data on the target platform to obtain the identity identifier of the target platform, including:

[0036] When the digital human crosses platforms from the current platform, the three-dimensional model, the basic voice segment, and the appearance data of the digital human on the target platform are acquired.

[0037] Based on the three-dimensional model of the target platform, the basic speech segment, and the appearance data, the geometric feature extraction, speech encoding, and appearance analysis are performed to obtain the target three-dimensional feature value, target speech feature value, and target appearance feature value corresponding to the target platform.

[0038] Based on the target's three-dimensional feature value, target's voice feature value, and target's appearance feature value, a hash operation is performed to obtain the target hash value of the digital human on the target platform;

[0039] The target hash value is compared with the current hash value in the identity identifier of the current platform, and the identity identifier of the current platform is updated according to the comparison result to obtain the identity identifier of the target platform.

[0040] Optionally, updating the identity identifier of the current platform based on the comparison result includes:

[0041] Based on the comparison results, it is determined whether the target hash value is consistent with the current hash value;

[0042] If the target hash value is consistent with the current hash value, then there is no need to update the identity identifier of the current platform, and the identity identifier of the current platform is used as the identity identifier of the target platform;

[0043] If the target hash value is inconsistent with the current hash value, the identity identifier of the current platform needs to be updated, the current hash value is replaced with the target hash value, and then the identity identifier of the target platform is generated based on the target hash value.

[0044] Secondly, the digital human identity management system of the present invention includes:

[0045] The data acquisition unit is used to acquire the digital human's 3D model, basic voice segments, and appearance data on the current platform;

[0046] A geometric feature extraction unit is used to extract geometric features based on the three-dimensional model to obtain the three-dimensional feature values ​​of the digital human.

[0047] A speech feature unit is used to perform speech encoding based on the basic speech segment to obtain the speech feature value of the digital human.

[0048] An appearance feature unit is used to perform appearance analysis based on the appearance data to obtain the appearance feature values ​​of the digital human.

[0049] The processing unit is used to perform hash operations on the three-dimensional feature values, voice feature values ​​and appearance feature values ​​of the digital human to obtain a hash value, and generate the identity identifier of the digital human on the current platform based on the hash value;

[0050] The cross-platform update unit is used to update the identity identifier of the current platform based on the three-dimensional model, basic voice segments and appearance data of the digital human on the target platform when the digital human crosses platforms from the current platform, so as to obtain the identity identifier of the target platform.

[0051] In a second aspect, the electronic device of the present invention includes a memory and a processor;

[0052] The memory is used to store computer programs;

[0053] The processor is configured to implement the digital human identity management method as described above when executing the computer program.

[0054] The digital human identity management method, system, and device of this invention significantly improve the correlation between digital humans and their identities and enhance management effectiveness by comprehensively utilizing the digital human's 3D model, basic voice segments, and appearance data to generate identity identifiers. First, by acquiring the digital human's 3D model, basic voice segments, and appearance data on the current platform, the core features of the digital human can be comprehensively captured. Through geometric feature extraction of the 3D model, speech encoding of the basic voice segments, and appearance analysis of the appearance data, a multi-level, multi-dimensional feature set is constructed, ensuring a close association between the identity identifier and the digital human itself, enabling each identity identifier to accurately reflect the uniqueness of the digital human. During the identity identifier generation process, the 3D feature values, voice feature values, and appearance feature values ​​are hashed to obtain a unique hash value, which is then used to generate the identity identifier. This hashing method based on multi-dimensional features not only guarantees the uniqueness of the identity identifier but also tightly binds the identity identifier to the core features of the digital human. Compared with traditional methods that generate identity identifiers based on random numbers or single features, the method of this invention effectively avoids identity identifier duplication and requires no additional database support during cross-platform verification, greatly improving the reliability and management efficiency of identity identifiers. When a digital human migrates across platforms, its identity on the current platform can be updated based on its 3D model, basic voice segments, and appearance data on the target platform, thus obtaining its identity on the target platform. This ensures the consistency and continuity of the digital human's identity across different platforms. Even if the digital human's characteristics change during the cross-platform process, the identity can be kept synchronized with the digital human's characteristics by recalculating the feature values ​​and updating the identity. This dynamic update mechanism not only improves the flexibility of the digital human's identity but also reduces the risk of identity invalidation or confusion due to feature changes. Furthermore, the identity management method of this invention has good scalability and compatibility. Since the identity is generated based on the digital human's core characteristics, identity verification across different platforms only requires recalculating the feature values ​​and performing hash operations to quickly verify the digital human's identity. This method is not only applicable to various existing digital human application scenarios but can also adapt to new platforms and scenarios. In this way, not only is the relevance and management effectiveness of digital human identity improved, but a solid technical support is also provided for the widespread application and future development of digital human technology. In summary, the digital human identity management method of this invention achieves close association and efficient management between digital humans and their identities by comprehensively utilizing multi-dimensional feature extraction, hash operations, and dynamic update mechanisms. This not only improves the uniqueness and reliability of identities but also enhances the consistency of digital human identities during cross-platform migration. It provides a more secure, efficient, and flexible identity management solution for the development and application of digital human technology. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating a digital human identity management method according to an embodiment of the present invention;

[0056] Figure 2 This is a flowchart illustrating the digital human identity management system in another embodiment of the present invention. Detailed Implementation

[0057] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0058] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0059] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0060] It should be noted that the terms "one" and "more" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0061] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0062] To address the problems existing in the aforementioned related technologies, this embodiment provides a digital human identity management method, system, and device.

[0063] Combination Figure 1As shown, the digital human identity management method provided in this embodiment of the invention includes:

[0064] Acquire the 3D model, basic voice clips, and appearance data of the digital human on the current platform.

[0065] Specifically, the digital human identity management method first requires acquiring the digital human's 3D model, basic voice clips, and appearance data on the current platform to ensure the accuracy and completeness of subsequent feature extraction and identity generation. Specifically, the 3D model is typically stored in formats such as .obj or .fbx, containing the digital human's geometric structure information; the basic voice clips are collected using recording equipment to ensure the integrity and clarity of the voice signal; and appearance data includes visual features such as the digital human's clothing, hairstyle, and makeup, which can be obtained through image recognition or manual annotation. Through these methods, the core characteristics of the digital human can be comprehensively captured, providing a data foundation for subsequent feature extraction and identity generation.

[0066] Geometric features are extracted from the three-dimensional model to obtain the three-dimensional feature values ​​of the digital human.

[0067] Specifically, since the geometric features of a 3D model are a crucial component of a digital human's identity, analyzing all vertex coordinates of the model and performing precision unification and sorting can eliminate coordinate precision differences caused by different software or export formats, ensuring the accuracy and consistency of feature extraction. Concatenating the sorted vertex coordinates into a string and converting it to a byte stream before performing hash operations generates 3D feature values ​​that are not only unique but also enhance data security and immutability through hashing. This effectively extracts and fixes the geometric features of the digital human, providing crucial geometric feature basis for subsequent identity generation.

[0068] Speech encoding is performed on the basic speech segments to obtain the speech feature values ​​of the digital human.

[0069] Specifically, since speech is one of the important features of digital human identity, this invention can obtain a static feature vector that stably reflects speech characteristics through denoising, extraction of MFCC coefficients, mean-variance standardization, and temporal aggregation. Furthermore, through principal component analysis (PCA) dimensionality reduction and uniform quantization, binary encoding is obtained as the speech feature value of the digital human. This not only ensures the accuracy and consistency of the speech feature value but also improves the data storage and transmission efficiency through quantization processing. It can effectively extract and fix the speech features of the digital human, providing important speech feature basis for subsequent identity generation.

[0070] Based on the appearance data, appearance analysis is performed to obtain the appearance feature values ​​of the digital human.

[0071] Specifically, appearance data can include visual features such as clothing, hairstyle, and makeup of the digital human, which are important components of the digital human's identity. By analyzing appearance data, multiple appearance attributes are extracted, and each attribute is transformed to obtain the digital human's appearance feature values. This not only ensures the accuracy and completeness of the appearance feature values ​​but also improves the storage and transmission efficiency of data through encoding methods. It can effectively extract and fix the digital human's appearance features, providing important appearance feature basis for subsequent identity generation.

[0072] A hash value is obtained by performing a hash operation on the digital human's three-dimensional feature value, voice feature value, and appearance feature value, and an identity identifier for the digital human on the current platform is generated based on the hash value.

[0073] Specifically, the extracted 3D feature values, voice feature values, and appearance feature values ​​are combined to generate a unique hash value, which is then used to generate the digital human's identity identifier on the current platform. By concatenating these feature values ​​into a feature string in a preset order and performing a hash operation on this feature string, a unique hash value can be generated, thus forming a structured identity identifier. This not only ensures the uniqueness and security of the identity identifier but also improves cross-platform compatibility through a standardized format. This process effectively integrates multi-dimensional features to generate a comprehensive, unique, and secure identity identifier, providing a reliable identification basis for the digital human's identity management on the current platform. This step is necessary because it integrates multi-dimensional features into a unique identifier, ensuring the accuracy and uniqueness of the identity identifier.

[0074] When the digital human crosses platforms from the current platform, the identity identifier of the digital human on the target platform is updated based on the digital human's 3D model, basic voice segments, and appearance data on the target platform to obtain the identity identifier of the target platform.

[0075] Specifically, when a digital human migrates from the current platform to a target platform, its 3D model, basic voice segments, and appearance data need to be reacquired, and the feature extraction and hashing steps need to be re-executed to generate feature values ​​and hash values ​​for the target platform. The target hash value is then compared with the hash value in the current platform's identity identifier to update the current platform's identity identifier. This improved dynamic update mechanism not only ensures the consistency of the digital human's identity identifier across different platforms but also enhances the flexibility and adaptability of the identity identifier. It effectively addresses potential feature changes that may occur during cross-platform migration, ensuring that the digital human's identity identifier remains consistent across different platforms and improving the reliability and management efficiency of the identity identifier.

[0076] The digital human identity management method, system, and device of this invention significantly improve the correlation between digital humans and their identities and enhance management effectiveness by comprehensively utilizing the digital human's 3D model, basic voice segments, and appearance data to generate identity identifiers. First, by acquiring the digital human's 3D model, basic voice segments, and appearance data on the current platform, the core features of the digital human can be comprehensively captured. Through geometric feature extraction of the 3D model, speech encoding of the basic voice segments, and appearance analysis of the appearance data, a multi-level, multi-dimensional feature set is constructed, ensuring a close association between the identity identifier and the digital human itself, enabling each identity identifier to accurately reflect the uniqueness of the digital human. During the identity identifier generation process, the 3D feature values, voice feature values, and appearance feature values ​​are hashed to obtain a unique hash value, which is then used to generate the identity identifier. This hashing method based on multi-dimensional features not only guarantees the uniqueness of the identity identifier but also tightly binds the identity identifier to the core features of the digital human. Compared with traditional methods that generate identity identifiers based on random numbers or single features, the method of this invention effectively avoids identity identifier duplication and requires no additional database support during cross-platform verification, greatly improving the reliability and management efficiency of identity identifiers. When a digital human migrates across platforms, its identity on the current platform can be updated based on its 3D model, basic voice segments, and appearance data on the target platform, thus obtaining its identity on the target platform. This ensures the consistency and continuity of the digital human's identity across different platforms. Even if the digital human's characteristics change during the cross-platform process, the identity can be kept synchronized with the digital human's characteristics by recalculating the feature values ​​and updating the identity. This dynamic update mechanism not only improves the flexibility of the digital human's identity but also reduces the risk of identity invalidation or confusion due to feature changes. Furthermore, the identity management method of this invention has good scalability and compatibility. Since the identity is generated based on the digital human's core characteristics, identity verification across different platforms only requires recalculating the feature values ​​and performing hash operations to quickly verify the digital human's identity. This method is not only applicable to various existing digital human application scenarios but can also adapt to new platforms and scenarios. In this way, not only is the relevance and management effectiveness of digital human identity improved, but a solid technical support is also provided for the widespread application and future development of digital human technology. In summary, the digital human identity management method of this invention achieves close association and efficient management between digital humans and their identities by comprehensively utilizing multi-dimensional feature extraction, hash operations, and dynamic update mechanisms. This not only improves the uniqueness and reliability of identities but also enhances the consistency of digital human identities during cross-platform migration. It provides a more secure, efficient, and flexible identity management solution for the development and application of digital human technology.

[0077] Optionally, the step of extracting geometric features from the three-dimensional model to obtain the three-dimensional features of the digital human includes:

[0078] The three-dimensional model is analyzed to obtain the vertex coordinates of all vertices of the three-dimensional model;

[0079] The vertex coordinates are standardized and sorted, and then the sorted vertex coordinates are concatenated according to a preset format to obtain the coordinate string of the three-dimensional model.

[0080] Convert the coordinate string into a byte stream;

[0081] A hash operation is performed on the byte stream to generate the three-dimensional feature values ​​of the digital human.

[0082] Specifically, in the digital human identity management method, geometric feature extraction is a crucial step in ensuring the uniqueness and accuracy of digital human identities. Since vertex coordinates are the core data of the 3D model, directly reflecting the geometric structure of the digital human, the 3D model is first analyzed to obtain the coordinate information of all vertices. Next, the vertex coordinates undergo precision unification and sorting to eliminate potential differences in coordinate precision caused by different software or export formats, ensuring that all vertex coordinates have a consistent format and precision. Sorting further ensures the ordered nature of the coordinate data. The sorted vertex coordinates are concatenated into a coordinate string according to a preset format. This coordinate string serves as input for a hash operation, which can use the MD5 hash algorithm. Due to the collision resistance of the MD5 algorithm, even if the vertex coordinates of two models differ only slightly (e.g., a difference of 0.000001 in the z-coordinate of a vertex), completely different hash values ​​will be generated, thus guaranteeing the uniqueness of the 3D model features. The byte stream described above is hashed using the standard MD5 algorithm to generate a 128-bit (16-byte) hash value, which is then converted into a 32-bit hexadecimal string (e.g., "d41d8cd98f00b204e9800998ecf8427e"). Converting the coordinate string to a byte stream is necessary to meet the input requirements of the hash algorithm, as hash algorithms typically process byte stream data. Finally, the byte stream is hashed again to generate the digital human's three-dimensional feature values. Hash operations not only generate a fixed-length unique identifier but also possess irreversible and collision-resistant properties, ensuring high security and uniqueness of the generated three-dimensional feature values. Through these steps, the geometric features of the digital human are accurately and securely extracted and represented, providing a reliable foundation for subsequent identity generation. This process can be implemented using built-in functions in programming languages ​​(such as Python's hashlib.md5()), eliminating the need for custom algorithms and ensuring compatibility.

[0083] In a preferred embodiment of the present invention, it is assumed that the digital human model contains 3 vertices, and the standardized coordinates are: (1.000000, 2.000000, 3.000000), (0.500000, 1.500000, 2.500000), (2.000000, 3.000000, 4.000000); the coordinates sorted by x in ascending order are (0.5..., 1.5..., 2.5...) → (1.0..., 2.0..., 3.0...) → (2.0... The concatenated string is: “0.500000,1.500000,2.500000;1.000000,2.000000,3.000000;2.000000,3.000000,4.000000”. MD5 hashing is then performed on this concatenated string using UTF-8 byte stream operations to obtain a 32-bit hexadecimal hash value (e.g., “a1b2c3d4e5f6a7b8c9d0e1f2a3b4c5d6”).

[0084] In this embodiment of the invention, the accuracy and consistency of the digital human's geometric features are ensured by accurately parsing the vertex coordinates of the 3D model and performing standardization processing. This standardization process effectively eliminates coordinate differences caused by different platforms or software, resulting in highly reliable and stable 3D feature values. Furthermore, the use of hash operations not only guarantees the uniqueness of feature values ​​but also enhances data security through its irreversibility and collision resistance.

[0085] Optionally, the step of performing speech encoding based on the basic speech segment to obtain the speech feature values ​​of the digital human includes:

[0086] The basic speech segment is denoised to obtain the effective speech signal in the basic speech segment.

[0087] The effective speech signal is extracted to obtain the MFCC coefficients of the basic speech segment, and the MFCC coefficients are normalized by mean-variance and time-series aggregation to obtain a static feature vector.

[0088] Principal component analysis (PCA) is used to reduce the dimensionality of the static feature vectors to obtain real number vectors.

[0089] The real number vector is uniformly quantized to obtain the binary code corresponding to the real number vector, and the binary code is used as the speech feature value of the digital human.

[0090] Specifically, firstly, the basic speech segments are denoised to remove background noise and other interference signals, thereby extracting a clean and effective speech signal, which improves the accuracy and stability of speech features. Next, MFCC coefficients are extracted from the effective speech signal. MFCC coefficients effectively reflect the spectral characteristics of speech and are widely used feature parameters in speech recognition and voiceprint authentication. To further improve the stability and consistency of the features, mean-variance standardization and temporal aggregation are performed on the MFCC coefficients to obtain a static feature vector. This eliminates the energy differences between different speech segments, making the feature vectors more comparable under different conditions. Subsequently, principal component analysis (PCA) is used to reduce the dimensionality of the static feature vector, resulting in a real-valued vector. Dimensionality reduction not only reduces the dimensionality of the data and improves computational efficiency but also retains the most important feature information. Finally, the real-valued vector is uniformly quantized to obtain the corresponding binary code, which is then used as the speech feature value of the digital human. Uniform quantization converts continuous real values ​​into discrete binary representations, improving data storage and transmission efficiency and making the feature values ​​more suitable for subsequent hash operations and identity generation. In a preferred embodiment of the invention, the speech segment is assumed to be a 10-second WAV file with a sampling rate of 16kHz and mono. First, background noise is removed through denoising processing to extract a clean and effective speech signal. Then, MFCC coefficients are extracted using Python's librosa library, with parameters set as follows: n_mfcc=13, n_fft=2048, hop_length=512. The extracted MFCC coefficients form a 13-dimensional feature matrix, where each dimension represents an MFCC coefficient, and each frame represents a feature at a given time point. Next, mean-variance standardization is performed on the MFCC coefficients to eliminate energy differences between different speech segments, making the feature vectors more comparable under different conditions. Then, temporal aggregation is performed on the standardized MFCC coefficients to obtain a 13-dimensional static feature vector. PCA is used to reduce the dimensionality of the 13-dimensional static feature vector to an 8-dimensional real vector, preserving the most important feature information. Finally, the 8-dimensional real vector is uniformly quantized to obtain a 32-bit binary code, for example: "10100011110001011110001101011001". This binary code is the digital human's voice feature value, which will be used in the subsequent identity generation process to ensure that the digital human's voice features are accurately reflected and verified in the identity.

[0091] In this embodiment of the invention, firstly, denoising and MFCC coefficient extraction effectively extract key information reflecting the speech characteristics of the digital human, improving the accuracy and stability of speech features. Secondly, mean-variance standardization and temporal aggregation further optimize the stability and consistency of feature vectors, making feature vectors of different speech segments more comparable under different conditions. PCA dimensionality reduction not only reduces the dimensionality of the data and improves computational efficiency but also retains the most important feature information, making the feature vectors more compact and efficient.

[0092] Optionally, the step of performing appearance analysis based on the appearance data to obtain the appearance feature values ​​of the digital human includes:

[0093] The appearance data is analyzed to obtain multiple appearance attributes of the data person;

[0094] Enumerate and encode all the appearance attributes to obtain the binary code corresponding to each appearance attribute;

[0095] The binary codes of all the appearance attributes are concatenated in a first preset order to obtain a binary string of all the appearance attributes, and the binary string is used as the appearance feature value of the digital human.

[0096] Specifically, firstly, the appearance data is analyzed to extract multiple appearance attributes of the digital human, such as hairstyle, hair color, skin tone, eyebrow shape, eye makeup, lip makeup, upper garment, and lower garment. These appearance attributes comprehensively reflect the visual appearance of the digital human. Next, each appearance attribute is enumerated and encoded, converting each attribute into a corresponding binary code. This encoding can be done manually or by pre-establishing a mapping table between appearance attributes and binary codes. The extracted appearance attributes are then compared with those in the mapping table to calculate similarity, determining the binary codes corresponding to appearance attributes with similarity greater than a threshold. Enumeration encoding is a method of converting categorical data into numerical data. In this way, complex appearance attributes are simplified into binary form, facilitating subsequent processing and storage. Finally, the binary codes of all appearance attributes are concatenated in a preset order to form a complete binary string. This binary string, as the appearance feature value of the digital human, comprehensively reflects the digital human's visual characteristics, ensuring the accuracy and consistency of appearance features and providing crucial appearance feature basis for subsequent identity generation. In a preferred embodiment of the present invention, it is assumed that the appearance data of the digital human includes the following attributes: long hair, black hair color, fair skin, willow-leaf eyebrows, light eye makeup, red lips, T-shirt top, and jeans bottom. First, these appearance attributes are enumerated and encoded, assuming the encoding rules are: Hairstyle: Long hair = 0001; Hair color: Black = 0000; Skin color: Fair = 0001; Eyebrow shape: Willow-leaf eyebrows = 0010; Eye makeup: Light makeup = 0011; Lip makeup: Red lips = 0100; Top: T-shirt = 0001; Bottom: Jeans = 0100. These binary codes are then concatenated in a preset order to form a complete binary string: “0001 0000 0001 0010 0011 0100 0001 0100”. This binary string represents the appearance feature value of the digital human. This information will be used in the subsequent identity generation process to ensure that the digital human's physical characteristics are accurately reflected and verified in the identity. In this way, the digital human's physical characteristics can be comprehensively and accurately represented in the identity, improving the accuracy and uniqueness of the identity.

[0097] In this embodiment of the invention, through comprehensive analysis of appearance data, multiple appearance attributes of the digital human can be extracted, which comprehensively reflect the visual characteristics of the digital human. Enumeration encoding simplifies complex appearance attributes into binary form, which not only improves data storage and transmission efficiency but also makes feature values ​​more suitable for subsequent processing and calculation. By concatenating the binary codes of all appearance attributes to form a complete binary string, the integrity and consistency of appearance feature values ​​can be ensured.

[0098] Optionally, the step of performing a hash operation on the three-dimensional feature values, the voice feature values, and the appearance feature values ​​of the digital human to obtain a hash value, and generating an identity identifier for the digital human on the current platform based on the hash value, includes:

[0099] The three-dimensional feature value, the voice feature value, and the appearance feature value are concatenated in a second preset order to obtain a feature string containing the three-dimensional feature value, the voice feature value, and the appearance feature value;

[0100] Perform a hash operation on the feature string to generate the hash value corresponding to the digital human;

[0101] The hash value is formatted using a preset format to obtain the identity identifier of the digital human on the current platform.

[0102] Specifically, first, the three feature values ​​are concatenated in a preset order to form a feature string containing all feature information, ensuring the consistency of the feature value order and structure, and providing standardized input for subsequent hash operations. Next, a hash operation is performed on the feature string to generate a unique hash value. Hash operations not only convert input data of different lengths into fixed-length outputs, but also possess irreversibility and collision resistance, making the generated hash value highly secure and unique. Finally, the hash value is formatted using a preset format to generate the digital human's identity identifier on the current platform. The preset format typically includes a timestamp and version number. The timestamp and version number not only increase the uniqueness of the identity identifier but also provide the generation time and version information, facilitating subsequent management and verification, ensuring that the generated identity identifier comprehensively and accurately reflects the multi-dimensional characteristics of the digital human, while possessing high security and uniqueness. The 3D feature value, speech feature value, and appearance feature value are concatenated in the order of "3D model hash (128 bits) + speech code (32 bits) + appearance code (32 bits)" to form a 192-bit feature string. This string is then input into the SHA-256 algorithm to generate a 256-bit hash value. The formula is as follows:

[0103] Hash value = SHA-256 (3D model hash || voice encoding || appearance encoding); where || represents string concatenation.

[0104] In a preferred embodiment of the present invention, it is assumed that the three-dimensional feature values ​​of the digital human are:

[0105] The given value is “5e884898da28047151d0e56f8dc6292773603d0d6aabbdd62a11ef721d1542d8”, with a voice feature value of “10100011110001011110001101011001” and an appearance feature value of “00010000000100100011010000010100”.

[0106] First, these three feature values ​​are concatenated in a preset order to form a feature string:

[0107] "5e884898da28047151d0e56f8dc6292773603d0d6aabbdd62a11ef721d1542d8 1010001111000101111000110101100100010000000100100011010000010100”.

[0108] Next, a hash operation is performed on this feature string to generate a unique hash value. Assuming the SHA-256 hash algorithm is used, the generated hash value is: "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef".

[0109] Finally, the hash value is formatted according to a preset format. Assume the current time is June 24, 2024, the Unix timestamp is 1719235200, and the version number is 01. This information is then concatenated into a structured identity identifier:

[0110] "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef-1719235200-01".

[0111] This structured identity identifier serves as the digital human's identity on the current platform. It not only comprehensively reflects the multi-dimensional characteristics of the digital human but also possesses a high degree of uniqueness and security. This approach ensures the accuracy and consistency of the digital human's identity across different platforms, providing a reliable basis for the management and verification of digital human identities.

[0112] In this embodiment of the invention, by concatenating three-dimensional feature values, voice feature values, and appearance feature values ​​to form a comprehensive feature string, a multi-dimensional feature string can be formed, which can comprehensively reflect the multi-dimensional characteristics of the digital human, ensuring that the generated identity identifier has a high degree of relevance and accuracy. Secondly, the use of hash operations not only guarantees the uniqueness of the feature string but also improves data security through its irreversibility and collision resistance properties.

[0113] Optionally, the step of combining the hash value with a preset format to obtain the identity identifier of the digital human on the current platform includes:

[0114] Obtain the timestamp that formats the hash value and the current version number of the digital human;

[0115] Convert the timestamp into a timestamp string;

[0116] The hash value, the timestamp string, and the current version number are concatenated with preset characters to obtain a structured string, which is then used as the identity identifier of the digital human on the current platform.

[0117] Specifically, the process involves obtaining the timestamp that formats the hash value and the current version number of the digital human. The timestamp records the specific time the identity was generated, while the version number reflects the current state of the digital human's characteristics. Next, the timestamp is converted into a timestamp string for concatenation with the hash value. Finally, the hash value, timestamp string, and version number are concatenated using preset characters to form a structured string. This structured string not only contains the digital human's multi-dimensional characteristic information but also provides generation time and version information, ensuring that the identity has high uniqueness and traceability across different platforms and time points. This guarantees that the generated identity not only comprehensively reflects the digital human's characteristics but also possesses a high degree of standardization and security.

[0118] In a preferred embodiment of the present invention, the hash value of the digital human is "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef", the current time is June 24, 2024, the Unix timestamp is 1719235200, and the version number is 01. First, the timestamp and version number are obtained, and the timestamp is converted into the string format 1719235200. Then, the hash value, the timestamp string, and the version number are concatenated with a preset character (such as "-") to form a structured identity identifier.

[0119] "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef-1719235200-01".

[0120] The aforementioned identity identifier is the identity identifier of the digital human on the current platform. It can comprehensively reflect the multi-dimensional characteristics of the digital human and also provides generation time and version information, making the management and verification of the identity identifier more convenient and secure on different platforms.

[0121] In this embodiment of the invention, by acquiring a timestamp and version number, the identity identifier can record the specific time of its generation and the current state of the digital human's characteristics, providing important timeline and version information for the management and verification of the identity identifier. Secondly, converting the timestamp into a string and concatenating it with other information to form a structured identity identifier not only improves the standardization of the identity identifier but also ensures its compatibility and consistency across different platforms.

[0122] Optionally, when the digital human crosses platforms from the current platform, the identity identifier of the digital human on the target platform is updated based on the digital human's 3D model, basic voice segments, and appearance data on the target platform to obtain the identity identifier of the target platform, including:

[0123] When the digital human crosses platforms from the current platform, the three-dimensional model, the basic voice segment, and the appearance data of the digital human on the target platform are acquired.

[0124] Based on the three-dimensional model of the target platform, the basic speech segment, and the appearance data, the geometric feature extraction, speech encoding, and appearance analysis are performed to obtain the target three-dimensional feature value, target speech feature value, and target appearance feature value corresponding to the target platform.

[0125] Based on the target's three-dimensional feature value, target's voice feature value, and target's appearance feature value, a hash operation is performed to obtain the target hash value of the digital human on the target platform;

[0126] The target hash value is compared with the current hash value in the identity identifier of the current platform, and the identity identifier of the current platform is updated according to the comparison result to obtain the identity identifier of the target platform.

[0127] Specifically, when a digital human migrates from the current platform to a target platform, the process first involves acquiring the digital human's 3D model, basic speech segments, and appearance data on the target platform. Next, based on the target platform's data, geometric feature extraction, speech encoding, and appearance analysis are performed to obtain the target platform's corresponding 3D feature values, speech feature values, and appearance feature values. This ensures the accuracy and consistency of the feature values, maintaining the integrity of the digital human's features even across different platforms. Then, hash operations are performed on these target feature values ​​to obtain the target platform's hash value. Finally, the target hash value is compared with the hash value in the current platform's identity identifier, and the current platform's identity identifier is updated based on the comparison result to obtain the target platform's identity identifier.

[0128] In a preferred embodiment of the present invention, the digital human migrates from the current platform to the target platform. On the current platform, the digital human's identity is identified as follows:

[0129] The string "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef-1719235200-01" contains the hash value:

[0130] The string "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef" contains a timestamp of 1719235200 and a version number of 01. When the digital human migrates to the target platform, the first step is to acquire the 3D model, basic speech fragments, and appearance data from the target platform. Assume the feature values ​​on the target platform are as follows:

[0131] The three-dimensional feature value is: “5e884898da28047151d0e56f8dc6292773603d0d6aabbdd62a11ef721d1542d8”;

[0132] Speech feature values ​​are:

[0133] “10100011110001011110001101011001”;

[0134] The appearance feature values ​​are:

[0135] “00010000000100100011010000010100”;

[0136] Based on these characteristic values, a hash operation is performed to generate the hash value for the target platform:

[0137] The hash value is "4f5a9b8cdef123456789abcdef123456789abcdef123456789abcdef123456789abcdef". The target hash value is compared with the hash value in the current platform's identity identifier. If they match, the current identity identifier is directly used as the target platform's identity identifier. If they do not match, a new version number 02 is generated by incrementing the version number in the current identity identifier, and the structured identity identifier is reconstructed using the current Unix timestamp 1719235200.

[0138] "4f5a9b8cdef123456789abcdef123456789abcdef123456789abcdef123456789abcdef-1719235200-02";

[0139] The structured identity identifiers mentioned above are the identity identifiers of digital humans on the target platform.

[0140] In this embodiment of the invention, by reacquiring the 3D model, basic voice fragments, and appearance data of the digital human on the target platform, and re-performing feature extraction and hashing operations, the consistency and continuity of the digital human's identity across different platforms are ensured. Secondly, comparing the target hash value with the hash value in the current platform's identity identifier allows for rapid verification of the digital human's identity consistency, improving the efficiency and accuracy of identity verification. Finally, updating the current platform's identity identifier based on the comparison results generates the target platform's identity identifier, which not only improves the flexibility of the identity identifier but also reduces the risk of identity identifier invalidation or confusion due to feature changes.

[0141] Optionally, updating the identity identifier of the current platform based on the comparison result includes:

[0142] Based on the comparison results, it is determined whether the target hash value is consistent with the current hash value;

[0143] If the target hash value is consistent with the current hash value, then there is no need to update the identity identifier of the current platform, and the identity identifier of the current platform is used as the identity identifier of the target platform;

[0144] If the target hash value is inconsistent with the current hash value, the identity identifier of the current platform needs to be updated, the current hash value is replaced with the target hash value, and then the identity identifier of the target platform is generated based on the target hash value.

[0145] Specifically, firstly, the characteristics of the digital human on the target platform are compared with the current hash value to determine if they are consistent with those on the current platform. Improving the comparison allows for quick and accurate identification of changes in the digital human's characteristics, thus ensuring the accuracy and reliability of the identity identifier. If the target hash value matches the current hash value, it means the digital human's characteristics have not changed during the cross-platform migration; therefore, there is no need to update the identity identifier on the current platform, and the current identity identifier can be directly used as the identity identifier on the target platform. This mechanism not only improves the efficiency of identity identifier management but also reduces unnecessary computational and storage overhead. If the target hash value does not match the current hash value, it indicates that the digital human's characteristics have changed during the cross-platform migration, and the identity identifier on the current platform needs to be updated. The update process includes replacing the current hash value with the target hash value and generating the identity identifier for the target platform based on the target hash value. This update mechanism not only ensures the dynamism and adaptability of the identity identifier but also provides a complete historical record of the identity identifier through version number and timestamp updates, facilitating subsequent management and verification.

[0146] In a preferred embodiment of the present invention, on the current platform, the identity identifier of the digital human is:

[0147] "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef-1719235200-01";

[0148] The hash value is:

[0149] The string "3f4a8b9cdef0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef" has a timestamp of "1719235200" and a version number of 01.

[0150] When the digital human migrates to the target platform, the system first acquires the 3D model, basic speech fragments, and appearance data from the target platform, and then re-executes feature extraction and hashing operations to generate a hash value for the target platform.

[0151] "4f5a9b8cdef123456789abcdef123456789abcdef123456789abcdef123456789abcdef";

[0152] Next, the target hash value is compared with the current hash value. If the target hash value matches the current hash value, there is no need to update the current platform's identity identifier; the current identity identifier can be directly used as the target platform's identity identifier. If the target hash value does not match the current hash value, the current platform's identity identifier needs to be updated. Specifically, first, the current hash value is replaced with the target hash value, the version number is updated, incrementing from 01 to 02, and the current timestamp is obtained, assumed to be 1719235200. The generated identity identifier for the target platform is:

[0153] "4f5a9b8cdef123456789abcdef123456789abcdef123456789abcdef123456789abcdef-1719235200-02".

[0154] In this embodiment of the invention, by comparing the target hash value with the current hash value, changes in digital human characteristics can be quickly and accurately identified, improving the accuracy and reliability of identity identification. If the hash values ​​match, the current identity is directly used as the identity identification for the target platform, reducing unnecessary computation and storage overhead and improving management efficiency. If the hash values ​​do not match, updating the hash value and generating a new identity ensures the dynamism and adaptability of the identity, effectively addressing potential characteristic changes in digital humans across platforms. Updates to the version number and timestamp provide a complete historical record of the identity, facilitating subsequent management and verification.

[0155] Combination Figure 2 As shown, the digital human identity management system provided in this embodiment of the invention includes:

[0156] The data acquisition unit is used to acquire the digital human's 3D model, basic voice segments, and appearance data on the current platform;

[0157] A geometric feature extraction unit is used to extract geometric features based on the three-dimensional model to obtain the three-dimensional feature values ​​of the digital human.

[0158] A speech feature unit is used to perform speech encoding based on the basic speech segment to obtain the speech feature value of the digital human.

[0159] An appearance feature unit is used to perform appearance analysis based on the appearance data to obtain the appearance feature values ​​of the digital human.

[0160] The processing unit is used to perform hash operations on the three-dimensional feature values, voice feature values ​​and appearance feature values ​​of the digital human to obtain a hash value, and generate the identity identifier of the digital human on the current platform based on the hash value;

[0161] The cross-platform update unit is used to update the identity identifier of the current platform based on the three-dimensional model, basic voice segments and appearance data of the digital human on the target platform when the digital human crosses platforms from the current platform, so as to obtain the identity identifier of the target platform.

[0162] The advantages of the digital human identity management system of the present invention compared with the prior art are the same as the advantages of the digital human identity management method compared with the prior art, and will not be repeated here.

[0163] The electronic device provided in this embodiment of the invention includes a memory and a processor;

[0164] The memory is used to store computer programs;

[0165] The processor is configured to implement the agent registration and service invocation method described above when executing the computer program.

[0166] The electronic device of the present invention has the same advantages over the prior art as the aforementioned digital human identity management method over the prior art, and will not be repeated here.

[0167] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A digital human identity management method, characterized by, The method comprises the following steps: acquiring a three-dimensional model, a basic voice segment and appearance data of a digital person on a current platform; extracting geometric features from the three-dimensional model to obtain three-dimensional feature values of the digital person; encoding the voice according to the basic voice segment to obtain voice feature values of the digital person; analyzing the appearance according to the appearance data to obtain appearance feature values of the digital person; performing hash operation on the three-dimensional feature values, the voice feature values and the appearance feature values of the digital person to obtain a hash value, and generating an identity of the digital person on the current platform according to the hash value; when the digital person performs cross-platform from the current platform, updating the identity of the current platform according to a three-dimensional model, a basic voice segment and appearance data of the digital person on a target platform to obtain an identity of the target platform.

2. The digital human identity management method of claim 1, wherein, The method of extracting geometric features from the three-dimensional model to obtain three-dimensional features of the digital person comprises the following steps: analyzing the three-dimensional model to obtain vertex coordinates of all vertices of the three-dimensional model; performing precision unification and sorting processing on the vertex coordinates, and then splicing the sorted vertex coordinates in a preset format to obtain a coordinate string of the three-dimensional model; converting the coordinate string into a byte stream; performing hash operation on the byte stream to generate the three-dimensional feature values of the digital person.

3. The digital human identity management method of claim 1, wherein, The method of encoding the voice according to the basic voice segment to obtain the voice feature values of the digital person comprises the following steps: performing noise reduction processing on the basic voice segment to obtain an effective voice signal in the basic voice segment; extracting MFCC coefficients of the basic voice segment according to the effective voice signal, and performing mean-variance standardization and time sequence aggregation on the MFCC coefficients to obtain a static feature vector; performing dimension reduction processing on the static feature vector by principal component analysis (PCA) to obtain a real number vector; uniformly quantizing the real number vector to obtain a binary code corresponding to the real number vector, and taking the binary code as the voice feature values of the digital person.

4. The digital human identity management method of claim 1, wherein, The method of analyzing the appearance according to the appearance data to obtain the appearance feature values of the digital person comprises the following steps: analyzing the appearance data to obtain a plurality of appearance attributes of the digital person; enumerating and encoding all the appearance attributes to obtain a binary code corresponding to each appearance attribute; splicing the binary codes of all the appearance attributes in a first preset order to obtain a binary string of all the appearance attributes, and taking the binary string as the appearance feature values of the digital person.

5. The digital human identity management method of claim 1, wherein, The method of performing hash operation on the three-dimensional feature values, the voice feature values and the appearance feature values of the digital person to obtain a hash value, and generating an identity of the digital person on the current platform according to the hash value comprises the following steps: splicing the three-dimensional feature values, the voice feature values and the appearance feature values in a second preset order to obtain a feature string containing the three-dimensional feature values, the voice feature values and the appearance feature values; hashing the feature string to generate the hash value corresponding to the digital human; performing format identification on the hash value in combination with a preset format to obtain the identity of the digital human on the current platform.

6. The digital human identity management method of claim 5, wherein, The combination of the preset format for the hash value format identification, obtaining the identity of the digital human on the current platform, includes: obtaining the time stamp for format identification of the hash value and the current version number of the digital human; convert the timestamp to a timestamp string; concatenate the hash value, the timestamp string and the current version number with a preset character to obtain a structured string, and take the structured string as the identity of the digital human on the current platform.

7. The digital human identity management method of claim 5, wherein, When the digital human crosses the platform from the current platform, the identity of the current platform is updated according to the three-dimensional model, the basic voice segment and the appearance data of the digital human on the target platform, and the identity of the target platform is obtained, including: When the digital human crosses the platform from the current platform, the three-dimensional model, the basic voice segment and the appearance data of the digital human on the target platform are obtained; According to the three-dimensional model, the basic voice segment and the appearance data of the target platform, the geometric feature extraction, voice coding and appearance analysis are performed to obtain the target three-dimensional feature value, the target voice feature value and the target appearance feature value corresponding to the target platform; According to the target three-dimensional feature value, the target voice feature value and the target appearance feature value, the hash value of the digital human on the target platform is obtained by hash operation; Compare the target hash value with the current hash value in the identity of the current platform, and update the identity of the current platform according to the comparison result to obtain the identity of the target platform.

8. The digital human identity management method of claim 7, wherein, The update of the identity of the current platform according to the comparison result includes: According to the comparison result, it is judged whether the target hash value is consistent with the current hash value; If the target hash value is consistent with the current hash value, the identity of the current platform is not updated, and the identity of the current platform is taken as the identity of the target platform; If the target hash value is not consistent with the current hash value, the identity of the current platform needs to be updated, and the current hash value is replaced by the target hash value, and then the identity of the target platform is generated according to the target hash value.

9. A digital human identity management system, characterized by, including: data acquisition unit, for obtaining three-dimensional model, basic voice segment and appearance data of digital human on current platform; geometric feature extraction unit, for extracting geometric features according to the three-dimensional model to obtain three-dimensional feature values of the digital human; speech feature unit, for performing speech coding according to the basic voice segment to obtain the speech feature value of the digital human; appearance feature unit, for performing appearance analysis according to the appearance data to obtain the appearance feature value of the digital human; An operation unit is configured to perform a hash operation on the three-dimensional feature value, the voice feature value, and the appearance feature value of the digital human to obtain a hash value, and generate an identity of the digital human on the current platform according to the hash value; A cross-platform updating unit is configured to, when the digital human performs cross-platform from the current platform, update the identity of the current platform according to a three-dimensional model, a basic voice segment, and appearance data of the digital human on a target platform to obtain an identity of the target platform.

10. An electronic device, comprising: comprising a memory and a processor; The memory is configured to store a computer program; The processor is configured to, when executing the computer program, implement the digital human identity management method according to any one of claims 1-8.

Images