Human setting, image, sound and life trajectory integrated construction method and system of digital human

Abstract

The invention discloses an integrated construction method and system for human settings, images, sounds and life tracks of a digital human, and belongs to the technical field of the Internet. The method comprises the following steps: generating a virtual digital human; constructing a life track, an image and voice of the virtual digital human; and performing behavior analysis on the virtual digital person, and updating the life track, image and voice of the virtual digital person. According to the invention, the defects of too long time and low efficiency of traditional digital human creation are overcome; the method solves the problems that the personality of the digital person in the traditional digital person creation depends on a model too singly, so that the characteristics of the digital person are too single, and the digital person cannot automatically learn and grow.

Description

Technical Field

[0001] The present invention relates to the field of Internet technology, and in particular to a method and system for integrating the personality, image, voice and life trajectory of a digital human. Background Art

[0002] With the advancement of science and technology, virtual reality technology has developed significantly, and the application of simulated digital humans in simulated work scenes has become increasingly widespread. The creation process of digital humans mainly involves the generation of digital human feature values, character voices, character portraits, skills, major deeds, and other related information descriptions, so that digital humans in actual work scenes are closer to real people. Therefore, the feature values ​​of digital humans often need to be truly close to the real person image. The current digital human generation scheme mainly relies on data collection, involving the collection of a large amount of information, data preprocessing and model training, and ultimately shaping the characteristics of digital humans through data reasoning.

[0003] The process of creating digital humans based on model generation is not only complex, but also time-consuming and resource-consuming. Limited by the creation method, the generated digital humans may lack uniqueness and have problems with repeated features. In addition, due to the lack of rich and realistic features, the appearance and behavior of digital humans may lack the necessary diversity and realism. More importantly, there are challenges for digital humans to adjust and grow themselves based on actual operations and experience. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for efficiently constructing the personality, image, voice and life trajectory of a digital human.

[0005] In order to solve the above technical problems, the present invention provides a method for integrating the personality, image, voice and life trajectory of a digital human, comprising the following steps:

[0006] Generate virtual digital people;

[0007] Construct the life trajectory, image and voice of virtual digital people;

[0008] Conduct behavioral analysis on virtual digital humans and update their life trajectory, image, and voice.

[0009] Preferably, generating a virtual digital human specifically includes the following steps:

[0010] Get prompt words or preset options;

[0011] According to the prompt words or preset options, the characteristics of the virtual digital human are constructed to generate the virtual digital human.

[0012] Preferably, the characteristics of the virtual digital person include name, gender, age, nationality, education background, personality, hobbies, skills, company affiliation, emotional state, religious beliefs and educational background.

[0013] Preferably, the method further comprises the following steps:

[0014] Conduct similarity analysis on the core features of virtual digital humans to ensure that the feature combination of each digital human is unique.

[0015] Preferably, constructing the life trajectory, image and voice of a virtual digital person specifically includes the following steps:

[0016] Generate important time nodes and important events as life trajectories based on the gender, age, personality, education, hobbies, skills, emotional state and educational background of the virtual digital person;

[0017] Generate an image of a virtual digital person based on the virtual digital person's gender, age, hobbies, education, emotional state, family background, and image prompt words;

[0018] The timbre is generated according to the fixed timbre prompt words or the gender and personality of the virtual digital person, and then the text information is generated according to the hobbies or important events of the life trajectory of the virtual digital person, and the voice is generated according to the timbre and text information.

[0019] Preferably, the method further comprises the following steps:

[0020] Based on the image and voice of the virtual digital person, important pictures and important voices of important time nodes and important events are generated.

[0021] Preferably, the behavior analysis of the virtual digital human and updating of the life trajectory, image and voice of the virtual digital human specifically include the following steps:

[0022] In a real work scenario, set anchor points;

[0023] After the anchor is triggered, the event subject is recorded;

[0024] Generate event theme voice or event theme image according to the event theme;

[0025] Update the characteristics of the virtual digital person according to the event theme to obtain the updated characteristics of the virtual digital person;

[0026] Based on the updated characteristics of the virtual digital person, the life trajectory, image and voice of the virtual digital person are reconstructed.

[0027] Preferably, the method further comprises the following steps:

[0028] Compare the differences of all anchor points to obtain the differences;

[0029] The anchor point with a difference less than the minimum difference threshold is taken as the first anchor point, and the newly set anchor points similar to the first anchor point are reduced;

[0030] The anchor point whose difference is greater than the maximum difference threshold is taken as the second anchor point, and a new anchor point similar to the second anchor point is added.

[0031] The present invention also provides a system for building a digital human's personality, image, voice and life trajectory in an integrated manner, including:

[0032] A character generation module is used to generate a virtual digital person;

[0033] Life trajectory, image and voice construction module, used to construct the life trajectory, image and voice of virtual digital people;

[0034] The behavior analysis module is used to analyze the behavior of the virtual digital human and update the life trajectory, image and voice of the virtual digital human.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The character generation module of the present invention creates certain fixed features of the digital human, the life trajectory module generates important events at certain time nodes based on the existing features of the digital human, the voice generation module can generate specific timbre and text information based on certain features, the image generation module generates a picture image that best suits the current digital human based on the feature values, the behavior analysis module records some landmark events after the successful creation of the digital human, and the data processing module continuously optimizes or generates feature values ​​by coordinating the functions of each module, thereby enhancing the realism of the digital human.

[0037] The advantages include:

[0038] 1. It solves the problem of traditional digital human creation that it takes too long and is inefficient. Traditional digital human creation often requires a large amount of data and models for training. It is better to directly use feature value prompts to create, and support participation in the setting of character settings in the generation stage, increasing the sense of participation and interaction. Then, the character features, life trajectory, voice, and image are generated through automatic processes, greatly improving the efficiency of creating digital humans.

[0039] 2. It solves the problem that the digital human personality in traditional digital human creation is too dependent on the model, resulting in the digital human's characteristics being too single and not being able to learn and grow automatically. Through real-time interaction and digital human behavior analysis, the present invention can continuously and dynamically improve the characteristic value according to the interaction between the digital human and the user or the development of events, and adjust the behavior and reaction in real time through the characteristic value, increase the depth and breadth of the digital human's characteristic value, and provide a more real and natural interactive experience.

[0040] 3. The design of the data processing module solves the problems of resource allocation and task calling in the case of multi-module calling, which is convenient for customization according to the application scenarios and needs of Tongtong, while reducing some resource consumption and improving the overall efficiency of the straight tube. At the same time, the complex redundancy of authentication, encryption, request processing, etc. is unified in design and maintenance, which is easy to maintain and upgrade, reducing maintenance costs and technical difficulties. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings.

[0042] Figure 1 It is a flow chart of a method for constructing the personality, image, voice and life trajectory of a digital human in an integrated manner according to the present invention;

[0043] Figure 2 It is a module schematic diagram of a system for constructing the personality, image, voice and life trajectory of a digital human in an integrated manner according to the present invention. DETAILED DESCRIPTION

[0044] Many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present invention, so the present invention is not limited to the specific implementation disclosed below.

[0045] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "said" and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0046] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of this specification, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0047] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0048] The present invention provides a method for integrating the personality, image, voice and life trajectory of a digital human, comprising the following steps:

[0049] Generate virtual digital people;

[0050] Construct the life trajectory, image and voice of virtual digital people;

[0051] Conduct behavioral analysis on virtual digital humans and update their life trajectory, image, and voice.

[0052] Preferably, constructing the life trajectory, image and voice of a virtual digital person;

[0053] Generate important time nodes and important events as life trajectories based on the gender, age, personality, education, hobbies, skills, emotional state and educational background of the virtual digital person;

[0054] Generate an image of a virtual digital person based on the virtual digital person's gender, age, hobbies, education, emotional state, family background, and image prompt words;

[0055] The timbre is generated according to the fixed timbre prompt words or the gender and personality of the virtual digital person, and then the text information is generated according to the hobbies or important events of the life trajectory of the virtual digital person, and the voice is generated according to the timbre and text information.

[0056] Preferably, the method further comprises the following steps:

[0057] Based on the image and voice of the virtual digital person, important pictures and important voices of important time nodes and important events are generated.

[0058] Preferably, generating a virtual digital human specifically includes the following steps:

[0059] Get prompt words or preset options;

[0060] Construct the characteristics of the virtual digital person based on prompt words or preset options.

[0061] Preferably, the characteristics of the virtual digital person include name, gender, age, nationality, education, personality, hobbies, skills, company affiliation, emotional state, religious beliefs and educational background.

[0062] Preferably, the method further comprises the following steps:

[0063] Conduct similarity analysis on the core features of virtual digital humans to ensure that the feature combination of each digital human is unique.

[0064] Preferably, the behavior analysis of the virtual digital human and updating of the life trajectory, image and voice of the virtual digital human specifically include the following steps:

[0065] In a real work scenario, set anchor points;

[0066] After the anchor is triggered, the event subject is recorded;

[0067] Generate event theme voice or event theme image according to the event theme;

[0068] Update the characteristics of the virtual digital person according to the event theme to obtain the updated characteristics of the virtual digital person;

[0069] Based on the updated characteristics of the virtual digital person, the life trajectory, image and voice of the virtual digital person are reconstructed.

[0070] Preferably, the method further comprises the following steps:

[0071] Compare the differences of all anchor points to obtain the differences;

[0072] The anchor point with a difference less than the minimum difference threshold is taken as the first anchor point, and the newly set anchor points similar to the first anchor point are reduced;

[0073] The anchor point whose difference is greater than the maximum difference threshold is taken as the second anchor point, and a new anchor point similar to the second anchor point is added.

[0074] The present invention also provides a system for building a digital human's personality, image, voice and life trajectory in an integrated manner, including:

[0075] A character generation module is used to generate a virtual digital person;

[0076] Life trajectory, image and voice construction module, used to construct the life trajectory, image and voice of virtual digital people;

[0077] The behavior analysis module is used to analyze the behavior of the virtual digital human and update the life trajectory, image and voice of the virtual digital human.

[0078] In order to better illustrate the technical effect of the present invention, the present invention provides the following specific embodiments to illustrate the above technical process:

[0079] Embodiment 1: A method for constructing a digital human's personality, image, voice and life trajectory in an integrated manner, which is applicable to work scenarios such as social media management and content marketing of virtual digital humans. Figure 1 As shown, the main steps of the method include:

[0080] 1. The character generation module can easily build the core features of virtual digital people using specific prompt words or preset options, including name, gender, age, nationality, education, personality, hobbies, skills, company affiliation, emotional state, religious beliefs and educational background. Subsequently, the authenticity and accuracy of this information is ensured through the review of precise algorithms. In addition, the system will also perform similarity analysis on the generated feature values ​​to ensure that the feature combination of each digital person is unique, thereby ensuring the uniqueness of the digital person's feature values;

[0081] 2. The life trajectory module selects some important time nodes based on certain characteristic values ​​of the generated digital person, such as gender, age, personality, education, hobbies, skills, emotional state, educational background and other information, generates events of important nodes in the life of this digital person, and calls the image service obtained in step 3 and the voice service obtained in step 4 to generate pictures and voices of these nodes;

[0082] Important node events refer to some special events in the life of the digital person generated by the life trajectory, such as starting to go to school at the age of 4, entering a secondary school at the age of 12, entering a university at the age of 18, getting engaged at the age of 24, and getting married at the age of 26. When the digital person actually chats later, the character's characteristics and personality can be analyzed again based on the character events and special anchor points triggered, making the character image more realistic.

[0083] 3. The image generation module generates a picture that best fits the current digital human image based on certain feature values ​​of the digital human, such as gender, age, hobbies, education, emotional state, family background, etc., or by directly adding some prompt words, such as appearance description prompt words such as "there is a beauty mole on the side of the left eye";

[0084] 4. The voice generation module can generate voices based on fixed voice prompt words, such as tenor, or based on certain characteristic values ​​of the digital person, such as gender, personality, etc., and then generate a text message based on the digital person's hobbies or important events in his life trajectory, and generate voices based on the text message and voice;

[0085] 5. The behavior analysis module mainly sets some anchor points based on the real work scenarios of social media management and content marketing of digital people after they are successfully created, and records fixed information based on these anchor points. For example, if a video is particularly funny, a text record is generated, a fixed voice is generated, and fixed character events are generated at special times, such as birthdays, receiving some gifts, generating some pictures and recording them in personal events, and even generating new feature values, which in turn drives changes in the entire digital person's personality, life trajectory, including image, voice, etc.;

[0086] 5.1 Setting of special anchor points, preferably some iconic events;

[0087] It can be based on the company's specific business needs. For example, when a digital person browses a post about beautiful scenery, an anchor point can be triggered. At this time, events can be used to report data and report the digital person information and post information.

[0088] 5.2 Anchor trigger: record the event theme after triggering, such as a particularly beautiful poem triggered in social media;

[0089] 5.3 Anchor event recording. After the anchor is triggered, some voice or image can be generated according to the theme of the event;

[0090] 5.4 Event influence: determine whether the anchor event will have an impact on the character's personality. For example, if a divorced person browses divorce-related content again, will it affect his emotions and generate new personality feature values?

[0091] 5.5 Analyze the anchor point events, compare the differences of multiple anchor point events, and then cancel the anchor point or set a new anchor point;

[0092] The anchor point will generate some data information, analyze the anchor point data of a digital person over a period of time, compare the differences between these events, and then reduce the frequency of data collection reported by similar anchor points, and increase the amount of reported data for those with excessive differences.

[0093] 6. The data processing module mainly handles the calls between modules, ensuring the calling methods between modules and the specific methods of calling implementation. Since it involves certain computing power issues, such a design can save some computing power resources and perform unified data maintenance;

[0094] 6.1 The calling method adopts the workflow method because the image, voice, and life trajectory services all rely on the feature values ​​generated by the personality service first, so the order of service calling needs to be controlled;

[0095] 6.2 Implementation method Using the publish-subscribe method, the request duration, computing resources, and processing speed issues between services can save a certain amount of computing resources;

[0096] 6.3 Authentication method: Authentication is performed through token request verification;

[0097] 6.4 Set up request task waiting queues between modules, reasonably schedule computing resources of each module, and give priority to completing necessary tasks. For example, for the personality generation service, the feature values ​​generated by the personality service can be interacted with in real time. This can also enhance the user's sense of participation in the generation phase, and at the same time improve the customization and personalization capabilities of the system. When resources are tight, submit the request task to the task waiting queue for storage. When the resources of the module service are idle, complete the tasks in the waiting queue in sequence and notify the publisher. This can effectively improve server resource utilization and control development investment costs.

[0098] The key technical points of this embodiment are:

[0099] 1. Personality generation: Through specific prompt words or preset options, the system can construct the core features of a virtual digital person, including name, gender, age, etc., and ensure the authenticity and accuracy of the information through sophisticated algorithms.

[0100] 2. Life trajectory generation: Based on the characteristic values ​​of the digital person, the system selects important time nodes, generates important events, and creates content for related nodes through image services and voice services.

[0101] 3. Image speech generation: the image generation module generates images based on the feature values ​​of the digital human, and the speech generation module generates speech based on the feature values ​​or prompt words.

[0102] 4. Behavioral analysis: by setting anchor points to record the behavior of digital humans and generating new human feature values ​​based on anchor point events, the entire digital human image will be made more realistic.

[0103] 5. Data processing, handling calls between modules, optimizing resource allocation, and improving system efficiency.

[0104] 6. Workflow call, the workflow method is used to ensure that the image, voice, and life trajectory services rely on the feature values ​​first generated by the personality service, and at the same time, the resources between each module can be reasonably allocated.

[0105] 7. Publish-subscribe implementation, using the publish-subscribe method to save computing resources and improve service efficiency.

[0106] 8. Authentication mechanism: authentication is performed through token request verification to ensure the security of the system.

[0107] 9. Task queue scheduling, set up a waiting queue for requested tasks, reasonably schedule computing resources, give priority to completing necessary tasks, and improve system response speed.

[0108] The data processing module can directly call the personality generation module, but the life trajectory module, voice generation module, and image generation module depend on the personality generation module. The behavior analysis module can generate new keywords, and then influence the life trajectory and voice and image modules through the data processing module;

[0109] like Figure 2 As shown, digital people are created on the business platform based on the integration of personality, image, voice and life trajectory:

[0110] 1) Deployment environment. Based on the existing AI large model, it integrates personality, image, voice, life trajectory, data processing, and behavior analysis functions. These modules are deployed on a high-performance server. The server is configured with two AMD EPYC7763 processors, 256G memory, and NVIDIA RTX A6000 graphics card with 48GB GDDR6 video memory. The existing Ollama large language model service tool is deployed based on phi3 and qwen2 as text models, Fooocus as image generation, whisper as speech recognition, fish-speech text-to-speech service, and data processing service to comprehensively call the services between various modules and realize data persistence.

[0111] 2) Add business modules for authentication and forwarding monitoring requests. Integrate Alibaba Cloud Object Storage as the image storage server, use MongoDB as the data persistence storage service, use Redis as the data cache storage, and implement subscription and publishing. Transform the original business modules, set anchor points, and connect to the behavior analysis service.

[0112] 3) The business platform extends its function to synchronously record the duration of creating the character’s characteristic values, image, life trajectory, voice, anchor point triggering frequency and other indicators during the digital human creation process.

[0113] 4) Browser display platform, which provides an entry point for creating digital humans and interactive pages during the creation process, as well as page data display effects after the creation is completed, provides some feature value filtering items, and provides data indicators after the creation is completed.

[0114] Example 1: Refined monitoring and analysis of a single creation:

[0115] The setting of creating a digital person is partially customized, setting the country of creation to Thailand, gender to male, age to 20-35, number of creations to single creation, generating the characteristic value of the person setting in 4 seconds, generating the image in 5 seconds, generating the voice in 5 seconds, generating the life trajectory in 4 seconds, and because of the use of publish-subscribe and asynchronous processing between modules in the creation process, the actual time in the process of creating a single digital person is 9 seconds, and the data has been persistently stored in this process and can be displayed synchronously on the page, and at this time the digital person has completely created all the data that basically meets the conditions. It is suitable for accurately creating digital people in a fixed area to facilitate the digital people to handle business in the area and make subsequent behavioral analysis more realistic.

[0116] Example 2: Generalized monitoring analysis of multiple creations:

[0117] The setting of creating digital people is partially customized. The country of creation is set to Thailand, the gender is male, the age is 20-35, the number of creations is batch creation, and the number of creations is 100. After actual testing, it is known that it takes 300 seconds to complete the generation of all digital people's characteristic values, and at this time, the image, voice and life trajectory of 30 digital people have been generated. The last data is completed at the 712th second. Since the request overhead in the actual request process is reduced during batch creation, the completion time is also much shorter than the normal single creation time*creation number. In addition, all requests do not always maintain long request links during the creation process, which further reduces resource overhead. The signal of whether the creation is completed is solved through the subscription and publishing solution. At the same time, this solution is implemented through caching, which is also more convenient and quick. It is suitable for creating a batch of digital people in large quantities to quickly meet business needs. If the server allows, distributed services can be configured for model support, then the speed of creating digital people will be almost the same, and the time interval in the middle will be greatly improved.

[0118] Example 3: Flexible monitoring and analysis triggered by anchor points:

[0119] After the anchor point is set and triggered, some of the character data of the current digital person is extracted, and the image, voice, and life trajectory are generated again based on the existing ones plus the new character feature values. The image generation time is 9 seconds, and the generated image needs to be re-analyzed. Based on the existing ones, the new feature values ​​are combined to generate new style pictures, and the voice generation time is 5 seconds. The generation of a new single life trajectory is 1 second. At this time, a new feature value and life trajectory of a certain digital person, as well as new pictures and voices are generated, which also directly and clearly verifies the real evolution process of this digital person. When there are enough anchor points set and the learning model connected is powerful enough, the digital person will become more and more realistic.

[0120] 3.1 Control Group A

[0121] Set the data processing service to proceed synchronously, maintain a long link between requests, and set it to random creation. The time for generating the characteristic value of the character is 4 seconds, the time for generating the image is 5 seconds, the time for generating the voice is 5 seconds, and the time for generating the life trajectory is 4 seconds. Then it takes 18 seconds to complete the creation. To complete a single creation, if 100 characters are created, the service for creating the character is single-threaded, and it takes 1800 seconds to complete the creation. Then within these 1800 seconds, service requests will be created 100*the number of service modules (times), and a long link must be maintained within this time range. Then the module running time is the service waiting time, resulting in a huge waste of time.

[0122] 3.2 Control group B

[0123] Set the data processing service to be asynchronous, maintain the subscription and publishing mode between requests, and set it to random creation. The time for generating the characteristic value of the character is 4 seconds, the time for generating the image is 5 seconds, the time for generating the voice is 5 seconds, and the time for generating the life trajectory is 4 seconds. Then, it takes 9 seconds to complete the creation. To complete a single creation, if 100 characters are created, and the service for creating the character is set to be single-threaded, it takes 300 seconds to create the character, and 712 seconds to complete the creation. Then, within these 712 seconds, the number of service requests is equal to the number of service modules, which saves 60% of the time and space overhead compared to the control group A, and the consumed service performance is almost negligible.

[0124] 3.3 Control group C

[0125] Set the AI ​​business support to distributed support, and set the concurrency to 3, prohibit the asynchronous data processing service, maintain the subscription and release mode between requests, and set it to random creation. The time for generating the characteristic value of the character is 4 seconds, the time for generating the image is 5 seconds, the time for generating the voice is 5 seconds, and the time for generating the life trajectory is 4 seconds. Then, it takes 9 seconds to complete the creation. To complete a single creation, if 100 characters are created, it only takes 143 seconds to complete the creation, and it takes 415 seconds to complete the creation. Then, within these 415 seconds, the number of service requests is equal to the number of service modules, which is the same as the control group B, but the server cost is doubled, saving 20% ​​of time. Compared with the control group A, the server cost is also doubled, saving 80% of time.

[0126] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules, modules or units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units, modules or components can be combined or integrated into another device, or some features can be ignored or not executed.

[0127] The units may or may not be physically separated, and the components displayed as units may be one physical unit or multiple physical units, that is, they may be located in one place or distributed in multiple different places. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0128] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0129] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part, and / or installed from a removable medium. When the computer program is executed by the central processing unit (CPU), the above-mentioned functions defined in the method of the present invention are executed. It should be noted that the above-mentioned computer-readable medium of the present invention can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of an electrical, magnetic, optical, electromagnetic, infrared segment, or semiconductor, or any combination of the above.

[0130] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0131] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for integrating the personality, image, voice and life trajectory of a digital human, characterized in that: The following steps are involved: Generate virtual digital people; Construct the life trajectory, image and voice of virtual digital people; Conduct behavioral analysis on virtual digital humans and update their life trajectory, image, and voice.

2. The method for constructing the digital human's personality, image, voice and life trajectory in an integrated manner according to claim 1, characterized in that: Generating a virtual digital human includes the following steps: Get prompt words or preset options; According to the prompt words or preset options, the characteristics of the virtual digital human are constructed to generate the virtual digital human.

3. The method for constructing the digital human's personality, image, voice and life trajectory in an integrated manner according to claim 2, characterized in that: The characteristics of the virtual digital person include name, gender, age, nationality, education background, personality, hobbies, skills, company affiliation, emotional state, religious beliefs and educational background.

4. The method for constructing the digital human's personality, image, voice and life trajectory in an integrated manner according to claim 3, characterized in that: The following steps are also included: Conduct similarity analysis on the core features of virtual digital humans to ensure that the feature combination of each digital human is unique.

5. The method for constructing the digital human's personality, image, voice and life trajectory in an integrated manner according to claim 4, characterized in that: Constructing the life trajectory, image and voice of a virtual digital person includes the following steps: Generate important time nodes and important events as life trajectories based on the gender, age, personality, education, hobbies, skills, emotional state and educational background of the virtual digital person; Generate an image of a virtual digital person based on the virtual digital person's gender, age, hobbies, education, emotional state, family background, and image prompt words; The timbre is generated according to the fixed timbre prompt words or the gender and personality of the virtual digital person, and then the text information is generated according to the hobbies or important events of the life trajectory of the virtual digital person, and the voice is generated according to the timbre and text information.

6. The method for constructing the digital human's personality, image, voice and life trajectory in an integrated manner according to claim 5, characterized in that: The following steps are also included: Based on the image and voice of the virtual digital person, important pictures and important voices of important time nodes and important events are generated.

7. The method for constructing the digital human's personality, image, voice and life trajectory in an integrated manner according to claim 6, characterized in that: Conducting behavioral analysis on virtual digital humans and updating their life trajectory, image, and voice specifically includes the following steps: In a real work scenario, set anchor points; After the anchor is triggered, the event subject is recorded; Generate an event theme image or event theme voice based on the event theme, the image and voice of the virtual digital person; Update the characteristics of the virtual digital person according to the event theme to obtain the updated characteristics of the virtual digital person; Based on the updated characteristics of the virtual digital person, the life trajectory, image and voice of the virtual digital person are reconstructed.

8. The method for constructing the digital human's personality, image, voice and life trajectory in an integrated manner according to claim 7, characterized in that: The following steps are also included: Compare the differences of all anchor points to obtain the differences; The anchor point with a difference less than the minimum difference threshold is taken as the first anchor point, and the newly set anchor points similar to the first anchor point are reduced; The anchor point whose difference is greater than the maximum difference threshold is taken as the second anchor point, and a new anchor point similar to the second anchor point is added.

9. A system for integrating the character setting, image, voice and life trajectory of a digital human, used to implement the method for integrating the character setting, image, voice and life trajectory of a digital human as claimed in any one of claims 1 to 8, characterized in that: include: A character generation module is used to generate a virtual digital person; Life trajectory, image and voice construction module, used to construct the life trajectory, image and voice of virtual digital people; The behavior analysis module is used to analyze the behavior of the virtual digital human and update the life trajectory, image and voice of the virtual digital human.

Images