Virtual human intelligent system with information iteration function

By introducing intelligent information perception, understanding and expression modules into the virtual human system, combined with deep learning algorithms, the problem of insufficient intelligent understanding capabilities of virtual humans is solved, the information iteration function is realized, and the interactive experience and application effect are improved.

CN119025962APending Publication Date: 2024-11-26WUHAN TRUE COLOR INTELLIGENT MFG TECH CO LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202411101650.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-12
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing virtual human system has insufficient intelligent understanding capabilities and cannot fully identify information, resulting in poor interactive experience and lack of information iteration functions, affecting the actual application effect.

Method used

A virtual human intelligent system with information iteration function was designed, including intelligent information perception and acceptance module, intelligent information understanding module, intelligent information screening module and intelligent expression module, and the model was performed using gradient descent algorithm, Adam algorithm, RMSprop algorithm and Adagrad algorithm. Training optimization combines speech, facial expressions, body movements and image information recognition to realize adaptive learning and information iteration.

Benefits of technology

It improves the virtual human perception of environmental information, enhances the interaction effect with humans, can understand human intentions and performs automatic information iteration, significantly improving the interactive experience and the practical application effect of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025962A_ABST
    Figure CN119025962A_ABST
Patent Text Reader

Abstract

The invention discloses a virtual human intelligent system with an information iteration function, and belongs to the technical field of virtual humans, the virtual human intelligent system comprises an intelligent information sensing and receiving module, the output end of the intelligent information sensing and receiving module is connected with the input end of an intelligent information understanding module; according to the system, the intelligent information sensing and receiving module is arranged in the system and includes voice information sensing, facial expression sensing, body movement sensing, text information recognition and image information recognition, so that when the system is applied, most of information in the environment can be sensed, the information sensing amount of the virtual human can be increased, and the user experience is improved. The system is further internally provided with an intelligent information understanding module, and the module can understand meanings represented behind various kinds of information, has human thinking, can easily understand human intentions, can automatically complete information iteration, and is powerful in function.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of virtual humans, and in particular relates to a virtual human intelligent system with an information iteration function. Background Art

[0002] Virtual human is a comprehensive product that integrates a variety of advanced technologies. It simulates the real human body or human image through digital technology and has rich interactive capabilities and a wide range of application scenarios.

[0003] A Chinese patent discloses (CN105574767A) an anthropomorphic aquarium virtual social intelligent system, including an intelligent fish tank, which is equipped with a temperature sensor, a light sensor, an automatic fish feeding control device and a camera; the temperature sensor, the light sensor, the automatic fish feeding control device are connected to the intelligent control chip; the intelligent control chip and the camera are connected to a PC; the PC is connected to a remote terminal through a server; the system collects various indicator data in the intelligent fish tank in real time, currently including: temperature, light intensity, automatic feeding interval and time, and images in the fish tank, and transmits these data to the PC end, the PC end application saves the data to the database, and according to the built-in algorithm of the software, performs natural language simulation conversion to form anthropomorphic language, the PC end encapsulates the anthropomorphic language into a message and sends it to the server; it has the characteristics of simple structure and easy operation. Although today's virtual human systems have good virtual simulation effects, their intelligent comprehension capabilities are insufficient and they are unable to identify information more comprehensively, resulting in a poor interaction experience. At the same time, they do not have information iteration functions, and the actual application effect of the overall system is not good. In order to solve the above problems, a virtual human intelligent system with information iteration functions is urgently needed. Summary of the invention

[0004] The purpose of the present invention is to solve the problem that although the current virtual human system has a good virtual simulation effect, its intelligent comprehension ability is insufficient and it is unable to recognize information more comprehensively, which leads to a bad experience in interaction. At the same time, it does not have the information iteration function and the actual application effect of the overall system is poor. A virtual human intelligent system with information iteration function is proposed.

[0005] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a virtual human intelligent system with information iteration function, including an intelligent information perception and receiving module, the output end of the intelligent information perception and receiving module is connected to the input end of the intelligent information understanding module, the output end of the intelligent information understanding module is connected to the input end of the intelligent information screening module, the output end of the intelligent information screening module is connected to the input end of the intelligent expression module, and the output end of the intelligent expression module is connected to the input end of the adaptive learning module.

[0006] As a further description of the above technical solution:

[0007] The output end of the adaptive learning module is connected to the input end of the privacy data automatic encryption module, and the output end of the privacy data automatic encryption module is connected to the input end of the model training optimization module. The algorithm that needs to be applied in the model training optimization module is one or more of the gradient descent algorithm, Adam algorithm, RMSprop algorithm, and Adagrad algorithm.

[0008] As a further description of the above technical solution:

[0009] The intelligent information perception and receiving module includes a voice intelligent automatic recognition and acquisition module and a facial expression intelligent capture and acquisition module, the output end of the voice intelligent automatic recognition and acquisition module is connected to the input end of the facial expression intelligent capture and acquisition module, and the output end of the facial expression intelligent capture and acquisition module is connected to the input end of the body movement intelligent capture and acquisition module.

[0010] As a further description of the above technical solution:

[0011] The output end of the body movement intelligent capture and acquisition module is connected to the input end of the text information intelligent recognition and acquisition module, and the output end of the text information intelligent recognition and acquisition module is connected to the input end of the image information intelligent recognition and acquisition module.

[0012] As a further description of the above technical solution:

[0013] The intelligent information understanding module includes a knowledge base graph import module and a speech semantic understanding module, wherein the output end of the knowledge base graph import module is connected to the input end of the speech semantic understanding module, and the output end of the speech semantic understanding module is connected to the input end of the facial expression feature extraction module.

[0014] As a further description of the above technical solution:

[0015] The output end of the facial expression feature extraction module is connected to the input end of the facial expression analysis and understanding module, the output end of the facial expression analysis and understanding module is connected to the input end of the body movement feature extraction module, and the output end of the body movement feature extraction module is connected to the input end of the body language analysis and understanding module.

[0016] As a further description of the above technical solution:

[0017] The intelligent information screening module includes an information cleaning module and a user information analysis module, and the output end of the information cleaning module is connected to the input end of the user information analysis module.

[0018] As a further description of the above technical solution:

[0019] The output end of the user information analysis module is connected to the input end of the information acquisition and screening module, the output end of the information acquisition and screening module is connected to the input end of the machine deep learning module, and the output end of the machine deep learning module is connected to the input end of the information real-time update and iteration module.

[0020] As a further description of the above technical solution:

[0021] The intelligent expression module includes a machine language conversion module and a natural speech generation module, the output end of the machine language conversion module is connected to the input end of the natural speech generation module, the output end of the natural speech generation module is connected to the input end of the context automatic simulation module, the output end of the context automatic simulation module is connected to the input end of the virtual human expression generation module, and the output end of the virtual human expression generation module is connected to the input end of the user interaction module.

[0022] As a further description of the above technical solution:

[0023] The model training optimization module includes a speech recognition training module, the output end of the speech recognition training module is connected to the input end of the body movement recognition training module, the output end of the body movement recognition training module is connected to the input end of the facial expression recognition training module, and the output end of the facial expression recognition training module is connected to the input end of the text image information recognition training module.

[0024] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0025] In the present invention, an intelligent information perception and receiving module is provided inside, and the module covers voice information perception, facial expression perception, body movement perception, text information recognition and image information recognition, so that the system can perceive most of the information in the environment when applied, and can expand the information perception of the virtual person, thereby improving the interaction effect between the virtual person and the person. The system is also provided with an intelligent information understanding module, which can understand the meaning behind various types of information, has human thinking, can easily understand human intentions, can automatically complete information iteration, and has powerful functions. The system is also provided with an intelligent expression module, which can enable the virtual person to interact with people according to the situation, and can also make virtual expressions, which can greatly improve the interactive experience and greatly improve the actual application effect of the overall system. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 The figure is a schematic diagram of the module structure of a virtual human intelligent system with information iteration function.

[0027] Figure 2The diagram is a submodule structure diagram of an intelligent information perception and receiving module in a virtual human intelligent system with information iteration function.

[0028] Figure 3 The diagram is a submodule structure diagram of an intelligent information understanding module in a virtual human intelligent system with information iteration function.

[0029] Figure 4 The present invention is a schematic diagram of the submodule structure of an intelligent information screening module in a virtual human intelligent system with information iteration function.

[0030] Figure 5 The figure is a schematic diagram of the submodule structure of an intelligent expression module in a virtual human intelligent system with information iteration function.

[0031] Figure 6 This is a schematic diagram of the submodule structure of a model training optimization module in a virtual human intelligent system with information iteration function.

[0032] Legend:

[0033] 1. Intelligent information perception and acceptance module; 2. Intelligent information understanding module; 3. Intelligent information screening module; 4. Intelligent expression module; 5. Model training and optimization module. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] Example 1

[0036] See also Figure 1-6 The present invention provides a technical solution: a virtual human intelligent system with information iteration function, comprising an intelligent information perception and receiving module 1, wherein the output end of the intelligent information perception and receiving module 1 is connected to the input end of the intelligent information understanding module 2, the output end of the intelligent information understanding module 2 is connected to the input end of the intelligent information screening module 3, the output end of the intelligent information screening module 3 is connected to the input end of the intelligent expression module 4, the output end of the intelligent expression module 4 is connected to the input end of the adaptive learning module, the output end of the adaptive learning module is connected to the input end of the privacy data automatic encryption module, the output end of the privacy data automatic encryption module is connected to the input end of the model training optimization module, and the algorithms to be applied in the model training optimization module are gradient descent algorithm, Adam algorithm, RMSprop algorithm and Adagrad algorithm;

[0037] The intelligent information perception and receiving module 1 includes a voice intelligent automatic recognition and acquisition module and a facial expression intelligent capture and acquisition module, the output end of the voice intelligent automatic recognition and acquisition module is connected to the input end of the facial expression intelligent capture and acquisition module, the output end of the facial expression intelligent capture and acquisition module is connected to the input end of the body movement intelligent capture and acquisition module, the output end of the body movement intelligent capture and acquisition module is connected to the input end of the text information intelligent recognition and acquisition module, and the output end of the text information intelligent recognition and acquisition module is connected to the input end of the image information intelligent recognition and acquisition module.

[0038] The intelligent information understanding module 2 includes a knowledge base graph import module and a speech semantic understanding module, the output end of the knowledge base graph import module is connected to the input end of the speech semantic understanding module, the output end of the speech semantic understanding module is connected to the input end of the facial expression feature extraction module, the output end of the facial expression feature extraction module is connected to the input end of the facial expression analysis and understanding module, the output end of the facial expression analysis and understanding module is connected to the input end of the limb movement feature extraction module, and the output end of the limb movement feature extraction module is connected to the input end of the body language analysis and understanding module.

[0039] The intelligent information screening module 3 includes an information cleaning module and a user information analysis module. The output end of the information cleaning module is connected to the input end of the user information analysis module, the output end of the user information analysis module is connected to the input end of the information acquisition screening module, the output end of the information acquisition screening module is connected to the input end of the machine deep learning module, and the output end of the machine deep learning module is connected to the input end of the information real-time update and iteration module.

[0040] The intelligent expression module 4 includes a machine language conversion module and a natural speech generation module, the output end of the machine language conversion module is connected to the input end of the natural speech generation module, the output end of the natural speech generation module is connected to the input end of the context automatic simulation module, the output end of the context automatic simulation module is connected to the input end of the virtual human expression generation module, and the output end of the virtual human expression generation module is connected to the input end of the user interaction module.

[0041] The model training optimization module 5 includes a speech recognition training module, the output end of the speech recognition training module is connected to the input end of the body movement recognition training module, the output end of the body movement recognition training module is connected to the input end of the facial expression recognition training module, and the output end of the facial expression recognition training module is connected to the input end of the text image information recognition training module.

[0042] In this embodiment, an intelligent information perception and receiving module is provided inside, which includes voice information perception, facial expression perception, body movement perception, text information recognition and image information recognition. This enables the system to perceive most of the information in the environment when used, and to expand the information perception of the virtual person, thereby improving the interaction effect between the system and people.

[0043] Example 2

[0044] See also Figure 1-6 The present invention provides a technical solution: a virtual human intelligent system with information iteration function, comprising an intelligent information perception and receiving module 1, wherein the output end of the intelligent information perception and receiving module 1 is connected to the input end of the intelligent information understanding module 2, the output end of the intelligent information understanding module 2 is connected to the input end of the intelligent information screening module 3, the output end of the intelligent information screening module 3 is connected to the input end of the intelligent expression module 4, the output end of the intelligent expression module 4 is connected to the input end of the adaptive learning module, the output end of the adaptive learning module is connected to the input end of the privacy data automatic encryption module, the output end of the privacy data automatic encryption module is connected to the input end of the model training optimization module, and the algorithms to be applied in the model training optimization module are the gradient descent algorithm and the Adagrad algorithm;

[0045] The intelligent information perception and receiving module 1 includes a voice intelligent automatic recognition and acquisition module and a facial expression intelligent capture and acquisition module, the output end of the voice intelligent automatic recognition and acquisition module is connected to the input end of the facial expression intelligent capture and acquisition module, the output end of the facial expression intelligent capture and acquisition module is connected to the input end of the body movement intelligent capture and acquisition module, the output end of the body movement intelligent capture and acquisition module is connected to the input end of the text information intelligent recognition and acquisition module, and the output end of the text information intelligent recognition and acquisition module is connected to the input end of the image information intelligent recognition and acquisition module.

[0046] The intelligent information understanding module 2 includes a knowledge base graph import module and a speech semantic understanding module, the output end of the knowledge base graph import module is connected to the input end of the speech semantic understanding module, the output end of the speech semantic understanding module is connected to the input end of the facial expression feature extraction module, the output end of the facial expression feature extraction module is connected to the input end of the facial expression analysis and understanding module, the output end of the facial expression analysis and understanding module is connected to the input end of the limb movement feature extraction module, and the output end of the limb movement feature extraction module is connected to the input end of the body language analysis and understanding module.

[0047] The intelligent information screening module 3 includes an information cleaning module and a user information analysis module. The output end of the information cleaning module is connected to the input end of the user information analysis module, the output end of the user information analysis module is connected to the input end of the information acquisition screening module, the output end of the information acquisition screening module is connected to the input end of the machine deep learning module, and the output end of the machine deep learning module is connected to the input end of the information real-time update and iteration module.

[0048] The intelligent expression module 4 includes a machine language conversion module and a natural speech generation module, the output end of the machine language conversion module is connected to the input end of the natural speech generation module, the output end of the natural speech generation module is connected to the input end of the context automatic simulation module, the output end of the context automatic simulation module is connected to the input end of the virtual human expression generation module, and the output end of the virtual human expression generation module is connected to the input end of the user interaction module.

[0049] The model training optimization module 5 includes a speech recognition training module, the output end of the speech recognition training module is connected to the input end of the body movement recognition training module, the output end of the body movement recognition training module is connected to the input end of the facial expression recognition training module, and the output end of the facial expression recognition training module is connected to the input end of the text image information recognition training module.

[0050] In this embodiment, an intelligent expression module is also provided in the system, which enables the virtual human to interact with people according to the situation and can also make virtual expressions, which can greatly improve the interactive experience and greatly improve the actual application effect of the overall system.

[0051] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A virtual human intelligent system with information iteration function, comprising an intelligent information perception and receiving module (1), characterized in that: The output end of the intelligent information perception and receiving module (1) is connected to the input end of the intelligent information understanding module (2), the output end of the intelligent information understanding module (2) is connected to the input end of the intelligent information screening module (3), the output end of the intelligent information screening module (3) is connected to the input end of the intelligent expression module (4), and the output end of the intelligent expression module (4) is connected to the input end of the adaptive learning module.

2. A virtual human intelligent system with information iteration function according to claim 1, characterized in that: The output end of the adaptive learning module is connected to the input end of the privacy data automatic encryption module, and the output end of the privacy data automatic encryption module is connected to the input end of the model training optimization module. The algorithm that needs to be applied in the model training optimization module is one or more of the gradient descent algorithm, Adam algorithm, RMSprop algorithm, and Adagrad algorithm.

3. A virtual human intelligent system with information iteration function according to claim 2, characterized in that: The intelligent information perception and receiving module (1) comprises a voice intelligent automatic recognition and acquisition module and a facial expression intelligent capture and acquisition module, wherein the output end of the voice intelligent automatic recognition and acquisition module is connected to the input end of the facial expression intelligent capture and acquisition module, and the output end of the facial expression intelligent capture and acquisition module is connected to the input end of the body movement intelligent capture and acquisition module.

4. A virtual human intelligent system with information iteration function according to claim 3, characterized in that: The output end of the body movement intelligent capture and acquisition module is connected to the input end of the text information intelligent recognition and acquisition module, and the output end of the text information intelligent recognition and acquisition module is connected to the input end of the image information intelligent recognition and acquisition module.

5. A virtual human intelligent system with information iteration function according to claim 4, characterized in that: The intelligent information understanding module (2) comprises a knowledge base graph import module and a speech semantic understanding module, wherein the output end of the knowledge base graph import module is connected to the input end of the speech semantic understanding module, and the output end of the speech semantic understanding module is connected to the input end of the facial expression feature extraction module.

6. A virtual human intelligent system with information iteration function according to claim 5, characterized in that: The output end of the facial expression feature extraction module is connected to the input end of the facial expression analysis and understanding module, the output end of the facial expression analysis and understanding module is connected to the input end of the body movement feature extraction module, and the output end of the body movement feature extraction module is connected to the input end of the body language analysis and understanding module.

7. A virtual human intelligent system with information iteration function according to claim 6, characterized in that: The intelligent information screening module (3) comprises an information cleaning module and a user information analysis module, and the output end of the information cleaning module is connected to the input end of the user information analysis module.

8. A virtual human intelligent system with information iteration function according to claim 7, characterized in that: The output end of the user information analysis module is connected to the input end of the information acquisition and screening module, the output end of the information acquisition and screening module is connected to the input end of the machine deep learning module, and the output end of the machine deep learning module is connected to the input end of the information real-time update and iteration module.

9. A virtual human intelligent system with information iteration function according to claim 8, characterized in that: The intelligent expression module (4) comprises a machine language conversion module and a natural speech generation module, wherein the output end of the machine language conversion module is connected to the input end of the natural speech generation module, the output end of the natural speech generation module is connected to the input end of the context automatic simulation module, the output end of the context automatic simulation module is connected to the input end of the virtual human expression generation module, and the output end of the virtual human expression generation module is connected to the input end of the user interaction module.

10. A virtual human intelligent system with information iteration function according to claim 9, characterized in that: The model training optimization module (5) comprises a speech recognition training module, the output end of the speech recognition training module is connected to the input end of the body movement recognition training module, the output end of the body movement recognition training module is connected to the input end of the facial expression recognition training module, and the output end of the facial expression recognition training module is connected to the input end of the text image information recognition training module.

Citation Information

Patent Citations

  • System and method for controlling virtual portrait through signals and medium

    CN111897434A

  • Virtual human emotion interaction system and method based on five-dimensional emotion model

    CN115543089A

  • Holographic digital human security modeling method and system

    CN116975883A

  • Virtual clothing display method and system for multi-user scene, equipment and medium

    CN117495495A

  • Personalized three-dimensional digital human holographic interaction forming system and method

    CN117523088A