Interactive agent cooperation system and control method thereof

Abstract

The invention relates to the technical field of intelligent interaction, in particular to an interaction agent collaboration system and a control method thereof, and solves the problems that in the interaction process of existing artificial intelligence, emotion perception cannot be performed according to interaction data of a user, so that accurate capture of user requirements is inconvenient to realize, and user experience is poor. The system comprises a multi-port intelligent agent, the multi-port intelligent agent is composed of a software module and a hardware module, the software module is embedded in the hardware module, the software module is an M-Creation engine, and the hardware module is an M-Creation engine. The M-Creation engine comprises a data set, a storage module, a process arrangement platform, a plug-in tool set, a large language model, a back-end service platform and a cache module. Through a large amount of emotion cognition training, the multi-port intelligent agent can conveniently customize the preset role identity and the emotion setting operation in the interaction process, the user demand can be accurately captured, the response emotion and mood can be automatically adapted, and the application of different scenes can be met through the virtual digital human.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent interaction, and specifically provides an interactive intelligent agent collaboration system and a control method therefor. Background Art

[0002] Intelligent interaction refers to simulating human intelligent behavior through computer programs to achieve natural interaction with humans. With the continuous progress of technology, intelligent interaction will pay more attention to multi-modal interaction, such as voice, expression, semantic analysis, etc., to provide a more natural and immersive user experience. The main application scenarios of intelligent interaction include voice assistants, smart homes, smart cars, robots, etc. By collecting and analyzing user interaction data, such as voice commands, touch operations, etc., to understand user needs and habits, so as to optimize the interaction design and function implementation of the system. By analyzing the user's voice or text input, identifying the user's intentions and key information, and generating appropriate responses or operation instructions. Among them, intelligent voice interaction is a new generation of interaction mode based on voice input, and feedback results can be obtained by just speaking.

[0003] In the existing artificial intelligence during the interaction process, it is unable to perform emotion perception based on user interaction data, thus making it inconvenient to accurately capture user needs, and it is also inconvenient to automatically adapt and respond to emotions and tones; therefore, it does not meet the existing requirements. For this reason, we propose an interactive intelligent agent collaboration system and a control method therefor. Summary of the Invention

[0004] The purpose of the present invention is to provide an interactive intelligent agent collaboration system and a control method therefor, so as to solve the problem that the existing artificial intelligence in the interaction process cannot perform emotion perception based on user interaction data, thus making it inconvenient to accurately capture user needs, and it is also inconvenient to automatically adapt and respond to emotions and tones as mentioned in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An interactive intelligent agent collaboration system, including a multi-port intelligent agent, the multi-port intelligent agent is composed of a software module and a hardware module, the software module is embedded inside the hardware module, the software module is the M-Creation engine, and the M-Creation engine includes a data set, a storage module, a process orchestration platform, a plug-in tool set, a large language model, a backend service platform, and a cache module;

[0006] The hardware module includes an interactive display, an M-Creation workstation, an M-Creation multi-modal box, and an interactive robot.

[0007] Preferably, the data set is used to extract, transform, and load interaction data, and the process orchestration platform performs process orchestration on the interaction data through the data set.

[0008] Preferably, the core of the M-Creation engine is a multi-port agent, which has an interaction function. The interaction function of the multi-port agent consists of a custom preset role identity, a RAG knowledge base, a Q&A knowledge base, external plugins, and an emotion setting module.

[0009] Preferably, the storage module is built-in with a vector database, a relational database, and a knowledge graph. The storage and plugin toolset are electrically connected. The plugin toolset consists of multiple external sockets, and each external socket is plugged with an external plugin.

[0010] Preferably, the large language model includes a commercial large language model, an open-source large language model, and a model service platform. The large language model is connected to the process orchestration platform through a cache module. The backend service platform makes data requests to the large language model through the process orchestration platform.

[0011] Preferably, the application scenarios of the multi-port agent are respectively life assistance, interactive learning, medical consultation, tour guide, interactive live broadcast, and video marketing. The interactive live broadcast includes dynamic conversation generation, TTS voice synthesis, AI intelligent field control, distributed TTS voice live broadcast, and AI intelligent review.

[0012] Preferably, the video marketing includes AI scripts, one-click video creation, automatic distribution, matrix building, and intelligent customer service. The multi-port agent conducts data interaction by establishing virtual digital humans. The virtual digital humans include virtual hosts, virtual IP spokespersons, virtual human video broadcasts, and virtual human customer services.

[0013] A control method for an interactive agent collaboration system includes the following steps:

[0014] S1: Select the hardware module of the interactive agent according to the application scenario of the interactive agent. The hardware module of the interactive agent includes an interactive display, an M-Creation workstation, an M-Creation multi-modal box, and an interactive robot. Embed the software module of the interactive agent, that is, the M-Creation engine, into the hardware module, and select the hardware module according to the application scenario;

[0015] S2: The application scenarios of the interactive agent include but are not limited to life assistance, interactive learning, medical consultation, tour guide, interactive live broadcast, and video marketing. Apply the core of the M-Creation engine, that is, the multi-port agent, so that the multi-port agent extracts, transforms, and loads the interactive data using the dataset and transmits it to the process orchestration platform. Through the dataset, the interactive data can be processed, reducing the operating pressure of the process orchestration platform;

[0016] S3: Based on different application scenarios, the process orchestration platform controls the output of the processed interaction data through the plug-in toolset and the large language model to meet the query requests in different interaction processes. At the same time, the multi-port intelligent agent conducts data interaction by establishing virtual digital humans, and through the virtual digital humans, the outbound actions in multiple scenarios can be realized, including virtual hosts, virtual IP spokespersons, virtual human video broadcasts, and virtual human customer services, reducing the labor intensity of users;

[0017] S4: The M-Creation engine trained through multiple scenario interactions can integrate and optimize various interaction data. At the same time, through a large number of emotional cognition trainings, it is convenient for the multi-port intelligent agent to customize the preset role identity and emotional setting operations during the interaction process, and through the interaction input of data such as voice, image, and text, it can accurately capture the user's needs and automatically adapt to the response emotions and tones.

[0018] Compared with the prior art, the beneficial effects of the present invention are:

[0019] 1. In the present invention, by embedding the M-Creation engine in multiple hardware applicable to different scenarios, the multiple hardware can achieve multi-modal interaction effects through the M-Creation engine. And by establishing virtual digital humans, it is convenient to reduce the user's labor volume. Through the virtual digital humans, the outbound actions in multiple scenarios can be realized, improving the accuracy of data output;

[0020] 2. Through the integration and optimization of multi-modal interaction capabilities and a large number of emotional cognition trainings, the present invention realizes fast and accurate information input and feedback, significantly improving the naturalness and fluency of user interaction. The model has an emotion perception function, can accurately detect and analyze facial expressions, accurately capture the user's needs through multi-dimensional information such as voice, image, and text, and automatically adapt to emotional responses and tones, enabling users to obtain a real and considerate interaction experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the architecture of the multi-port intelligent agent of the present invention;

[0022] Figure 2 It is a schematic diagram of the comparison of enterprise IT before and after being empowered by the large model of the present invention;

[0023] Figure 3 It is a schematic diagram of the structure of the interactive intelligent agent of the present invention applied to campus management;

[0024] Figure 4 It is a schematic diagram of the structure of the interactive intelligent agent of the present invention applied to school-level education;

[0025] Figure 5 It is the overall control flow chart of the present invention. Detailed implementation manners

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0027] Please refer to Figures 1 to 5 , an embodiment provided by the present invention: an interactive agent collaboration system, including a multi-port agent. The multi-port agent is composed of a software module and a hardware module. The software module is embedded inside the hardware module. The software module is the M-Creation engine. The M-Creation engine includes a data set, a storage module, a process orchestration platform, a plug-in tool set, a large language model, a backend service platform, and a cache module. The hardware module includes an interactive display, an M-Creation workstation, an M-Creation multi-modal box, and an interactive robot. The hardware module is selected according to the application scenario to meet the corresponding settings and data output of the M-Creation engine according to different application scenarios;

[0028] The storage module is built-in with a vector database, a relational database, and a knowledge graph. The data set is used to extract, transform, and load interactive data. The process orchestration platform orchestrates the interactive data through the data set. The large language model includes a commercial large language model, an open-source large language model, and a model service platform. The large language model is connected to the process orchestration platform through the cache module. The backend service platform requests data from the large language model through the process orchestration platform, so that the multi-port agent extracts, transforms, and loads the interactive data using the data set and transmits it to the process orchestration platform. The interactive data can be processed through the data set, reducing the operating pressure of the process orchestration platform.

[0029] Please refer to Figure 1 , the core of the M-Creation engine is the multi-port agent. The multi-port agent has an interactive function. The interactive function of the multi-port agent is composed of a custom preset role identity, a RAG knowledge base, a question-and-answer knowledge base, external plug-ins, and an emotion setting module. The storage is electrically connected to the plug-in tool set. The plug-in tool set consists of multiple external sockets, and each external socket is plugged with an external plug-in. The process orchestration platform controls the output of the processed interactive data through the plug-in tool set and the large language model according to different application scenarios to meet the query requests in different interactive processes.

[0030] Among them, the application scenarios of the multi-port intelligent agent are respectively life assistance, interactive learning, medical consultation, tour guide, interactive live broadcast and video marketing. The interactive live broadcast includes dynamic speech generation, TTS voice synthesis, AI intelligent field control, distributed TTS voice live broadcast and AI intelligent review. The video marketing includes AI scripts, one-click video creation, automatic distribution, matrix building and intelligent customer service. The multi-port intelligent agent conducts data interaction by establishing virtual digital humans, which include virtual hosts, virtual IP spokespersons, virtual human video broadcasts and virtual human customer services. Through the virtual digital humans, the outbound actions of multiple scenarios can be realized, reducing the labor intensity of users.

[0031] Please refer to Figure 5 , a control method for an interactive intelligent agent collaboration system, comprising the following steps:

[0032] S1: Select the hardware module of the interactive intelligent agent according to the application scenario of the interactive intelligent agent. The hardware module of the interactive intelligent agent includes an interactive display, an M-Creation workstation, an M-Creation multimodal box and an interactive robot. Embed the software module of the interactive intelligent agent, namely the M-Creation engine, into the hardware module, and select the hardware module according to the application scenario;

[0033] S2: The application scenarios of the interactive intelligent agent include but are not limited to life assistance, interactive learning, medical consultation, tour guide, interactive live broadcast and video marketing. Apply the multi-port intelligent agent, which is the core of the M-Creation engine, so that the multi-port intelligent agent extracts, transforms and loads the interactive data using the data set and transmits it to the process orchestration platform. Through the data set, the interactive data can be processed, reducing the operating pressure of the process orchestration platform;

[0034] S3: The process orchestration platform controls the output of the processed interactive data through the plug-in tool set and the large language model according to different application scenarios to meet the query requests in different interactive processes. At the same time, the multi-port intelligent agent conducts data interaction by establishing virtual digital humans. Through the virtual digital humans, the outbound actions of multiple scenarios can be realized, including virtual hosts, virtual IP spokespersons, virtual human video broadcasts and virtual human customer services, reducing the labor intensity of users;

[0035] S4: The M-Creation engine trained through multiple scenario interactions can integrate and optimize various interactive data. At the same time, through a large number of emotional cognition trainings, it is convenient for the multi-port intelligent agent to customize the preset role identity and emotional setting operations during the interaction process, and through the interactive input of data such as voice, image and text, it can accurately capture the user's needs and automatically adapt to respond emotions and tones.

[0036] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above-described exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, in any regard, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. An interactive agent collaboration system, including a multi-port agent, the multi-port agent is composed of a software module and a hardware module, and the software module is embedded inside the hardware module, characterized in that: The software module is the M-Creation engine, and the M-Creation engine includes a data set, a storage module, a process orchestration platform, a plug-in tool set, a large language model, a back-end service platform, and a cache module; The hardware module includes an interactive display, an M-Creation workstation, an M-Creation multimodal box, and an interactive robot.

2. The interactive intelligent agent collaboration system according to claim 1, wherein: The data set is used to extract, transform, and load interactive data, and the process orchestration platform orchestrates the interactive data through the data set.

3. The interactive intelligent agent cooperation system according to claim 1, characterized in that: The core of the M-Creation engine is a multi-port agent. The multi-port agent has an interaction function, and the interaction function of the multi-port agent consists of a custom preset role identity, a RAG knowledge base, a question-and-answer knowledge base, external plug-ins, and an emotion setting module.

4. An interactive intelligent agent collaboration system according to claim 1, characterized in that: The storage module is built-in with a vector database, a relational database, and a knowledge graph. The storage is electrically connected to the plug-in tool set. The plug-in tool set consists of multiple external sockets, and each external socket is plugged with an external plug-in.

5. An interactive intelligent agent collaboration system according to claim 1, characterized in that: The large language model includes a commercial large language model, an open-source large language model, and a model service platform. The large language model is connected to the process orchestration platform through the cache module, and the back-end service platform requests data from the large language model through the process orchestration platform.

6. An interactive agent collaboration system according to claim 1, characterized in that: The application scenarios of the multi-port agent are respectively life assistance, interactive learning, medical consultation, tour guide, interactive live broadcast, and video marketing. The interactive live broadcast includes dynamic script generation, TTS voice synthesis, AI intelligent field control, distributed TTS voice live broadcast, and AI intelligent review.

7. An interactive intelligent agent collaboration system according to claim 1, wherein: The video marketing includes AI scripts, one-click video creation, automatic distribution, matrix building, and intelligent customer service. The multi-port agent conducts data interaction by creating virtual digital humans. The virtual digital humans include virtual hosts, virtual IP spokespersons, virtual human video broadcasts, and virtual human customer services.

8. A control method for an interactive agent collaboration system, characterized in that, Including the following steps: S1: Select the hardware module of the interactive agent according to the application scenario of the interactive agent. The hardware module of the interactive agent includes an interactive display, an M-Creation workstation, an M-Creation multimodal box, and an interactive robot. Embed the software module of the interactive agent, that is, the M-Creation engine, into the hardware module, and select the hardware module according to the application scenario; S2: The application scenarios of the interactive agent include but are not limited to life assistance, interactive learning, medical consultation, tour guide, interactive live broadcast, and video marketing. Apply the core of the M-Creation engine, that is, the multi-port agent, so that the multi-port agent extracts, transforms, and loads the interactive data using the data set and transmits it to the process orchestration platform. Through the data set, the interactive data can be processed, reducing the operating pressure of the process orchestration platform; S3: According to different application scenarios, the process orchestration platform controls the output of the processed interaction data through the plug-in toolset and the large language model to meet the query requests in different interaction processes. At the same time, the multi-port intelligent agent conducts data interaction by establishing virtual digital humans, and through the virtual digital humans, outbound actions in multiple scenarios can be realized, including virtual hosts, virtual IP spokespersons, virtual human video broadcasts, and virtual human customer service, reducing the labor intensity of users; S4: The M-Creation engine trained through multiple scenario interactions can integrate and optimize various interaction data. At the same time, through a large number of emotional cognition trainings, it is convenient for the multi-port intelligent agent to customize preset role identities and emotional setting operations during the interaction process, and through the interaction input of data such as voice, images, and texts, it can accurately capture user needs and automatically adapt to respond with emotions and tones.

Images