system

Abstract

We provide the system. [Solution] A means of capturing user speech data, A means for analyzing the aforementioned speech data to extract the user's conversation patterns, A means comprising a generative model that generates response candidates based on the user's current contextual information, A means of presenting the generated response candidates to the user, A means for outputting the aforementioned response candidates as audio based on the user's selection, A system that includes this.

Description

Technical Field

[0001] The technology of the present disclosure relates to a system.

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, and includes steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] People who have difficulty speaking in specific situations are often hindered from smooth communication in daily life. This communication barrier becomes a factor of obstacles and stress in the social life of the parties concerned, and there is also a possibility of social isolation. The present invention aims to solve the problem of realizing smooth communication by automatically generating an appropriate response as an alternative means of speaking and supporting the natural conversation of users in such a situation.

Means for Solving the Problems

[0005] This invention provides a system that captures and analyzes user speech data to extract user conversation patterns and generates response candidates using a generative model based on these patterns. This system includes means for presenting multiple generated response candidates to the user and outputting the response as audio based on the user's selection, thereby enabling users to communicate naturally even in situations where speaking is difficult.

[0006] "User" refers to an individual who uses the system to receive assistance with speech in specific situations.

[0007] "Speech data" refers to the record of conversations conducted by users in voice or text format, and encompasses all information captured by the system.

[0008] "Analysis" refers to the process of processing speech data captured by the system to extract user conversation patterns.

[0009] "Conversation patterns" refer to characteristics that indicate the choice of words a user typically uses, the frequency of their utterances, and the flow of conversation according to the context.

[0010] "Contextual information" refers to information about the situation and environment in which a user makes a speech, including information such as the time, place, and the person they are speaking to.

[0011] "Response candidates" refer to multiple possible responses that the system generates based on utterance data and contextual information.

[0012] A "generative model" refers to an algorithm or program within a system used to automatically generate response candidates from user speech data.

[0013] "Voice output" refers to the process of converting selected response candidates into synthesized speech and communicating them to those around you. [Brief explanation of the drawing]

[0014] [Figure 1]It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which multiple emotions are mapped. [Figure 10] It shows an emotion map to which multiple emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

MODE FOR CARRYING OUT THE INVENTION

[0015] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0018] In the following embodiments, the numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0019] In the following embodiments, the numbered storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0020] In the following embodiments, the numbered communication I / F (Interface) is an interface including a communication processor and an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark), and the like.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0022] [First Embodiment]

[0023] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0024] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0025] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0027] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0030] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0034] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0035] The communication assistance system according to the present invention is designed to enable users with selective mutism to engage in smooth conversations in various situations. This system operates by combining speech analysis technology, natural language processing (NLP), generative AI models, and speech synthesis technology.

[0036] When a user uses a device, the system first captures the user's speech data. This speech data is recorded in text or audio format and securely stored on the device. The collected data is then encrypted and transmitted to a server.

[0037] The server analyzes the received data and processes it to identify the user's conversation patterns. Specifically, the server utilizes natural language processing techniques to extract frequently occurring themes and the user's speaking tendencies in specific contexts. Based on this information, the server updates the user's profile and uses it to understand their individual conversation style.

[0038] Based on the analysis results, the server uses a generative AI model to generate response candidates for various conversation scenarios. This prepares responses that enable users to speak smoothly in specific situations. The response candidates are optimized based on the user's preferences and habits, and multiple options are presented.

[0039] The device presents the user with a list of generated response options. The user selects the response that best reflects their intention from the displayed list. The selected response is converted into natural-sounding speech using the device's built-in speech synthesis function and output in the user's voice. This process effectively communicates the user's intentions to others.

[0040] As a concrete example, consider a scenario where a user orders at a cafe. Based on past data, the server predicts the drinks the user frequently orders and generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please" from the presented options, and the terminal synthesizes the selection into speech and conveys it to the staff.

[0041] Thus, the system of the present invention provides support to enable users of selective mutism to express their intentions appropriately in various social situations and has the effect of reducing communication barriers.

[0042] The following describes the processing flow.

[0043] Step 1:

[0044] The device captures speech data through voice input or text messages during everyday conversations. This data is processed in the background so as not to interrupt the user's actions.

[0045] Step 2:

[0046] The device temporarily stores the captured speech data in local storage and, when certain conditions are met (for example, when connected to Wi-Fi), encrypts the data and sends it to the server.

[0047] Step 3:

[0048] The server receives encrypted data sent from the terminal and stores it in a database. This data is then used as reference data for subsequent analysis processes.

[0049] Step 4:

[0050] The server analyzes the received speech data using natural language processing technology. Specifically, it extracts frequently occurring words and phrases, identifies patterns in the user's topics and the flow of conversation, and updates the user's profile accordingly.

[0051] Step 5:

[0052] The server uses a generative AI model based on the analyzed data to generate response candidates to support user communication. These candidates are optimized based on the user's past conversation history and contextual information.

[0053] Step 6:

[0054] The terminal presents the user with a list of response options sent from the server. The user interface is intuitively designed, allowing users to easily make selections.

[0055] Step 7:

[0056] The user selects the response that best reflects their intention from the presented options. This selection is typically made using touch controls or a visual interface.

[0057] Step 8:

[0058] The device uses its speech synthesis function to convert the user's selected response into speech, outputting it as a natural-sounding voice. This ensures that the user's intentions are communicated to others.

[0059] Step 9:

[0060] Users can provide feedback to the device on whether the response was as intended. This feedback can be used to improve the quality of future response generation.

[0061] This processing flow will enable users to express their intentions smoothly even in difficult situations.

[0062] (Example 1)

[0063] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0064] It is crucial to enable individuals with selective mutism to communicate effectively in specific situations. However, current technology does not adequately provide speech support tailored to each individual's characteristics. As a result, individuals face challenges in naturally conveying their intentions and having their opportunities for social participation limited.

[0065] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0066] In this invention, the server includes a device for acquiring speech information, a device for analyzing the speech information to extract the user's dialogue tendencies, and a device for generating response candidates using a generative artificial intelligence model based on the analysis results. This makes it possible to generate responses optimized for the individual characteristics of the user and support smooth communication in specific situations.

[0067] "Speech information" refers to the content of communications made by users in text or voice format.

[0068] A "device for acquiring data" is a device that has the function of detecting and recording voice and text data.

[0069] "Dialogue tendencies" refer to patterns in speaking style and content that users have shown in past communications.

[0070] A "speech analyzer" is a device that has the function of analyzing the user's dialogue tendencies from the acquired speech information.

[0071] A "generative artificial intelligence model" is a model that uses artificial intelligence technology to generate responses based on the user's speech.

[0072] "Response candidates" refer to multiple response options that are generated based on a specific situation or context and presented to the user.

[0073] A "speech conversion device" is a device that has the function of converting selected response candidates into a natural speech format and outputting it.

[0074] This invention is a system that supports users with selective mutism in communicating smoothly in a variety of situations. Specific embodiments of this system are described below.

[0075] Users use a dedicated terminal to acquire their own speech information in real time. The terminal has a microphone and keyboard as input devices, and the speech information is recorded in voice and text format. The terminal encrypts the recorded information and then sends it to the server via a secure protocol.

[0076] The server analyzes the received speech information using natural language processing techniques. These techniques include morphological analysis and contextual understanding. Through this analysis, the server extracts the user's conversational tendencies and main themes, and updates the user profile based on this information.

[0077] The generation AI model on the server considers the analysis results and user profile to generate a variety of response candidates. Here, the generation AI model improves the appropriateness of the responses by also referring to the user's past preference history. An example of a prompt is "Generate drink candidates that the user might like."

[0078] The terminal presents the user with a list of response options sent from the server. The user selects the most suitable response from the options, and the terminal converts the selected response into natural-sounding speech using speech synthesis technology. This speech is then output as the user's voice, allowing them to clearly communicate their intentions.

[0079] As a concrete example, consider ordering a drink at a cafe. Based on past data of drinks the user frequently orders, the server generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please," and the terminal announces this verbally.

[0080] In this way, the system plays a role in promoting social interaction by providing communication support tailored to the individual needs of users.

[0081] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0082] Step 1:

[0083] The user activates the device and inputs spoken information via voice or text. The device converts the input voice into text data using speech recognition technology. The input voice data and the converted text data are encrypted to ensure security and recorded. As a result, the spoken information is output in both text and voice formats.

[0084] Step 2:

[0085] The terminal sends encrypted speech information to the server. This transmission uses a secure communication protocol (e.g., HTTPS). The input consists of speech and text data, which are then transferred to the server.

[0086] Step 3:

[0087] The server analyzes the received speech information using natural language processing techniques. This analysis involves morphological analysis and contextual understanding to identify the user's dialogue tendencies and themes. The input is text data, and the output is data on dialogue tendencies and related themes.

[0088] Step 4:

[0089] Based on the analysis results, the server generates response candidates using a generative artificial intelligence model. This process uses prompts to input analysis data into the AI ​​model, generating a variety of response options. For example, a prompt like "Generate drink candidates that the user would like" might be used. The output is a list of response candidates.

[0090] Step 5:

[0091] The terminal presents the user with a list of response options sent from the server. The user reviews the options displayed on the screen and selects the most appropriate response. Input for selection can be via touch or voice commands.

[0092] Step 6:

[0093] The device uses speech synthesis technology to convert the user's selected response into speech. This process transforms text data into natural-sounding speech data. The output is the actual audio that is played back. This ensures that the user's intentions are communicated in a way that is easily understood by others.

[0094] (Application Example 1)

[0095] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0096] There is a need for support tools that enable users with communication difficulties to communicate smoothly in public places and face-to-face interactions. Current technology fails to fully utilize users' individual contextual information and preference history, making it difficult to provide timely and optimal responses to facilitate the flow of conversation. In particular, there is a high need for technology that effectively and naturally assists speech for individuals with selective mutism or other communication disorders.

[0097] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0098] In this invention, the server includes means for capturing user speech information, means for analyzing the speech information to extract conversational tendencies, and means for a generative model that generates response candidates based on current situation information and preference history. This makes it possible for users to be presented with responses effectively and quickly in face-to-face interactions, and for those responses to be communicated in a natural manner using speech synthesis technology.

[0099] "User" refers to an individual who receives communication support using this system.

[0100] "Speech information" refers to the content of a conversation expressed by the user in written or spoken form.

[0101] "Means of capture" refers to components that have the function of acquiring and storing speech information.

[0102] "Means of analysis" refers to techniques that analyze acquired speech information and extract the user's conversational tendencies and characteristics.

[0103] "Conversational tendencies" refer to the linguistic characteristics and speech patterns that users exhibit during conversations.

[0104] "Preference history" refers to data about responses and preferences that a user has selected in the past.

[0105] A "generative model" is a module that uses artificial intelligence technology to generate response candidates based on the user's current situation information and preference history.

[0106] A "response option" refers to a set of utterances presented to the user as choices they can use in a specific situation.

[0107] "Speech synthesis" is a technology that converts text information into natural-sounding speech.

[0108] "Means of communication to others" refers to technologies for effectively conveying synthesized speech responses to people other than the user.

[0109] This invention is a communication assistance system aimed at enabling users with selective mutism or other communication difficulties to communicate smoothly in face-to-face interactions. The system is implemented via a smartphone or similar user terminal.

[0110] The system's core process combines the acquisition, analysis, and generation of speech information, followed by speech synthesis output. Users input speech information via a terminal in either voice or text format. The terminal converts the voice data to text using speech recognition technology, encrypts the data, and then sends it to the server. Speech recognition technologies such as Google® Cloud Speech-to-Text and other similar services are available.

[0111] The server analyzes the received data using natural language processing (NLP) techniques to extract the user's conversational tendencies. NLP tools such as NLTK and SpaCy are used here. Then, using a generative AI model such as GPT-4 (registered trademark), appropriate response candidates are generated based on the user's past preference history and current situation. The generated responses are optimized based on important topics and the user's preferences and presented as multiple options.

[0112] On the device, generated response suggestions are displayed to the user. The response selected by the user is converted into natural-sounding speech using speech synthesis technology such as Google Text-to-Speech and communicated to the other party. This process facilitates smoother communication.

[0113] As a concrete example of use, consider a scenario where a user orders tea at a cafe. Based on the user's past purchase history, the system recommends a type of tea and generates a response such as, "I'd like Darjeeling tea, please." Once the user selects this, the terminal verbally communicates this to the staff.

[0114] As an example of a prompt, you can instruct the AI ​​as follows: "Generate a list of recommended teas based on the user's preferences and past purchase history. Please suggest three options, taking into account current preferences and trends."

[0115] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0116] Step 1:

[0117] The user inputs speech information into the device. The input format can be voice or text. In the case of voice input, the device uses Google Cloud Speech-to-Text to convert the voice data into text format. The input data is then sent to the next step either in its original format or as text.

[0118] Step 2:

[0119] The terminal encrypts the input data to store it securely. AES encryption technology is used here. This process protects user privacy while preparing the data for transmission to the next step. The encrypted data is output and sent to the server.

[0120] Step 3:

[0121] The server decrypts the received encrypted data and analyzes its contents. It uses natural language processing techniques to analyze the text data and extract user conversational trends and important topics. By utilizing tools such as NLTK and SpaCy, it outputs trend data from the input text.

[0122] Step 4:

[0123] The server uses a generative AI model, such as GPT-4, to generate response candidates that take into account the user's context and preference history. The generative AI model generates an appropriate response based on past data and the current prompt, and a list of response candidates is output.

[0124] Step 5:

[0125] The device presents the user with the received response options. The user selects the response that matches their intent. The user's selection is output, and the result is passed on to the next step.

[0126] Step 6:

[0127] The selected response is converted into natural-sounding speech using speech synthesis technology on the device. Text data is converted to speech data using tools such as Google Text-to-Speech. The converted speech is played through the device's speaker and transmitted to the other party.

[0128] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0129] The communication assistance system according to the present invention is designed to facilitate smooth communication in situations where it is difficult for the user to speak. This system enables more natural and appropriate communication by combining an emotion engine at each stage of speech data capture, analysis, response generation, and voice output.

[0130] First, as users use their devices, speech data is continuously captured. This data is entered in either text or voice format, and the device collects it. The collected data is encrypted and sent to the server.

[0131] The server analyzes the received speech data to identify the user's conversation patterns. Natural language processing techniques are used in this analysis to extract the user's everyday speaking tendencies and frequently occurring words. Simultaneously, an emotion engine recognizes emotions from the user's speech and selected words, and evaluates their emotional state. This information is also added to the user profile maintained by the server and learned as a comprehensive characteristic of the user.

[0132] Based on this data, the server generates response candidates using a generative AI model. During this process, the tone and content of the responses are adjusted to match the user's current emotional state, taking into account the user's emotional condition. For example, if the server determines that the user is stressed, responses that promote relaxation are prioritized.

[0133] The device presents the user with a list of generated response options. The user selects the response that best suits them through an intuitive interface. The selected response is then converted into speech, tailored to the context and emotions, and output in the user's natural voice. By using speech synthesis technology, the user's voice tone and emotions are also reflected, resulting in a more realistic conversation.

[0134] As a concrete example, consider a scenario where a user is giving a presentation in a public place. If the user shows signs of nervousness, the server can detect this using its emotion engine and offer a response such as, "Relax, you're doing great." If the user selects this option, the device will generate a voice in a comforting tone to support the user.

[0135] This invention is a system that supports users in maintaining the most appropriate and effective communication by performing comprehensive information processing, including emotions. This technology is useful not only for people with selective mutism but also for all people who require special communication support.

[0136] The following describes the processing flow.

[0137] Step 1:

[0138] The device captures voice or text data as soon as the user begins a conversation. This includes what the user says to the device and messages they type on the screen.

[0139] Step 2:

[0140] The device temporarily stores the collected speech data, and once it confirms that the communication environment is stable, it encrypts the data and sends it to the server. This data also includes emotional information based on the user's consent.

[0141] Step 3:

[0142] The server converts the received data into a format suitable for analysis, analyzes the conversation content using natural language processing techniques, and extracts the user's conversation patterns. This analysis includes frequently occurring words, the flow of the conversation, and the user's preferences.

[0143] Step 4:

[0144] The server uses an emotion engine to recognize the user's emotional state from their voice tone and selected words. For example, it analyzes the tone and tempo of the voice data to determine whether the user is tense or relaxed.

[0145] Step 5:

[0146] The server uses a generative AI model to generate response candidates, taking into account the user's conversation patterns and emotional state. This generation process utilizes the user's past conversation history and current contextual information to prepare multiple appropriate responses.

[0147] Step 6:

[0148] The device displays generated response options to the user. The UI is designed to be intuitive and easy for the user to operate, and sometimes a response that helps to ease tension is displayed at the top.

[0149] Step 7:

[0150] The user selects a response from the presented options that best matches their intentions and feelings. The selected response is optimized to suit the individual's emotions and situation.

[0151] Step 8:

[0152] The device uses a speech synthesis engine to output the selected response as speech that reflects the user's voice and tone. This makes it sound to others as if the user were speaking naturally.

[0153] Step 9:

[0154] Users can evaluate whether the outputted response was intended in the communication and provide feedback to the system.

[0155] This allows the system to use this feedback as an indicator to further improve the accuracy of future responses.

[0156] (Example 2)

[0157] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0158] In modern society, effective communication is required in a variety of situations, but users sometimes have difficulty conveying their intentions appropriately. In such situations, there is a need for technologies that support smooth dialogue tailored to the user's mental state and linguistic characteristics. In particular, responses that respond to emotional states are important, and there is a challenge in that support in this area is insufficient.

[0159] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0160] In this invention, the server includes a device for capturing user speech information, a device for analyzing the speech information to extract the user's language tendencies, and a generative AI mechanism for analyzing the user's emotional state and generating response candidates based on the language tendencies and emotional state. This enables appropriate and effective communication support that takes emotions into consideration.

[0161] "User speech information" refers to the content that users utter in voice or text format, and this information is collected as data.

[0162] "Linguistic tendencies" refer to specific language usage patterns and characteristics of frequently occurring words and phrases observed in users' conversations.

[0163] "Emotional state" refers to the psychological or emotional condition analyzed from the user's utterances and selected words.

[0164] A "generative AI mechanism" is a system that incorporates artificial intelligence technology used to generate appropriate responses based on the user's language tendencies and emotional state.

[0165] A "communication support system" refers to a device or program that provides a technical mechanism to assist users in communicating their intentions smoothly and appropriately.

[0166] This invention is a communication support system designed to provide more natural and appropriate responses when users communicate. The system aims to efficiently analyze speech information while considering the user's emotional state and generate appropriate responses.

[0167] First, the content spoken by the user, either by voice or text, is captured by the device. Here, voice processing software, commonly used as speech recognition technology, is used to convert the voice data into text. The text data is temporarily stored on the device and then transmitted to the server using encryption technology to ensure security.

[0168] The server utilizes natural language processing techniques to analyze the received text data. Here, general analysis software is used to extract the user's language tendencies. Simultaneously, a sentiment analysis engine is employed to evaluate the user's emotional state. These evaluation results are stored as individual user profiles.

[0169] Based on the analysis results and emotional state, the server generates response candidates using a generative AI model. The generative AI model enables natural dialogue that matches the user's psychological state. An example of a prompt would be, "Generate a natural, relaxing response for when the user is nervous during a presentation."

[0170] The generated responses are presented to the user by the device, and the user can select one through an intuitive interface. The selected response is then output naturally in the user's voice using speech synthesis technology. This enables the user to engage in smooth, emotionally sensitive communication.

[0171] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0172] Step 1:

[0173] The device receives voice or text input from the user. In the case of voice input, the device uses speech recognition software to convert the voice data into text. The resulting text data is then output. The speech recognition process includes an acoustic model and a language model for converting voice samples into text.

[0174] Step 2:

[0175] The terminal encrypts the generated text data and sends it to the server. It receives text data as input, protects it using an encryption algorithm, and outputs encrypted data. This ensures confidentiality during data transfer.

[0176] Step 3:

[0177] The server receives encrypted text data, decrypts it, and returns it to its original text data. This decrypted data becomes the input, and raw text ready for analysis is output.

[0178] Step 4:

[0179] The server analyzes the decoded text data using natural language processing techniques. This extracts the user's language tendencies and the intent expressed in the text. The input for this step is raw text, and the output is the analyzed language features and conversational intent.

[0180] Step 5:

[0181] The server uses a sentiment analysis engine to evaluate the user's emotional state. The input is pre-analyzed text, and the output is the detected emotional state and psychological tendencies. This allows for the clear identification of subtle emotions hidden within the text.

[0182] Step 6:

[0183] The server uses a generative AI model to generate response candidates based on the analysis results and emotional state. The generative AI model takes linguistic features and emotional state as input and generates emotionally sensitive response candidates as output. This process creates natural-sounding utterances that are appropriate to the user's current psychological state.

[0184] Step 7:

[0185] The terminal presents the user with generated response options. This can be done by displaying the response on the screen or by providing voice guidance. The input here is response data from the server, which is output as choices for the user to select.

[0186] Step 8:

[0187] The user selects the most appropriate response from the presented options. The selected response is recognized as input, and the next action is planned accordingly.

[0188] Step 9:

[0189] The device outputs the selected response in speech format using speech synthesis technology. The input here is the selected text response, and the output is synthesized speech with a tone similar to the user's voice. This enables more natural communication.

[0190] (Application Example 2)

[0191] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0192] Conventional communication assistance systems have the problem of being unable to conduct conversations that fully consider the user's emotions, and thus are unable to communicate naturally and smoothly. Furthermore, in customer service at physical stores, it is necessary to appropriately understand the customer's emotions and respond accordingly, but there has been no effective technology to support this.

[0193] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0194] In this invention, the server includes means for capturing user speech data, means for analyzing the speech data to extract the user's conversation patterns, and means for a generative model that generates response candidates based on the user's current contextual information and emotional state. This makes it possible to appropriately grasp the customer's emotions and support optimal communication in interactions with customers in physical stores.

[0195] "User speech data" refers to the audio and text information that users produce when communicating.

[0196] "Conversation patterns" refer to the collective characteristics and tendencies extracted from conversations a user has had in the past.

[0197] "Contextual information" refers to information that includes the situation and background in which an utterance is made, as well as the emotional state of the speaker.

[0198] A "generative model" refers to a structure or algorithm used to generate a response based on input data, employing natural language processing techniques.

[0199] A "display device" refers to an electronic device used to visually present generated information to a user.

[0200] "Speech synthesis" is a technology that mechanically generates human speech from text.

[0201] "Emotional state" refers to the psychological state that can be inferred from the user's statements and actions.

[0202] To implement this invention, a terminal device, such as smart glasses worn by the user, and a server connected to it are required. The terminal device is equipped with a camera and microphone, and has the function to transmit real-time data acquired by these devices to the server. The server receives the user's speech data and facial expression data and uses natural language processing technology and facial recognition technology to analyze them. Specifically, it is common to use Google Cloud Speech-to-Text API or OpenCV.

[0203] The server first converts the acquired audio data into text, and then extracts the user's conversation patterns and emotional state from the text data and facial expression data. Based on this information, a generative AI model generates the optimal response. The generated response is then presented to the terminal, with its tone and selected language adjusted to match the user's emotional state. The terminal visually communicates this response to the user and provides audio feedback as needed.

[0204] For example, in a physical store, if a customer's anxious expression or speech is detected during customer service, a suggestion such as "Please let us know if you need any assistance" will be displayed on the staff member's smart glasses. This allows staff members to provide more appropriate and effective customer service.

[0205] A concrete example of a prompt for a generative AI model is: "Generate the most appropriate response when the customer looks anxious. Choose the right words based on the user's emotional state." This prompt allows the system to provide a response appropriate to the emotional state, improving the quality of communication.

[0206] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0207] Step 1:

[0208] The device captures the user's speech and facial expressions in real time using its camera and microphone. The input is the user's voice and video data, and the output is the transmission of this data to a server. Specifically, sensors within the device record the voice, and the camera captures the facial expressions.

[0209] Step 2:

[0210] The server converts the received audio data into text. This process uses the Google Cloud Speech-to-Text API. The input is audio data, and the output is text data. The operation here is to process the audio signal as a text string.

[0211] Step 3:

[0212] The server analyzes facial expression data to infer the user's emotions. It uses OpenCV to detect facial features and then analyzes them with an emotion engine. The input is facial image data, and the output is the inferred emotional state. Specifically, it measures changes and patterns in facial expressions and assigns emotion labels.

[0213] Step 4:

[0214] The server uses natural language processing technology to extract conversation patterns from text data. A generative AI model then generates response candidates based on this. The input is text data and emotion states, and the output is the text of the response candidates. In operation, prompt sentences are sent to the generative AI model, which automatically generates responses.

[0215] Step 5:

[0216] The terminal presents the user with generated response options visually or audibly. The user selects the most appropriate response, which is then output via speech synthesis. The input is the text data of the response options, and the output is either an audio message or text on a display. The operation here is to display the response on a visual device or play the audio through a speaker.

[0217] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0218] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0219] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0220] [Second Embodiment]

[0221] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0222] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0223] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0224] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0225] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0226] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0227] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0228] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0229] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0230] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0231] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0232] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0233] The communication assistance system according to the present invention is designed to enable users with selective mutism to engage in smooth conversations in various situations. This system operates by combining speech analysis technology, natural language processing (NLP), generative AI models, and speech synthesis technology.

[0234] When a user uses a device, the system first captures the user's speech data. This speech data is recorded in text or audio format and securely stored on the device. The collected data is then encrypted and transmitted to a server.

[0235] The server analyzes the received data and processes it to identify the user's conversation patterns. Specifically, the server utilizes natural language processing techniques to extract frequently occurring themes and the user's speaking tendencies in specific contexts. Based on this information, the server updates the user's profile and uses it to understand their individual conversation style.

[0236] Based on the analysis results, the server uses a generative AI model to generate response candidates for various conversation scenarios. This prepares responses that enable users to speak smoothly in specific situations. The response candidates are optimized based on the user's preferences and habits, and multiple options are presented.

[0237] The device presents the user with a list of generated response options. The user selects the response that best reflects their intention from the displayed list. The selected response is converted into natural-sounding speech using the device's built-in speech synthesis function and output in the user's voice. This process effectively communicates the user's intentions to others.

[0238] As a concrete example, consider a scenario where a user orders at a cafe. Based on past data, the server predicts the drinks the user frequently orders and generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please" from the presented options, and the terminal synthesizes the selection into speech and conveys it to the staff.

[0239] Thus, the system of the present invention provides support to enable users of selective mutism to express their intentions appropriately in various social situations and has the effect of reducing communication barriers.

[0240] The following describes the processing flow.

[0241] Step 1:

[0242] The device captures speech data through voice input or text messages during everyday conversations. This data is processed in the background so as not to interrupt the user's actions.

[0243] Step 2:

[0244] The device temporarily stores the captured speech data in local storage and, when certain conditions are met (for example, when connected to Wi-Fi), encrypts the data and sends it to the server.

[0245] Step 3:

[0246] The server receives encrypted data sent from the terminal and stores it in a database. This data is then used as reference data for subsequent analysis processes.

[0247] Step 4:

[0248] The server analyzes the received speech data using natural language processing technology. Specifically, it extracts frequently occurring words and phrases, identifies patterns in the user's topics and the flow of conversation, and updates the user's profile accordingly.

[0249] Step 5:

[0250] The server uses a generative AI model based on the analyzed data to generate response candidates to support user communication. These candidates are optimized based on the user's past conversation history and contextual information.

[0251] Step 6:

[0252] The terminal presents the user with a list of response options sent from the server. The user interface is intuitively designed, allowing users to easily make selections.

[0253] Step 7:

[0254] The user selects the response that best reflects their intention from the presented options. This selection is typically made using touch controls or a visual interface.

[0255] Step 8:

[0256] The device uses its speech synthesis function to convert the user's selected response into speech, outputting it as a natural-sounding voice. This ensures that the user's intentions are communicated to others.

[0257] Step 9:

[0258] Users can provide feedback to the device on whether the response was as intended. This feedback can be used to improve the quality of future response generation.

[0259] This processing flow will enable users to express their intentions smoothly even in difficult situations.

[0260] (Example 1)

[0261] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0262] It is crucial to enable individuals with selective mutism to communicate effectively in specific situations. However, current technology does not adequately provide speech support tailored to each individual's characteristics. As a result, individuals face challenges in naturally conveying their intentions and having their opportunities for social participation limited.

[0263] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0264] In this invention, the server includes a device for acquiring speech information, a device for analyzing the speech information to extract the user's dialogue tendencies, and a device for generating response candidates using a generative artificial intelligence model based on the analysis results. This makes it possible to generate responses optimized for the individual characteristics of the user and support smooth communication in specific situations.

[0265] "Speech information" refers to the content of communications made by users in text or voice format.

[0266] A "device for acquiring data" is a device that has the function of detecting and recording voice and text data.

[0267] "Dialogue tendencies" refer to patterns in speaking style and content that users have shown in past communications.

[0268] A "speech analyzer" is a device that has the function of analyzing the user's dialogue tendencies from the acquired speech information.

[0269] A "generative artificial intelligence model" is a model that uses artificial intelligence technology to generate responses based on the user's speech.

[0270] "Response candidates" refer to multiple response options that are generated based on a specific situation or context and presented to the user.

[0271] A "speech conversion device" is a device that has the function of converting selected response candidates into a natural speech format and outputting it.

[0272] This invention is a system that supports users with selective mutism in communicating smoothly in a variety of situations. Specific embodiments of this system are described below.

[0273] Users use a dedicated terminal to acquire their own speech information in real time. The terminal has a microphone and keyboard as input devices, and the speech information is recorded in voice and text format. The terminal encrypts the recorded information and then sends it to the server via a secure protocol.

[0274] The server analyzes the received speech information using natural language processing techniques. These techniques include morphological analysis and contextual understanding. Through this analysis, the server extracts the user's conversational tendencies and main themes, and updates the user profile based on this information.

[0275] The generation AI model on the server considers the analysis results and user profile to generate a variety of response candidates. Here, the generation AI model improves the appropriateness of the responses by also referring to the user's past preference history. An example of a prompt is "Generate drink candidates that the user might like."

[0276] The terminal presents the user with a list of response options sent from the server. The user selects the most suitable response from the options, and the terminal converts the selected response into natural-sounding speech using speech synthesis technology. This speech is then output as the user's voice, allowing them to clearly communicate their intentions.

[0277] As a concrete example, consider ordering a drink at a cafe. Based on past data of drinks the user frequently orders, the server generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please," and the terminal announces this verbally.

[0278] In this way, the system plays a role in promoting social interaction by providing communication support tailored to the individual needs of users.

[0279] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0280] Step 1:

[0281] The user starts the terminal and inputs speech information either by voice or text. The terminal converts the input voice into text data using voice recognition technology. The input voice data and the converted text data are encrypted to ensure security, and these are recorded. As a result, the speech information is output in both text and voice formats.

[0282] Step 2:

[0283] The terminal sends the encrypted speech information to the server. The transmission is carried out using a secure communication protocol (e.g., HTTPS). There is voice and text data as input, and this is transferred to the server.

[0284] Step 3:

[0285] The server analyzes the received speech information using natural language processing technology. In this analysis operation, morphological analysis and context understanding are carried out, and the user's conversation tendency and theme are identified. The input is text data, and the output is data on the conversation tendency and related themes.

[0286] Step 4:

[0287] Based on the analysis results, the server uses a generated artificial intelligence model to generate response candidates. In this operation, a prompt sentence is used to input the analysis data into the AI model, and various response options are generated. For example, a prompt such as "Generate candidates for the user's favorite drinks" is used. The output is a list of response candidates.

[0288] Step 5:

[0289] The terminal presents the response candidates sent from the server to the user. The user checks the candidates displayed on the screen and selects the response that seems to be the most appropriate from them. The input for selection is a touch operation or a voice command.

[0290] Step 6:

[0291] The device uses speech synthesis technology to convert the user's selected response into speech. This process transforms text data into natural-sounding speech data. The output is the actual audio that is played back. This ensures that the user's intentions are communicated in a way that is easily understood by others.

[0292] (Application Example 1)

[0293] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0294] There is a need for support tools that enable users with communication difficulties to communicate smoothly in public places and face-to-face interactions. Current technology fails to fully utilize users' individual contextual information and preference history, making it difficult to provide timely and optimal responses to facilitate the flow of conversation. In particular, there is a high need for technology that effectively and naturally assists speech for individuals with selective mutism or other communication disorders.

[0295] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0296] In this invention, the server includes means for capturing user speech information, means for analyzing the speech information to extract conversational tendencies, and means for a generative model that generates response candidates based on current situation information and preference history. This makes it possible for users to be presented with responses effectively and quickly in face-to-face interactions, and for those responses to be communicated in a natural manner using speech synthesis technology.

[0297] "User" refers to an individual who receives communication support using this system.

[0298] "Speech information" refers to the content of a conversation expressed by the user in written or spoken form.

[0299] "The means for capturing" refers to a component having a function for acquiring and storing utterance information.

[0300] "The means for analyzing" refers to a technique for analyzing the acquired utterance information and extracting the conversation tendency and characteristics of the user.

[0301] "Conversation tendency" indicates the linguistic features and utterance patterns shown by the user in the conversation.

[0302] "Preference history" refers to data regarding the responses and preferences selected by the user in the past.

[0303] "Generation model" is a module using artificial intelligence technology for generating response candidates based on the current situation information and preference history of the user.

[0304] "Response candidate" refers to an utterance presented as a series of options that can be used by the user in a specific situation.

[0305] "Voice synthesis" is a technology for converting text information into natural voice.

[0306] "The means for transmitting to others" is a technology for effectively conveying the voice-synthesized response to people other than the user.

[0307] The present invention is a communication assistance system aimed at enabling users who have difficulties in scene silence and other communications to smoothly convey their intentions in face-to-face interactions. The system is implemented via a smartphone or a similar user terminal.

[0308] The system's core process combines the acquisition, analysis, and generation of speech information, followed by speech synthesis output. Users input speech information via a terminal in either voice or text format. The terminal converts the voice data to text using speech recognition technology, encrypts the data, and then sends it to the server. Speech recognition technologies such as Google Cloud Speech-to-Text and other similar services are available.

[0309] The server analyzes the received data using natural language processing (NLP) techniques to extract the user's conversational tendencies. NLP tools such as NLTK and SpaCy are used here. Then, using a generative AI model such as GPT-4, appropriate response candidates are generated based on the user's past preference history and current situation. The generated responses are optimized based on important topics and the user's preferences and presented as multiple options.

[0310] On the device, generated response suggestions are displayed to the user. The response selected by the user is converted into natural-sounding speech using speech synthesis technology such as Google Text-to-Speech and communicated to the other party. This process facilitates smoother communication.

[0311] As a concrete example of use, consider a scenario where a user orders tea at a cafe. Based on the user's past purchase history, the system recommends a type of tea and generates a response such as, "I'd like Darjeeling tea, please." Once the user selects this, the terminal verbally communicates this to the staff.

[0312] As an example of a prompt, you can instruct the AI ​​as follows: "Generate a list of recommended teas based on the user's preferences and past purchase history. Please suggest three options, taking into account current preferences and trends."

[0313] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0314] Step 1:

[0315] The user inputs speech information into the device. The input format can be voice or text. In the case of voice input, the device uses Google Cloud Speech-to-Text to convert the voice data into text format. The input data is then sent to the next step either in its original format or as text.

[0316] Step 2:

[0317] The terminal encrypts the input data to store it securely. AES encryption technology is used here. This process protects user privacy while preparing the data for transmission to the next step. The encrypted data is output and sent to the server.

[0318] Step 3:

[0319] The server decrypts the received encrypted data and analyzes its contents. It uses natural language processing techniques to analyze the text data and extract user conversational trends and important topics. By utilizing tools such as NLTK and SpaCy, it outputs trend data from the input text.

[0320] Step 4:

[0321] The server uses a generative AI model, such as GPT-4, to generate response candidates that take into account the user's context and preference history. The generative AI model generates an appropriate response based on past data and the current prompt, and a list of response candidates is output.

[0322] Step 5:

[0323] The device presents the user with the received response options. The user selects the response that matches their intent. The user's selection is output, and the result is passed on to the next step.

[0324] Step 6:

[0325] The selected response is converted into natural-sounding speech using speech synthesis technology on the device. Text data is converted to speech data using tools such as Google Text-to-Speech. The converted speech is played through the device's speaker and transmitted to the other party.

[0326] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0327] The communication assistance system according to the present invention is designed to facilitate smooth communication in situations where it is difficult for the user to speak. This system enables more natural and appropriate communication by combining an emotion engine at each stage of speech data capture, analysis, response generation, and voice output.

[0328] First, as users use their devices, speech data is continuously captured. This data is entered in either text or voice format, and the device collects it. The collected data is encrypted and sent to the server.

[0329] The server analyzes the received speech data to identify the user's conversation patterns. Natural language processing techniques are used in this analysis to extract the user's everyday speaking tendencies and frequently occurring words. Simultaneously, an emotion engine recognizes emotions from the user's speech and selected words, and evaluates their emotional state. This information is also added to the user profile maintained by the server and learned as a comprehensive characteristic of the user.

[0330] Based on this data, the server generates response candidates using a generative AI model. During this process, the tone and content of the responses are adjusted to match the user's current emotional state, taking into account the user's emotional condition. For example, if the server determines that the user is stressed, responses that promote relaxation are prioritized.

[0331] The device presents the user with a list of generated response options. The user selects the response that best suits them through an intuitive interface. The selected response is then converted into speech, tailored to the context and emotions, and output in the user's natural voice. By using speech synthesis technology, the user's voice tone and emotions are also reflected, resulting in a more realistic conversation.

[0332] As a concrete example, consider a scenario where a user is giving a presentation in a public place. If the user shows signs of nervousness, the server can detect this using its emotion engine and offer a response such as, "Relax, you're doing great." If the user selects this option, the device will generate a voice in a comforting tone to support the user.

[0333] This invention is a system that supports users in maintaining the most appropriate and effective communication by performing comprehensive information processing, including emotions. This technology is useful not only for people with selective mutism but also for all people who require special communication support.

[0334] The following describes the processing flow.

[0335] Step 1:

[0336] The device captures voice or text data as soon as the user begins a conversation. This includes what the user says to the device and messages they type on the screen.

[0337] Step 2:

[0338] The device temporarily stores the collected speech data, and once it confirms that the communication environment is stable, it encrypts the data and sends it to the server. This data also includes emotional information based on the user's consent.

[0339] Step 3:

[0340] The server converts the received data into a format suitable for analysis, analyzes the conversation content using natural language processing techniques, and extracts the user's conversation patterns. This analysis includes frequently occurring words, the flow of the conversation, and the user's preferences.

[0341] Step 4:

[0342] The server uses an emotion engine to recognize the user's emotional state from their voice tone and selected words. For example, it analyzes the tone and tempo of the voice data to determine whether the user is tense or relaxed.

[0343] Step 5:

[0344] The server uses a generative AI model to generate response candidates, taking into account the user's conversation patterns and emotional state. This generation process utilizes the user's past conversation history and current contextual information to prepare multiple appropriate responses.

[0345] Step 6:

[0346] The device displays generated response options to the user. The UI is designed to be intuitive and easy for the user to operate, and sometimes a response that helps to ease tension is displayed at the top.

[0347] Step 7:

[0348] The user selects a response from the presented options that best matches their intentions and feelings. The selected response is optimized to suit the individual's emotions and situation.

[0349] Step 8:

[0350] The device uses a speech synthesis engine to output the selected response as speech that reflects the user's voice and tone. This makes it sound to others as if the user were speaking naturally.

[0351] Step 9:

[0352] Users can evaluate whether the outputted response was intended in the communication and provide feedback to the system.

[0353] This allows the system to use this feedback as an indicator to further improve the accuracy of future responses.

[0354] (Example 2)

[0355] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0356] In modern society, effective communication is required in a variety of situations, but users sometimes have difficulty conveying their intentions appropriately. In such situations, there is a need for technologies that support smooth dialogue tailored to the user's mental state and linguistic characteristics. In particular, responses that respond to emotional states are important, and there is a challenge in that support in this area is insufficient.

[0357] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0358] In this invention, the server includes a device for capturing user speech information, a device for analyzing the speech information to extract the user's language tendencies, and a generative AI mechanism for analyzing the user's emotional state and generating response candidates based on the language tendencies and emotional state. This enables appropriate and effective communication support that takes emotions into consideration.

[0359] "User speech information" refers to the content that users utter in voice or text format, and this information is collected as data.

[0360] "Linguistic tendencies" refer to specific language usage patterns and characteristics of frequently occurring words and phrases observed in users' conversations.

[0361] "Emotional state" refers to the psychological or emotional condition analyzed from the user's utterances and selected words.

[0362] A "generative AI mechanism" is a system that incorporates artificial intelligence technology used to generate appropriate responses based on the user's language tendencies and emotional state.

[0363] A "communication support system" refers to a device or program that provides a technical mechanism to assist users in communicating their intentions smoothly and appropriately.

[0364] This invention is a communication support system designed to provide more natural and appropriate responses when users communicate. The system aims to efficiently analyze speech information while considering the user's emotional state and generate appropriate responses.

[0365] First, the content spoken by the user, either by voice or text, is captured by the device. Here, voice processing software, commonly used as speech recognition technology, is used to convert the voice data into text. The text data is temporarily stored on the device and then transmitted to the server using encryption technology to ensure security.

[0366] The server utilizes natural language processing techniques to analyze the received text data. Here, general analysis software is used to extract the user's language tendencies. Simultaneously, a sentiment analysis engine is employed to evaluate the user's emotional state. These evaluation results are stored as individual user profiles.

[0367] Based on the analysis results and emotional state, the server generates response candidates using a generative AI model. The generative AI model enables natural dialogue that matches the user's psychological state. An example of a prompt would be, "Generate a natural, relaxing response for when the user is nervous during a presentation."

[0368] The generated responses are presented to the user by the device, and the user can select one through an intuitive interface. The selected response is then output naturally in the user's voice using speech synthesis technology. This enables the user to engage in smooth, emotionally sensitive communication.

[0369] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0370] Step 1:

[0371] The device receives voice or text input from the user. In the case of voice input, the device uses speech recognition software to convert the voice data into text. The resulting text data is then output. The speech recognition process includes an acoustic model and a language model for converting voice samples into text.

[0372] Step 2:

[0373] The terminal encrypts the generated text data and sends it to the server. It receives text data as input, protects it using an encryption algorithm, and outputs encrypted data. This ensures confidentiality during data transfer.

[0374] Step 3:

[0375] The server receives encrypted text data, decrypts it, and returns it to its original text data. This decrypted data becomes the input, and raw text ready for analysis is output.

[0376] Step 4:

[0377] The server analyzes the decoded text data using natural language processing techniques. This extracts the user's language tendencies and the intent expressed in the text. The input for this step is raw text, and the output is the analyzed language features and conversational intent.

[0378] Step 5:

[0379] The server uses a sentiment analysis engine to evaluate the user's emotional state. The input is pre-analyzed text, and the output is the detected emotional state and psychological tendencies. This allows for the clear identification of subtle emotions hidden within the text.

[0380] Step 6:

[0381] The server uses a generative AI model to generate response candidates based on the analysis results and emotional state. The generative AI model takes linguistic features and emotional state as input and generates emotionally sensitive response candidates as output. This process creates natural-sounding utterances that are appropriate to the user's current psychological state.

[0382] Step 7:

[0383] The terminal presents the user with generated response options. This can be done by displaying the response on the screen or by providing voice guidance. The input here is response data from the server, which is output as choices for the user to select.

[0384] Step 8:

[0385] The user selects the most appropriate response from the presented options. The selected response is recognized as input, and the next action is planned accordingly.

[0386] Step 9:

[0387] The device outputs the selected response in speech format using speech synthesis technology. The input here is the selected text response, and the output is synthesized speech with a tone similar to the user's voice. This enables more natural communication.

[0388] (Application Example 2)

[0389] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0390] Conventional communication assistance systems have the problem of being unable to conduct conversations that fully consider the user's emotions, and thus are unable to communicate naturally and smoothly. Furthermore, in customer service at physical stores, it is necessary to appropriately understand the customer's emotions and respond accordingly, but there has been no effective technology to support this.

[0391] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0392] In this invention, the server includes means for capturing user speech data, means for analyzing the speech data to extract the user's conversation patterns, and means for a generative model that generates response candidates based on the user's current contextual information and emotional state. This makes it possible to appropriately grasp the customer's emotions and support optimal communication in interactions with customers in physical stores.

[0393] "User speech data" refers to the audio and text information that users produce when communicating.

[0394] "Conversation patterns" refer to the collective characteristics and tendencies extracted from conversations a user has had in the past.

[0395] "Contextual information" refers to information that includes the situation and background in which an utterance is made, as well as the emotional state of the speaker.

[0396] A "generative model" refers to a structure or algorithm used to generate a response based on input data, employing natural language processing techniques.

[0397] A "display device" refers to an electronic device used to visually present generated information to a user.

[0398] "Speech synthesis" is a technology that mechanically generates human speech from text.

[0399] "Emotional state" refers to the psychological state that can be inferred from the user's statements and actions.

[0400] To implement this invention, a terminal device, such as smart glasses worn by the user, and a server connected to it are required. The terminal device is equipped with a camera and microphone, and has the function to transmit real-time data acquired by these devices to the server. The server receives the user's speech data and facial expression data and uses natural language processing technology and facial recognition technology to analyze them. Specifically, it is common to use Google Cloud Speech-to-Text API or OpenCV.

[0401] The server first converts the acquired audio data into text, and then extracts the user's conversation patterns and emotional state from the text data and facial expression data. Based on this information, a generative AI model generates the optimal response. The generated response is then presented to the terminal, with its tone and selected language adjusted to match the user's emotional state. The terminal visually communicates this response to the user and provides audio feedback as needed.

[0402] For example, in a physical store, if a customer's anxious expression or speech is detected during customer service, a suggestion such as "Please let us know if you need any assistance" will be displayed on the staff member's smart glasses. This allows staff members to provide more appropriate and effective customer service.

[0403] A concrete example of a prompt for a generative AI model is: "Generate the most appropriate response when the customer looks anxious. Choose the right words based on the user's emotional state." This prompt allows the system to provide a response appropriate to the emotional state, improving the quality of communication.

[0404] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0405] Step 1:

[0406] The device captures the user's speech and facial expressions in real time using its camera and microphone. The input is the user's voice and video data, and the output is the transmission of this data to a server. Specifically, sensors within the device record the voice, and the camera captures the facial expressions.

[0407] Step 2:

[0408] The server converts the received audio data into text. This process uses the Google Cloud Speech-to-Text API. The input is audio data, and the output is text data. The operation here is to process the audio signal as a text string.

[0409] Step 3:

[0410] The server analyzes facial expression data to infer the user's emotions. It uses OpenCV to detect facial features and then analyzes them with an emotion engine. The input is facial image data, and the output is the inferred emotional state. Specifically, it measures changes and patterns in facial expressions and assigns emotion labels.

[0411] Step 4:

[0412] The server uses natural language processing technology to extract conversation patterns from text data. A generative AI model then generates response candidates based on this. The input is text data and emotion states, and the output is the text of the response candidates. In operation, prompt sentences are sent to the generative AI model, which automatically generates responses.

[0413] Step 5:

[0414] The terminal presents the user with generated response options visually or audibly. The user selects the most appropriate response, which is then output via speech synthesis. The input is the text data of the response options, and the output is either an audio message or text on a display. The operation here is to display the response on a visual device or play the audio through a speaker.

[0415] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0416] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0417] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0418] [Third Embodiment]

[0419] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0420] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0421] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0422] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0423] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0424] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0425] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0426] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0427] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0428] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0429] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0430] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0431] The communication assistance system according to the present invention is designed to enable users with selective mutism to engage in smooth conversations in various situations. This system operates by combining speech analysis technology, natural language processing (NLP), generative AI models, and speech synthesis technology.

[0432] When a user uses a device, the system first captures the user's speech data. This speech data is recorded in text or audio format and securely stored on the device. The collected data is then encrypted and transmitted to a server.

[0433] The server analyzes the received data and processes it to identify the user's conversation patterns. Specifically, the server utilizes natural language processing techniques to extract frequently occurring themes and the user's speaking tendencies in specific contexts. Based on this information, the server updates the user's profile and uses it to understand their individual conversation style.

[0434] Based on the analysis results, the server uses a generative AI model to generate response candidates for various conversation scenarios. This prepares responses that enable users to speak smoothly in specific situations. The response candidates are optimized based on the user's preferences and habits, and multiple options are presented.

[0435] The device presents the user with a list of generated response options. The user selects the response that best reflects their intention from the displayed list. The selected response is converted into natural-sounding speech using the device's built-in speech synthesis function and output in the user's voice. This process effectively communicates the user's intentions to others.

[0436] As a concrete example, consider a scenario where a user orders at a cafe. Based on past data, the server predicts the drinks the user frequently orders and generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please" from the presented options, and the terminal synthesizes the selection into speech and conveys it to the staff.

[0437] Thus, the system of the present invention provides support to enable users of selective mutism to express their intentions appropriately in various social situations and has the effect of reducing communication barriers.

[0438] The following describes the processing flow.

[0439] Step 1:

[0440] The device captures speech data through voice input or text messages during everyday conversations. This data is processed in the background so as not to interrupt the user's actions.

[0441] Step 2:

[0442] The device temporarily stores the captured speech data in local storage and, when certain conditions are met (for example, when connected to Wi-Fi), encrypts the data and sends it to the server.

[0443] Step 3:

[0444] The server receives encrypted data sent from the terminal and stores it in a database. This data is then used as reference data for subsequent analysis processes.

[0445] Step 4:

[0446] The server analyzes the received speech data using natural language processing technology. Specifically, it extracts frequently occurring words and phrases, identifies patterns in the user's topics and the flow of conversation, and updates the user's profile accordingly.

[0447] Step 5:

[0448] The server uses a generative AI model based on the analyzed data to generate response candidates to support user communication. These candidates are optimized based on the user's past conversation history and contextual information.

[0449] Step 6:

[0450] The terminal presents the user with a list of response options sent from the server. The user interface is intuitively designed, allowing users to easily make selections.

[0451] Step 7:

[0452] The user selects the response that best reflects their intention from the presented options. This selection is typically made using touch controls or a visual interface.

[0453] Step 8:

[0454] The device uses its speech synthesis function to convert the user's selected response into speech, outputting it as a natural-sounding voice. This ensures that the user's intentions are communicated to others.

[0455] Step 9:

[0456] Users can provide feedback to the device on whether the response was as intended. This feedback can be used to improve the quality of future response generation.

[0457] This processing flow will enable users to express their intentions smoothly even in difficult situations.

[0458] (Example 1)

[0459] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0460] It is crucial to enable individuals with selective mutism to communicate effectively in specific situations. However, current technology does not adequately provide speech support tailored to each individual's characteristics. As a result, individuals face challenges in naturally conveying their intentions and having their opportunities for social participation limited.

[0461] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0462] In this invention, the server includes a device for acquiring speech information, a device for analyzing the speech information to extract the user's dialogue tendencies, and a device for generating response candidates using a generative artificial intelligence model based on the analysis results. This makes it possible to generate responses optimized for the individual characteristics of the user and support smooth communication in specific situations.

[0463] "Speech information" refers to the content of communications made by users in text or voice format.

[0464] A "device for acquiring data" is a device that has the function of detecting and recording voice and text data.

[0465] "Dialogue tendencies" refer to patterns in speaking style and content that users have shown in past communications.

[0466] A "speech analyzer" is a device that has the function of analyzing the user's dialogue tendencies from the acquired speech information.

[0467] A "generative artificial intelligence model" is a model that uses artificial intelligence technology to generate responses based on the user's speech.

[0468] "Response candidates" refer to multiple response options that are generated based on a specific situation or context and presented to the user.

[0469] A "speech conversion device" is a device that has the function of converting selected response candidates into a natural speech format and outputting it.

[0470] This invention is a system that supports users with selective mutism in communicating smoothly in a variety of situations. Specific embodiments of this system are described below.

[0471] Users use a dedicated terminal to acquire their own speech information in real time. The terminal has a microphone and keyboard as input devices, and the speech information is recorded in voice and text format. The terminal encrypts the recorded information and then sends it to the server via a secure protocol.

[0472] The server analyzes the received speech information using natural language processing techniques. These techniques include morphological analysis and contextual understanding. Through this analysis, the server extracts the user's conversational tendencies and main themes, and updates the user profile based on this information.

[0473] The generation AI model on the server considers the analysis results and user profile to generate a variety of response candidates. Here, the generation AI model improves the appropriateness of the responses by also referring to the user's past preference history. An example of a prompt is "Generate drink candidates that the user might like."

[0474] The terminal presents the user with a list of response options sent from the server. The user selects the most suitable response from the options, and the terminal converts the selected response into natural-sounding speech using speech synthesis technology. This speech is then output as the user's voice, allowing them to clearly communicate their intentions.

[0475] As a concrete example, consider ordering a drink at a cafe. Based on past data of drinks the user frequently orders, the server generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please," and the terminal announces this verbally.

[0476] In this way, the system plays a role in promoting social interaction by providing communication support tailored to the individual needs of users.

[0477] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0478] Step 1:

[0479] The user activates the device and inputs spoken information via voice or text. The device converts the input voice into text data using speech recognition technology. The input voice data and the converted text data are encrypted to ensure security and recorded. As a result, the spoken information is output in both text and voice formats.

[0480] Step 2:

[0481] The terminal sends encrypted speech information to the server. This transmission uses a secure communication protocol (e.g., HTTPS). The input consists of speech and text data, which are then transferred to the server.

[0482] Step 3:

[0483] The server analyzes the received speech information using natural language processing techniques. This analysis involves morphological analysis and contextual understanding to identify the user's dialogue tendencies and themes. The input is text data, and the output is data on dialogue tendencies and related themes.

[0484] Step 4:

[0485] Based on the analysis results, the server generates response candidates using a generative artificial intelligence model. This process uses prompts to input analysis data into the AI ​​model, generating a variety of response options. For example, a prompt like "Generate drink candidates that the user would like" might be used. The output is a list of response candidates.

[0486] Step 5:

[0487] The terminal presents the user with a list of response options sent from the server. The user reviews the options displayed on the screen and selects the most appropriate response. Input for selection can be via touch or voice commands.

[0488] Step 6:

[0489] The device uses speech synthesis technology to convert the user's selected response into speech. This process transforms text data into natural-sounding speech data. The output is the actual audio that is played back. This ensures that the user's intentions are communicated in a way that is easily understood by others.

[0490] (Application Example 1)

[0491] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0492] There is a need for support tools that enable users with communication difficulties to communicate smoothly in public places and face-to-face interactions. Current technology fails to fully utilize users' individual contextual information and preference history, making it difficult to provide timely and optimal responses to facilitate the flow of conversation. In particular, there is a high need for technology that effectively and naturally assists speech for individuals with selective mutism or other communication disorders.

[0493] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0494] In this invention, the server includes means for capturing user speech information, means for analyzing the speech information to extract conversational tendencies, and means for a generative model that generates response candidates based on current situation information and preference history. This makes it possible for users to be presented with responses effectively and quickly in face-to-face interactions, and for those responses to be communicated in a natural manner using speech synthesis technology.

[0495] "User" refers to an individual who receives communication support using this system.

[0496] "Speech information" refers to the content of a conversation expressed by the user in written or spoken form.

[0497] "Means of capture" refers to components that have the function of acquiring and storing speech information.

[0498] "Means of analysis" refers to techniques that analyze acquired speech information and extract the user's conversational tendencies and characteristics.

[0499] "Conversational tendencies" refer to the linguistic characteristics and speech patterns that users exhibit during conversations.

[0500] "Preference history" refers to data about responses and preferences that a user has selected in the past.

[0501] A "generative model" is a module that uses artificial intelligence technology to generate response candidates based on the user's current situation information and preference history.

[0502] A "response option" refers to a set of utterances presented to the user as choices they can use in a specific situation.

[0503] "Speech synthesis" is a technology that converts text information into natural-sounding speech.

[0504] "Means of communication to others" refers to technologies for effectively conveying synthesized speech responses to people other than the user.

[0505] This invention is a communication assistance system aimed at enabling users with selective mutism or other communication difficulties to communicate smoothly in face-to-face interactions. The system is implemented via a smartphone or similar user terminal.

[0506] The system's core process combines the acquisition, analysis, and generation of speech information, followed by speech synthesis output. Users input speech information via a terminal in either voice or text format. The terminal converts the voice data to text using speech recognition technology, encrypts the data, and then sends it to the server. Speech recognition technologies such as Google Cloud Speech-to-Text and other similar services are available.

[0507] The server analyzes the received data using natural language processing (NLP) techniques to extract the user's conversational tendencies. NLP tools such as NLTK and SpaCy are used here. Then, using a generative AI model such as GPT-4, appropriate response candidates are generated based on the user's past preference history and current situation. The generated responses are optimized based on important topics and the user's preferences and presented as multiple options.

[0508] On the device, generated response suggestions are displayed to the user. The response selected by the user is converted into natural-sounding speech using speech synthesis technology such as Google Text-to-Speech and communicated to the other party. This process facilitates smoother communication.

[0509] As a concrete example of use, consider a scenario where a user orders tea at a cafe. Based on the user's past purchase history, the system recommends a type of tea and generates a response such as, "I'd like Darjeeling tea, please." Once the user selects this, the terminal verbally communicates this to the staff.

[0510] As an example of a prompt, you can instruct the AI ​​as follows: "Generate a list of recommended teas based on the user's preferences and past purchase history. Please suggest three options, taking into account current preferences and trends."

[0511] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0512] Step 1:

[0513] The user inputs speech information into the device. The input format can be voice or text. In the case of voice input, the device uses Google Cloud Speech-to-Text to convert the voice data into text format. The input data is then sent to the next step either in its original format or as text.

[0514] Step 2:

[0515] The terminal encrypts the input data to store it securely. AES encryption technology is used here. This process protects user privacy while preparing the data for transmission to the next step. The encrypted data is output and sent to the server.

[0516] Step 3:

[0517] The server decrypts the received encrypted data and analyzes its contents. It uses natural language processing techniques to analyze the text data and extract user conversational trends and important topics. By utilizing tools such as NLTK and SpaCy, it outputs trend data from the input text.

[0518] Step 4:

[0519] The server uses a generative AI model, such as GPT-4, to generate response candidates that take into account the user's context and preference history. The generative AI model generates an appropriate response based on past data and the current prompt, and a list of response candidates is output.

[0520] Step 5:

[0521] The device presents the user with the received response options. The user selects the response that matches their intent. The user's selection is output, and the result is passed on to the next step.

[0522] Step 6:

[0523] The selected response is converted into natural-sounding speech using speech synthesis technology on the device. Text data is converted to speech data using tools such as Google Text-to-Speech. The converted speech is played through the device's speaker and transmitted to the other party.

[0524] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0525] The communication assistance system according to the present invention is designed to facilitate smooth communication in situations where it is difficult for the user to speak. This system enables more natural and appropriate communication by combining an emotion engine at each stage of speech data capture, analysis, response generation, and voice output.

[0526] First, as users use their devices, speech data is continuously captured. This data is entered in either text or voice format, and the device collects it. The collected data is encrypted and sent to the server.

[0527] The server analyzes the received speech data to identify the user's conversation patterns. Natural language processing techniques are used in this analysis to extract the user's everyday speaking tendencies and frequently occurring words. Simultaneously, an emotion engine recognizes emotions from the user's speech and selected words, and evaluates their emotional state. This information is also added to the user profile maintained by the server and learned as a comprehensive characteristic of the user.

[0528] Based on this data, the server generates response candidates using a generative AI model. During this process, the tone and content of the responses are adjusted to match the user's current emotional state, taking into account the user's emotional condition. For example, if the server determines that the user is stressed, responses that promote relaxation are prioritized.

[0529] The device presents the user with a list of generated response options. The user selects the response that best suits them through an intuitive interface. The selected response is then converted into speech, tailored to the context and emotions, and output in the user's natural voice. By using speech synthesis technology, the user's voice tone and emotions are also reflected, resulting in a more realistic conversation.

[0530] As a concrete example, consider a scenario where a user is giving a presentation in a public place. If the user shows signs of nervousness, the server can detect this using its emotion engine and offer a response such as, "Relax, you're doing great." If the user selects this option, the device will generate a voice in a comforting tone to support the user.

[0531] This invention is a system that supports users in maintaining the most appropriate and effective communication by performing comprehensive information processing, including emotions. This technology is useful not only for people with selective mutism but also for all people who require special communication support.

[0532] The following describes the processing flow.

[0533] Step 1:

[0534] The device captures voice or text data as soon as the user begins a conversation. This includes what the user says to the device and messages they type on the screen.

[0535] Step 2:

[0536] The device temporarily stores the collected speech data, and once it confirms that the communication environment is stable, it encrypts the data and sends it to the server. This data also includes emotional information based on the user's consent.

[0537] Step 3:

[0538] The server converts the received data into a format suitable for analysis, analyzes the conversation content using natural language processing techniques, and extracts the user's conversation patterns. This analysis includes frequently occurring words, the flow of the conversation, and the user's preferences.

[0539] Step 4:

[0540] The server uses an emotion engine to recognize the user's emotional state from their voice tone and selected words. For example, it analyzes the tone and tempo of the voice data to determine whether the user is tense or relaxed.

[0541] Step 5:

[0542] The server uses a generative AI model to generate response candidates, taking into account the user's conversation patterns and emotional state. This generation process utilizes the user's past conversation history and current contextual information to prepare multiple appropriate responses.

[0543] Step 6:

[0544] The device displays generated response options to the user. The UI is designed to be intuitive and easy for the user to operate, and sometimes a response that helps to ease tension is displayed at the top.

[0545] Step 7:

[0546] The user selects a response from the presented options that best matches their intentions and feelings. The selected response is optimized to suit the individual's emotions and situation.

[0547] Step 8:

[0548] The device uses a speech synthesis engine to output the selected response as speech that reflects the user's voice and tone. This makes it sound to others as if the user were speaking naturally.

[0549] Step 9:

[0550] Users can evaluate whether the outputted response was intended in the communication and provide feedback to the system.

[0551] This allows the system to use this feedback as an indicator to further improve the accuracy of future responses.

[0552] (Example 2)

[0553] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0554] In modern society, effective communication is required in a variety of situations, but users sometimes have difficulty conveying their intentions appropriately. In such situations, there is a need for technologies that support smooth dialogue tailored to the user's mental state and linguistic characteristics. In particular, responses that respond to emotional states are important, and there is a challenge in that support in this area is insufficient.

[0555] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0556] In this invention, the server includes a device for capturing user speech information, a device for analyzing the speech information to extract the user's language tendencies, and a generative AI mechanism for analyzing the user's emotional state and generating response candidates based on the language tendencies and emotional state. This enables appropriate and effective communication support that takes emotions into consideration.

[0557] "User speech information" refers to the content that users utter in voice or text format, and this information is collected as data.

[0558] "Linguistic tendencies" refer to specific language usage patterns and characteristics of frequently occurring words and phrases observed in users' conversations.

[0559] "Emotional state" refers to the psychological or emotional condition analyzed from the user's utterances and selected words.

[0560] A "generative AI mechanism" is a system that incorporates artificial intelligence technology used to generate appropriate responses based on the user's language tendencies and emotional state.

[0561] A "communication support system" refers to a device or program that provides a technical mechanism to assist users in communicating their intentions smoothly and appropriately.

[0562] This invention is a communication support system designed to provide more natural and appropriate responses when users communicate. The system aims to efficiently analyze speech information while considering the user's emotional state and generate appropriate responses.

[0563] First, the content spoken by the user, either by voice or text, is captured by the device. Here, voice processing software, commonly used as speech recognition technology, is used to convert the voice data into text. The text data is temporarily stored on the device and then transmitted to the server using encryption technology to ensure security.

[0564] The server utilizes natural language processing techniques to analyze the received text data. Here, general analysis software is used to extract the user's language tendencies. Simultaneously, a sentiment analysis engine is employed to evaluate the user's emotional state. These evaluation results are stored as individual user profiles.

[0565] Based on the analysis results and emotional state, the server generates response candidates using a generative AI model. The generative AI model enables natural dialogue that matches the user's psychological state. An example of a prompt would be, "Generate a natural, relaxing response for when the user is nervous during a presentation."

[0566] The generated responses are presented to the user by the device, and the user can select one through an intuitive interface. The selected response is then output naturally in the user's voice using speech synthesis technology. This enables the user to engage in smooth, emotionally sensitive communication.

[0567] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0568] Step 1:

[0569] The device receives voice or text input from the user. In the case of voice input, the device uses speech recognition software to convert the voice data into text. The resulting text data is then output. The speech recognition process includes an acoustic model and a language model for converting voice samples into text.

[0570] Step 2:

[0571] The terminal encrypts the generated text data and sends it to the server. It receives text data as input, protects it using an encryption algorithm, and outputs encrypted data. This ensures confidentiality during data transfer.

[0572] Step 3:

[0573] The server receives encrypted text data, decrypts it, and returns it to its original text data. This decrypted data becomes the input, and raw text ready for analysis is output.

[0574] Step 4:

[0575] The server analyzes the decoded text data using natural language processing techniques. This extracts the user's language tendencies and the intent expressed in the text. The input for this step is raw text, and the output is the analyzed language features and conversational intent.

[0576] Step 5:

[0577] The server uses a sentiment analysis engine to evaluate the user's emotional state. The input is pre-analyzed text, and the output is the detected emotional state and psychological tendencies. This allows for the clear identification of subtle emotions hidden within the text.

[0578] Step 6:

[0579] The server uses a generative AI model to generate response candidates based on the analysis results and emotional state. The generative AI model takes linguistic features and emotional state as input and generates emotionally sensitive response candidates as output. This process creates natural-sounding utterances that are appropriate to the user's current psychological state.

[0580] Step 7:

[0581] The terminal presents the user with generated response options. This can be done by displaying the response on the screen or by providing voice guidance. The input here is response data from the server, which is output as choices for the user to select.

[0582] Step 8:

[0583] The user selects the most appropriate response from the presented options. The selected response is recognized as input, and the next action is planned accordingly.

[0584] Step 9:

[0585] The device outputs the selected response in speech format using speech synthesis technology. The input here is the selected text response, and the output is synthesized speech with a tone similar to the user's voice. This enables more natural communication.

[0586] (Application Example 2)

[0587] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0588] Conventional communication assistance systems have the problem of being unable to conduct conversations that fully consider the user's emotions, and thus are unable to communicate naturally and smoothly. Furthermore, in customer service at physical stores, it is necessary to appropriately understand the customer's emotions and respond accordingly, but there has been no effective technology to support this.

[0589] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0590] In this invention, the server includes means for capturing user speech data, means for analyzing the speech data to extract the user's conversation patterns, and means for a generative model that generates response candidates based on the user's current contextual information and emotional state. This makes it possible to appropriately grasp the customer's emotions and support optimal communication in interactions with customers in physical stores.

[0591] "User speech data" refers to the audio and text information that users produce when communicating.

[0592] "Conversation patterns" refer to the collective characteristics and tendencies extracted from conversations a user has had in the past.

[0593] "Contextual information" refers to information that includes the situation and background in which an utterance is made, as well as the emotional state of the speaker.

[0594] A "generative model" refers to a structure or algorithm used to generate a response based on input data, employing natural language processing techniques.

[0595] A "display device" refers to an electronic device used to visually present generated information to a user.

[0596] "Speech synthesis" is a technology that mechanically generates human speech from text.

[0597] "Emotional state" refers to the psychological state that can be inferred from the user's statements and actions.

[0598] To implement this invention, a terminal device, such as smart glasses worn by the user, and a server connected to it are required. The terminal device is equipped with a camera and microphone, and has the function to transmit real-time data acquired by these devices to the server. The server receives the user's speech data and facial expression data and uses natural language processing technology and facial recognition technology to analyze them. Specifically, it is common to use Google Cloud Speech-to-Text API or OpenCV.

[0599] The server first converts the acquired audio data into text, and then extracts the user's conversation patterns and emotional state from the text data and facial expression data. Based on this information, a generative AI model generates the optimal response. The generated response is then presented to the terminal, with its tone and selected language adjusted to match the user's emotional state. The terminal visually communicates this response to the user and provides audio feedback as needed.

[0600] For example, in a physical store, if a customer's anxious expression or speech is detected during customer service, a suggestion such as "Please let us know if you need any assistance" will be displayed on the staff member's smart glasses. This allows staff members to provide more appropriate and effective customer service.

[0601] A concrete example of a prompt for a generative AI model is: "Generate the most appropriate response when the customer looks anxious. Choose the right words based on the user's emotional state." This prompt allows the system to provide a response appropriate to the emotional state, improving the quality of communication.

[0602] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0603] Step 1:

[0604] The device captures the user's speech and facial expressions in real time using its camera and microphone. The input is the user's voice and video data, and the output is the transmission of this data to a server. Specifically, sensors within the device record the voice, and the camera captures the facial expressions.

[0605] Step 2:

[0606] The server converts the received audio data into text. This process uses the Google Cloud Speech-to-Text API. The input is audio data, and the output is text data. The operation here is to process the audio signal as a text string.

[0607] Step 3:

[0608] The server analyzes facial expression data to infer the user's emotions. It uses OpenCV to detect facial features and then analyzes them with an emotion engine. The input is facial image data, and the output is the inferred emotional state. Specifically, it measures changes and patterns in facial expressions and assigns emotion labels.

[0609] Step 4:

[0610] The server uses natural language processing technology to extract conversation patterns from text data. A generative AI model then generates response candidates based on this. The input is text data and emotion states, and the output is the text of the response candidates. In operation, prompt sentences are sent to the generative AI model, which automatically generates responses.

[0611] Step 5:

[0612] The terminal presents the user with generated response options visually or audibly. The user selects the most appropriate response, which is then output via speech synthesis. The input is the text data of the response options, and the output is either an audio message or text on a display. The operation here is to display the response on a visual device or play the audio through a speaker.

[0613] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0614] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0615] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0616] [Fourth Embodiment]

[0617] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0618] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0619] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0620] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0621] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0622] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0623] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0624] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0625] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0626] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0627] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0628] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0629] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0630] The communication assistance system according to the present invention is designed to enable users with selective mutism to engage in smooth conversations in various situations. This system operates by combining speech analysis technology, natural language processing (NLP), generative AI models, and speech synthesis technology.

[0631] When a user uses a device, the system first captures the user's speech data. This speech data is recorded in text or audio format and securely stored on the device. The collected data is then encrypted and transmitted to a server.

[0632] The server analyzes the received data and processes it to identify the user's conversation patterns. Specifically, the server utilizes natural language processing techniques to extract frequently occurring themes and the user's speaking tendencies in specific contexts. Based on this information, the server updates the user's profile and uses it to understand their individual conversation style.

[0633] Based on the analysis results, the server uses a generative AI model to generate response candidates for various conversation scenarios. This prepares responses that enable users to speak smoothly in specific situations. The response candidates are optimized based on the user's preferences and habits, and multiple options are presented.

[0634] The device presents the user with a list of generated response options. The user selects the response that best reflects their intention from the displayed list. The selected response is converted into natural-sounding speech using the device's built-in speech synthesis function and output in the user's voice. This process effectively communicates the user's intentions to others.

[0635] As a concrete example, consider a scenario where a user orders at a cafe. Based on past data, the server predicts the drinks the user frequently orders and generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please" from the presented options, and the terminal synthesizes the selection into speech and conveys it to the staff.

[0636] Thus, the system of the present invention provides support to enable users of selective mutism to express their intentions appropriately in various social situations and has the effect of reducing communication barriers.

[0637] The following describes the processing flow.

[0638] Step 1:

[0639] The device captures speech data through voice input or text messages during everyday conversations. This data is processed in the background so as not to interrupt the user's actions.

[0640] Step 2:

[0641] The device temporarily stores the captured speech data in local storage and, when certain conditions are met (for example, when connected to Wi-Fi), encrypts the data and sends it to the server.

[0642] Step 3:

[0643] The server receives encrypted data sent from the terminal and stores it in a database. This data is then used as reference data for subsequent analysis processes.

[0644] Step 4:

[0645] The server analyzes the received speech data using natural language processing technology. Specifically, it extracts frequently occurring words and phrases, identifies patterns in the user's topics and the flow of conversation, and updates the user's profile accordingly.

[0646] Step 5:

[0647] The server uses a generative AI model based on the analyzed data to generate response candidates to support user communication. These candidates are optimized based on the user's past conversation history and contextual information.

[0648] Step 6:

[0649] The terminal presents the user with a list of response options sent from the server. The user interface is intuitively designed, allowing users to easily make selections.

[0650] Step 7:

[0651] The user selects the response that best reflects their intention from the presented options. This selection is typically made using touch controls or a visual interface.

[0652] Step 8:

[0653] The device uses its speech synthesis function to convert the user's selected response into speech, outputting it as a natural-sounding voice. This ensures that the user's intentions are communicated to others.

[0654] Step 9:

[0655] Users can provide feedback to the device on whether the response was as intended. This feedback can be used to improve the quality of future response generation.

[0656] This processing flow will enable users to express their intentions smoothly even in difficult situations.

[0657] (Example 1)

[0658] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0659] It is crucial to enable individuals with selective mutism to communicate effectively in specific situations. However, current technology does not adequately provide speech support tailored to each individual's characteristics. As a result, individuals face challenges in naturally conveying their intentions and having their opportunities for social participation limited.

[0660] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0661] In this invention, the server includes a device for acquiring speech information, a device for analyzing the speech information to extract the user's dialogue tendencies, and a device for generating response candidates using a generative artificial intelligence model based on the analysis results. This makes it possible to generate responses optimized for the individual characteristics of the user and support smooth communication in specific situations.

[0662] "Speech information" refers to the content of communications made by users in text or voice format.

[0663] A "device for acquiring data" is a device that has the function of detecting and recording voice and text data.

[0664] "Dialogue tendencies" refer to patterns in speaking style and content that users have shown in past communications.

[0665] A "speech analyzer" is a device that has the function of analyzing the user's dialogue tendencies from the acquired speech information.

[0666] A "generative artificial intelligence model" is a model that uses artificial intelligence technology to generate responses based on the user's speech.

[0667] "Response candidates" refer to multiple response options that are generated based on a specific situation or context and presented to the user.

[0668] A "speech conversion device" is a device that has the function of converting selected response candidates into a natural speech format and outputting it.

[0669] This invention is a system that supports users with selective mutism in communicating smoothly in a variety of situations. Specific embodiments of this system are described below.

[0670] Users use a dedicated terminal to acquire their own speech information in real time. The terminal has a microphone and keyboard as input devices, and the speech information is recorded in voice and text format. The terminal encrypts the recorded information and then sends it to the server via a secure protocol.

[0671] The server analyzes the received speech information using natural language processing techniques. These techniques include morphological analysis and contextual understanding. Through this analysis, the server extracts the user's conversational tendencies and main themes, and updates the user profile based on this information.

[0672] The generation AI model on the server considers the analysis results and user profile to generate a variety of response candidates. Here, the generation AI model improves the appropriateness of the responses by also referring to the user's past preference history. An example of a prompt is "Generate drink candidates that the user might like."

[0673] The terminal presents the user with a list of response options sent from the server. The user selects the most suitable response from the options, and the terminal converts the selected response into natural-sounding speech using speech synthesis technology. This speech is then output as the user's voice, allowing them to clearly communicate their intentions.

[0674] As a concrete example, consider ordering a drink at a cafe. Based on past data of drinks the user frequently orders, the server generates response options such as "I'd like a cafe latte, please" and "I'd like an espresso, please." The user selects "I'd like a cafe latte, please," and the terminal announces this verbally.

[0675] In this way, the system plays a role in promoting social interaction by providing communication support tailored to the individual needs of users.

[0676] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0677] Step 1:

[0678] The user activates the device and inputs spoken information via voice or text. The device converts the input voice into text data using speech recognition technology. The input voice data and the converted text data are encrypted to ensure security and recorded. As a result, the spoken information is output in both text and voice formats.

[0679] Step 2:

[0680] The terminal sends encrypted speech information to the server. This transmission uses a secure communication protocol (e.g., HTTPS). The input consists of speech and text data, which are then transferred to the server.

[0681] Step 3:

[0682] The server analyzes the received speech information using natural language processing techniques. This analysis involves morphological analysis and contextual understanding to identify the user's dialogue tendencies and themes. The input is text data, and the output is data on dialogue tendencies and related themes.

[0683] Step 4:

[0684] Based on the analysis results, the server generates response candidates using a generative artificial intelligence model. This process uses prompts to input analysis data into the AI ​​model, generating a variety of response options. For example, a prompt like "Generate drink candidates that the user would like" might be used. The output is a list of response candidates.

[0685] Step 5:

[0686] The terminal presents the user with a list of response options sent from the server. The user reviews the options displayed on the screen and selects the most appropriate response. Input for selection can be via touch or voice commands.

[0687] Step 6:

[0688] The device uses speech synthesis technology to convert the user's selected response into speech. This process transforms text data into natural-sounding speech data. The output is the actual audio that is played back. This ensures that the user's intentions are communicated in a way that is easily understood by others.

[0689] (Application Example 1)

[0690] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0691] There is a need for support tools that enable users with communication difficulties to communicate smoothly in public places and face-to-face interactions. Current technology fails to fully utilize users' individual contextual information and preference history, making it difficult to provide timely and optimal responses to facilitate the flow of conversation. In particular, there is a high need for technology that effectively and naturally assists speech for individuals with selective mutism or other communication disorders.

[0692] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0693] In this invention, the server includes means for capturing user speech information, means for analyzing the speech information to extract conversational tendencies, and means for a generative model that generates response candidates based on current situation information and preference history. This makes it possible for users to be presented with responses effectively and quickly in face-to-face interactions, and for those responses to be communicated in a natural manner using speech synthesis technology.

[0694] "User" refers to an individual who receives communication support using this system.

[0695] "Speech information" refers to the content of a conversation expressed by the user in written or spoken form.

[0696] "Means of capture" refers to components that have the function of acquiring and storing speech information.

[0697] "Means of analysis" refers to techniques that analyze acquired speech information and extract the user's conversational tendencies and characteristics.

[0698] "Conversational tendencies" refer to the linguistic characteristics and speech patterns that users exhibit during conversations.

[0699] "Preference history" refers to data about responses and preferences that a user has selected in the past.

[0700] A "generative model" is a module that uses artificial intelligence technology to generate response candidates based on the user's current situation information and preference history.

[0701] A "response option" refers to a set of utterances presented to the user as choices they can use in a specific situation.

[0702] "Speech synthesis" is a technology that converts text information into natural-sounding speech.

[0703] "Means of communication to others" refers to technologies for effectively conveying synthesized speech responses to people other than the user.

[0704] This invention is a communication assistance system aimed at enabling users with selective mutism or other communication difficulties to communicate smoothly in face-to-face interactions. The system is implemented via a smartphone or similar user terminal.

[0705] The system's core process combines the acquisition, analysis, and generation of speech information, followed by speech synthesis output. Users input speech information via a terminal in either voice or text format. The terminal converts the voice data to text using speech recognition technology, encrypts the data, and then sends it to the server. Speech recognition technologies such as Google Cloud Speech-to-Text and other similar services are available.

[0706] The server analyzes the received data using natural language processing (NLP) techniques to extract the user's conversational tendencies. NLP tools such as NLTK and SpaCy are used here. Then, using a generative AI model such as GPT-4, appropriate response candidates are generated based on the user's past preference history and current situation. The generated responses are optimized based on important topics and the user's preferences and presented as multiple options.

[0707] On the device, generated response suggestions are displayed to the user. The response selected by the user is converted into natural-sounding speech using speech synthesis technology such as Google Text-to-Speech and communicated to the other party. This process facilitates smoother communication.

[0708] As a concrete example of use, consider a scenario where a user orders tea at a cafe. Based on the user's past purchase history, the system recommends a type of tea and generates a response such as, "I'd like Darjeeling tea, please." Once the user selects this, the terminal verbally communicates this to the staff.

[0709] As an example of a prompt, you can instruct the AI ​​as follows: "Generate a list of recommended teas based on the user's preferences and past purchase history. Please suggest three options, taking into account current preferences and trends."

[0710] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0711] Step 1:

[0712] The user inputs speech information into the device. The input format can be voice or text. In the case of voice input, the device uses Google Cloud Speech-to-Text to convert the voice data into text format. The input data is then sent to the next step either in its original format or as text.

[0713] Step 2:

[0714] The terminal encrypts the input data to store it securely. AES encryption technology is used here. This process protects user privacy while preparing the data for transmission to the next step. The encrypted data is output and sent to the server.

[0715] Step 3:

[0716] The server decrypts the received encrypted data and analyzes its contents. It uses natural language processing techniques to analyze the text data and extract user conversational trends and important topics. By utilizing tools such as NLTK and SpaCy, it outputs trend data from the input text.

[0717] Step 4:

[0718] The server uses a generative AI model, such as GPT-4, to generate response candidates that take into account the user's context and preference history. The generative AI model generates an appropriate response based on past data and the current prompt, and a list of response candidates is output.

[0719] Step 5:

[0720] The device presents the user with the received response options. The user selects the response that matches their intent. The user's selection is output, and the result is passed on to the next step.

[0721] Step 6:

[0722] The selected response is converted into natural-sounding speech using speech synthesis technology on the device. Text data is converted to speech data using tools such as Google Text-to-Speech. The converted speech is played through the device's speaker and transmitted to the other party.

[0723] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0724] The communication assistance system according to the present invention is designed to facilitate smooth communication in situations where it is difficult for the user to speak. This system enables more natural and appropriate communication by combining an emotion engine at each stage of speech data capture, analysis, response generation, and voice output.

[0725] First, as users use their devices, speech data is continuously captured. This data is entered in either text or voice format, and the device collects it. The collected data is encrypted and sent to the server.

[0726] The server analyzes the received speech data to identify the user's conversation patterns. Natural language processing techniques are used in this analysis to extract the user's everyday speaking tendencies and frequently occurring words. Simultaneously, an emotion engine recognizes emotions from the user's speech and selected words, and evaluates their emotional state. This information is also added to the user profile maintained by the server and learned as a comprehensive characteristic of the user.

[0727] Based on this data, the server generates response candidates using a generative AI model. During this process, the tone and content of the responses are adjusted to match the user's current emotional state, taking into account the user's emotional condition. For example, if the server determines that the user is stressed, responses that promote relaxation are prioritized.

[0728] The device presents the user with a list of generated response options. The user selects the response that best suits them through an intuitive interface. The selected response is then converted into speech, tailored to the context and emotions, and output in the user's natural voice. By using speech synthesis technology, the user's voice tone and emotions are also reflected, resulting in a more realistic conversation.

[0729] As a concrete example, consider a scenario where a user is giving a presentation in a public place. If the user shows signs of nervousness, the server can detect this using its emotion engine and offer a response such as, "Relax, you're doing great." If the user selects this option, the device will generate a voice in a comforting tone to support the user.

[0730] This invention is a system that supports users in maintaining the most appropriate and effective communication by performing comprehensive information processing, including emotions. This technology is useful not only for people with selective mutism but also for all people who require special communication support.

[0731] The following describes the processing flow.

[0732] Step 1:

[0733] The device captures voice or text data as soon as the user begins a conversation. This includes what the user says to the device and messages they type on the screen.

[0734] Step 2:

[0735] The device temporarily stores the collected speech data, and once it confirms that the communication environment is stable, it encrypts the data and sends it to the server. This data also includes emotional information based on the user's consent.

[0736] Step 3:

[0737] The server converts the received data into a format suitable for analysis, analyzes the conversation content using natural language processing techniques, and extracts the user's conversation patterns. This analysis includes frequently occurring words, the flow of the conversation, and the user's preferences.

[0738] Step 4:

[0739] The server uses an emotion engine to recognize the user's emotional state from their voice tone and selected words. For example, it analyzes the tone and tempo of the voice data to determine whether the user is tense or relaxed.

[0740] Step 5:

[0741] The server uses a generative AI model to generate response candidates, taking into account the user's conversation patterns and emotional state. This generation process utilizes the user's past conversation history and current contextual information to prepare multiple appropriate responses.

[0742] Step 6:

[0743] The device displays generated response options to the user. The UI is designed to be intuitive and easy for the user to operate, and sometimes a response that helps to ease tension is displayed at the top.

[0744] Step 7:

[0745] The user selects a response from the presented options that best matches their intentions and feelings. The selected response is optimized to suit the individual's emotions and situation.

[0746] Step 8:

[0747] The device uses a speech synthesis engine to output the selected response as speech that reflects the user's voice and tone. This makes it sound to others as if the user were speaking naturally.

[0748] Step 9:

[0749] Users can evaluate whether the outputted response was intended in the communication and provide feedback to the system.

[0750] This allows the system to use this feedback as an indicator to further improve the accuracy of future responses.

[0751] (Example 2)

[0752] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0753] In modern society, effective communication is required in a variety of situations, but users sometimes have difficulty conveying their intentions appropriately. In such situations, there is a need for technologies that support smooth dialogue tailored to the user's mental state and linguistic characteristics. In particular, responses that respond to emotional states are important, and there is a challenge in that support in this area is insufficient.

[0754] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0755] In this invention, the server includes a device for capturing user speech information, a device for analyzing the speech information to extract the user's language tendencies, and a generative AI mechanism for analyzing the user's emotional state and generating response candidates based on the language tendencies and emotional state. This enables appropriate and effective communication support that takes emotions into consideration.

[0756] "User speech information" refers to the content that users utter in voice or text format, and this information is collected as data.

[0757] "Linguistic tendencies" refer to specific language usage patterns and characteristics of frequently occurring words and phrases observed in users' conversations.

[0758] "Emotional state" refers to the psychological or emotional condition analyzed from the user's utterances and selected words.

[0759] A "generative AI mechanism" is a system that incorporates artificial intelligence technology used to generate appropriate responses based on the user's language tendencies and emotional state.

[0760] A "communication support system" refers to a device or program that provides a technical mechanism to assist users in communicating their intentions smoothly and appropriately.

[0761] This invention is a communication support system designed to provide more natural and appropriate responses when users communicate. The system aims to efficiently analyze speech information while considering the user's emotional state and generate appropriate responses.

[0762] First, the content spoken by the user, either by voice or text, is captured by the device. Here, voice processing software, commonly used as speech recognition technology, is used to convert the voice data into text. The text data is temporarily stored on the device and then transmitted to the server using encryption technology to ensure security.

[0763] The server utilizes natural language processing techniques to analyze the received text data. Here, general analysis software is used to extract the user's language tendencies. Simultaneously, a sentiment analysis engine is employed to evaluate the user's emotional state. These evaluation results are stored as individual user profiles.

[0764] Based on the analysis results and emotional state, the server generates response candidates using a generative AI model. The generative AI model enables natural dialogue that matches the user's psychological state. An example of a prompt would be, "Generate a natural, relaxing response for when the user is nervous during a presentation."

[0765] The generated responses are presented to the user by the device, and the user can select one through an intuitive interface. The selected response is then output naturally in the user's voice using speech synthesis technology. This enables the user to engage in smooth, emotionally sensitive communication.

[0766] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0767] Step 1:

[0768] The device receives voice or text input from the user. In the case of voice input, the device uses speech recognition software to convert the voice data into text. The resulting text data is then output. The speech recognition process includes an acoustic model and a language model for converting voice samples into text.

[0769] Step 2:

[0770] The terminal encrypts the generated text data and sends it to the server. It receives text data as input, protects it using an encryption algorithm, and outputs encrypted data. This ensures confidentiality during data transfer.

[0771] Step 3:

[0772] The server receives encrypted text data, decrypts it, and returns it to its original text data. This decrypted data becomes the input, and raw text ready for analysis is output.

[0773] Step 4:

[0774] The server analyzes the decoded text data using natural language processing techniques. This extracts the user's language tendencies and the intent expressed in the text. The input for this step is raw text, and the output is the analyzed language features and conversational intent.

[0775] Step 5:

[0776] The server uses a sentiment analysis engine to evaluate the user's emotional state. The input is pre-analyzed text, and the output is the detected emotional state and psychological tendencies. This allows for the clear identification of subtle emotions hidden within the text.

[0777] Step 6:

[0778] The server uses a generative AI model to generate response candidates based on the analysis results and emotional state. The generative AI model takes linguistic features and emotional state as input and generates emotionally sensitive response candidates as output. This process creates natural-sounding utterances that are appropriate to the user's current psychological state.

[0779] Step 7:

[0780] The terminal presents the user with generated response options. This can be done by displaying the response on the screen or by providing voice guidance. The input here is response data from the server, which is output as choices for the user to select.

[0781] Step 8:

[0782] The user selects the most appropriate response from the presented options. The selected response is recognized as input, and the next action is planned accordingly.

[0783] Step 9:

[0784] The device outputs the selected response in speech format using speech synthesis technology. The input here is the selected text response, and the output is synthesized speech with a tone similar to the user's voice. This enables more natural communication.

[0785] (Application Example 2)

[0786] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0787] Conventional communication assistance systems have the problem of being unable to conduct conversations that fully consider the user's emotions, and thus are unable to communicate naturally and smoothly. Furthermore, in customer service at physical stores, it is necessary to appropriately understand the customer's emotions and respond accordingly, but there has been no effective technology to support this.

[0788] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0789] In this invention, the server includes means for capturing user speech data, means for analyzing the speech data to extract the user's conversation patterns, and means for a generative model that generates response candidates based on the user's current contextual information and emotional state. This makes it possible to appropriately grasp the customer's emotions and support optimal communication in interactions with customers in physical stores.

[0790] "User speech data" refers to the audio and text information that users produce when communicating.

[0791] "Conversation patterns" refer to the collective characteristics and tendencies extracted from conversations a user has had in the past.

[0792] "Contextual information" refers to information that includes the situation and background in which an utterance is made, as well as the emotional state of the speaker.

[0793] A "generative model" refers to a structure or algorithm used to generate a response based on input data, employing natural language processing techniques.

[0794] A "display device" refers to an electronic device used to visually present generated information to a user.

[0795] "Speech synthesis" is a technology that mechanically generates human speech from text.

[0796] "Emotional state" refers to the psychological state that can be inferred from the user's statements and actions.

[0797] To implement this invention, a terminal device, such as smart glasses worn by the user, and a server connected to it are required. The terminal device is equipped with a camera and microphone, and has the function to transmit real-time data acquired by these devices to the server. The server receives the user's speech data and facial expression data and uses natural language processing technology and facial recognition technology to analyze them. Specifically, it is common to use Google Cloud Speech-to-Text API or OpenCV.

[0798] The server first converts the acquired audio data into text, and then extracts the user's conversation patterns and emotional state from the text data and facial expression data. Based on this information, a generative AI model generates the optimal response. The generated response is then presented to the terminal, with its tone and selected language adjusted to match the user's emotional state. The terminal visually communicates this response to the user and provides audio feedback as needed.

[0799] For example, in a physical store, if a customer's anxious expression or speech is detected during customer service, a suggestion such as "Please let us know if you need any assistance" will be displayed on the staff member's smart glasses. This allows staff members to provide more appropriate and effective customer service.

[0800] A concrete example of a prompt for a generative AI model is: "Generate the most appropriate response when the customer looks anxious. Choose the right words based on the user's emotional state." This prompt allows the system to provide a response appropriate to the emotional state, improving the quality of communication.

[0801] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0802] Step 1:

[0803] The device captures the user's speech and facial expressions in real time using its camera and microphone. The input is the user's voice and video data, and the output is the transmission of this data to a server. Specifically, sensors within the device record the voice, and the camera captures the facial expressions.

[0804] Step 2:

[0805] The server converts the received audio data into text. This process uses the Google Cloud Speech-to-Text API. The input is audio data, and the output is text data. The operation here is to process the audio signal as a text string.

[0806] Step 3:

[0807] The server analyzes facial expression data to infer the user's emotions. It uses OpenCV to detect facial features and then analyzes them with an emotion engine. The input is facial image data, and the output is the inferred emotional state. Specifically, it measures changes and patterns in facial expressions and assigns emotion labels.

[0808] Step 4:

[0809] The server uses natural language processing technology to extract conversation patterns from text data. A generative AI model then generates response candidates based on this. The input is text data and emotion states, and the output is the text of the response candidates. In operation, prompt sentences are sent to the generative AI model, which automatically generates responses.

[0810] Step 5:

[0811] The terminal presents the user with generated response options visually or audibly. The user selects the most appropriate response, which is then output via speech synthesis. The input is the text data of the response options, and the output is either an audio message or text on a display. The operation here is to display the response on a visual device or play the audio through a speaker.

[0812] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0813] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0814] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0815] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0816] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0817] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0818] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0819] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0820] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0821] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0822] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0823] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0824] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0825] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[0826] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0827] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0828] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0829] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0830] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0831] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0832] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[0833] The following is further disclosed regarding the embodiments described above.

[0834] (Claim 1)

[0835] A means of capturing user speech data,

[0836] A means for analyzing the aforementioned speech data to extract the user's conversation patterns,

[0837] A means comprising a generative model that generates response candidates based on the user's current contextual information,

[0838] A means of presenting the generated response candidates to the user,

[0839] A means for outputting the aforementioned response candidates as audio based on the user's selection,

[0840] A system that includes this.

[0841] (Claim 2)

[0842] The system according to claim 1, wherein the capture of the speech data includes recording a conversation made by a user in text or voice format.

[0843] (Claim 3)

[0844] The system according to claim 1, wherein the generation model has a function to adjust the appropriateness of the response considering the user's conversation patterns and past preference history.

[0845] "Example 1"

[0846] (Claim 1)

[0847] A device for acquiring speech information,

[0848] A device that analyzes the aforementioned speech information to extract the user's dialogue tendencies,

[0849] A device that generates response candidates using a generative artificial intelligence model based on analysis results,

[0850] A device that supplies the generated response candidates to the user,

[0851] A device that converts the aforementioned response candidates into speech based on the user's selection,

[0852] ...

[0853] A system that includes this.

[0854] (Claim 2)

[0855] The system according to claim 1, wherein the acquisition of the utterance information includes recording communications made by the user in text or voice format.

[0856] (Claim 3)

[0857] The system according to claim 1, wherein the generating artificial intelligence model has a function to adjust the appropriateness of the response, taking into account the user's conversational tendencies and past preference history.

[0858] "Application Example 1"

[0859] (Claim 1)

[0860] Means for capturing user speech information,

[0861] A means for analyzing the aforementioned speech information to extract the user's conversational tendencies,

[0862] A means comprising a generative model that generates response candidates based on the user's current status information and preference history,

[0863] A means of presenting the generated response candidates to the user,

[0864] A means for outputting the aforementioned response candidates as audio based on the user's selection,

[0865] A means of transmitting a response to another person using speech synthesis via a user terminal,

[0866] A system that includes this.

[0867] (Claim 2)

[0868] The system according to claim 1, wherein the capture of the speech information includes recording a dialogue performed by the user in text or voice format.

[0869] (Claim 3)

[0870] The system according to claim 1, wherein the generation model has a function to adjust the appropriateness of the response considering the user's conversational tendencies and past preference history, and has a function to transmit the response to others by speech synthesis.

[0871] "Example 2 of combining an emotion engine"

[0872] (Claim 1)

[0873] A device that captures user speech information,

[0874] A device that analyzes the aforementioned speech information to extract the user's language tendencies,

[0875] A generative AI mechanism that analyzes the user's emotional state and generates response candidates based on the aforementioned language tendencies and emotional state,

[0876] A device that presents the generated response candidates to the user,

[0877] A device that outputs the aforementioned response candidates as voice based on the user's selection,

[0878] A communication support system including this.

[0879] (Claim 2)

[0880] The communication support system according to claim 1, wherein the capture of the utterance information includes recording a conversation conducted by a user in voice or text format.

[0881] (Claim 3)

[0882] The communication support system according to claim 1, wherein the generation AI mechanism has a function to adjust the suitability of the response considering the user's language tendencies and past preference information.

[0883] "Application example 2 of combining emotional engines"

[0884] (Claim 1)

[0885] A means of capturing user speech data,

[0886] A means for analyzing the aforementioned speech data to extract the user's conversation patterns,

[0887] A means comprising a generative model that generates response candidates based on the user's current contextual information and emotional state,

[0888] A means for using a display device that presents generated response candidates to the user and suggests the optimal response according to the dialogue environment,

[0889] A means for outputting the aforementioned response candidates as voice based on the user's selection and using speech synthesis to reflect the speaker's emotions,

[0890] A system that includes this.

[0891] (Claim 2)

[0892] The system according to claim 1, wherein the capture of the speech data includes recording conversations made by the user in text or voice format, and detecting the user's facial expressions and voice characteristics.

[0893] (Claim 3)

[0894] The system according to claim 1, wherein the generation model has a function to adjust the appropriateness of the response by taking into account the user's conversation patterns and past preference history, as well as detected emotional information. [Explanation of Symbols]

[0895] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of capturing user speech data, A means for analyzing the aforementioned speech data to extract the user's conversation patterns, A means comprising a generative model that generates response candidates based on the user's current contextual information, A means of presenting the generated response candidates to the user, A means for outputting the aforementioned response candidates as audio based on the user's selection, A system that includes this.

2. The system according to claim 1, wherein the capture of the speech data includes recording a conversation made by a user in text or voice format.

3. The system according to claim 1, wherein the generation model has a function to adjust the appropriateness of the response considering the user's conversation patterns and past preference history.

Images