System

Abstract

A system is provided.SOLUTION: Receiving, by a server, a call request including character selection information from one or more terminals, and sending first time scenario information to a generation AI, generating, by the generation AI, conversation text based on the scenario information, converting, by the server, the generated conversation text into voice, and sending, by the server, a voice file to the one or more terminals; A system comprising: means for a terminal to play back a voice; means for a user to make a reply to the terminal, to convert the reply into a text, and to retransmit the text to a server; and means for the server to pass next scenario data to a generated AI based on the retransmitted text.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the 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] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] With current technology, it is difficult to enjoy real-time interactive conversations with famous animal characters. Furthermore, existing conversation systems often use scripted responses, preventing users from experiencing fresh, natural conversations. Children, in particular, lack the means to enjoy direct conversations with familiar characters. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by providing a system including: means for a user to select a character and make a call request using a terminal; means for a server to receive the call request including the character selection information from the terminal and send initial scenario data to a generation AI; means for the generation AI to generate conversation text based on the scenario data; means for the server to convert the generated conversation text into speech; means for sending an audio file from the server to the terminal and for the terminal to play the audio; means for the user to respond to the terminal, convert the response into text, and resend the text to the server; and means for the server to pass the next scenario data to the generation AI based on the resent text. This system allows users to enjoy natural conversations in real time with famous animal characters.

[0006] A "terminal" is a communication device used by a user, such as a smartphone or tablet.

[0007] The "server" is a central system that receives requests from terminals and processes them using the generation AI and text-to-speech engine.

[0008] A "user" is someone who operates a terminal and enjoys conversations with characters.

[0009] "Character selection information" is information for identifying a specific animal character that the user has selected as a conversation partner.

[0010] A "call request" is a request sent from a terminal to a server to start a conversation with a character selected by the user.

[0011] "Initial scenario data" is data that includes basic conversation content for starting the first conversation with a character.

[0012] "Generative AI" is artificial intelligence that generates natural conversational text based on received scenario data.

[0013] "Conversational text" is a string of characters that indicates what is being said by a character generated by the generation AI.

[0014] A "text-to-speech engine (TTS engine)" is a technology that converts text data into voice data.

[0015] "Audio File" means digital audio data containing the voice of a character generated by a text-to-speech engine.

[0016] "Audio playback" refers to the operation of outputting an audio file received by a terminal so that the user can hear it.

[0017] A "reply" is a voice response given by the user to a character's statement.

[0018] A "voice recognition engine" is a technology that converts a user's speech into text data.

[0019] "Scenario data" is a data set for controlling the content of conversations with characters. [Brief explanation of the drawings]

[0020] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0021] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0022] First, the terms used in the following description will be explained.

[0023] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0024] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0025] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0026] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, 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), Bluetooth (registered trademark), etc.

[0027] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0028] [First embodiment]

[0029] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0030] 1, a 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.

[0031] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0032] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0033] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the 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.

[0034] 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 of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0035] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0036] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0037] 2, in the data processing device 12, a specific process 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" according to the technology of the present 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 process 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.

[0038] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0039] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0040] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0041] This invention is a system that allows users to use a terminal to have real-time telephone conversations with famous animal characters. Below, we will explain the program processing of this system and specific examples.

[0042] Character selection and call request

[0043] First, the user opens the dedicated app on their device. Within the app, they select the animal character they want to talk to. Next, when the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. This request also includes the identification information of the selected character.

[0044] Receiving a call request and generating a first conversation scenario

[0045] The server receives a call request from the device and analyzes the request. Based on the character information obtained as a result of the analysis, the server passes the initial conversation scenario data to the generation AI. This scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[0046] Conversational text generation

[0047] The AI ​​generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing the user with natural and familiar conversations.

[0048] Conversational text-to-speech

[0049] The server receives the generated conversational text from the AI ​​generator and converts it into audio using a text-to-speech (TTS) engine. The converted audio data is then tuned to play in the character's unique voice.

[0050] Sending and playing audio files

[0051] The server sends the converted audio file to the device, which then receives the file and plays it for the user to listen to. The user can enjoy the conversation while listening to the character's voice.

[0052] Handling the user's response

[0053] The user listens to the conversation and responds. The device records the user's response and converts it into text data using a speech recognition engine. This text data is then sent back to the server.

[0054] Next conversation generation process

[0055] The server receives the text data sent by the user and passes it to the generation AI as the next conversation scenario data. The generation AI generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[0056] Specific examples

[0057] For example, suppose a user asks, "What did you do today?" The server passes this question to the generation AI, which then generates a response such as, "I took a walk in the park today. I could hear lots of birds and it was very pleasant." This response is then converted into audio by the text-to-speech engine and sent to and played back on the device. The user can then listen to this and continue the conversation.

[0058] In this way, the system allows users to have natural conversations in real time with their chosen animal characters, providing an engaging and relatable experience especially for children.

[0059] The processing flow will be explained below.

[0060] Step 1:

[0061] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[0062] Step 2:

[0063] The server analyzes the call request received from the device and extracts the selected character information.The server then sends the initial conversation scenario data to the generation AI, instructing it to generate an opening greeting and introductory sentence.

[0064] Step 3:

[0065] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[0066] Step 4:

[0067] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[0068] Step 5:

[0069] The server sends the converted audio file to the terminal, which then plays the received audio file for the user to listen to.

[0070] Step 6:

[0071] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[0072] Step 7:

[0073] The device resends the converted text data to the server, which analyzes the received user response and passes the next conversation scenario data based on that to the generation AI.

[0074] Step 8:

[0075] The generative AI generates new conversational text based on the user's response. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[0076] Step 9:

[0077] The server converts the generated new conversation text into voice data using the text-to-speech engine again, and transmits the voice data to the terminal.

[0078] Step 10:

[0079] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character.

[0080] Example 1

[0081] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0082] Conventional systems have struggled to enable users to have natural, real-time conversations with their chosen characters, resulting in limited user experience. Furthermore, character voice generation and user response processing often lacked smoothness, resulting in delayed conversations and unnatural responses. These issues pose a significant obstacle, particularly for children, where familiarity and naturalness are crucial.

[0083] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0084] In this invention, the server includes: means for a user to use a terminal to select a character and make a call request; means for the terminal to send a call request including character selection information to the server; means for the server to receive and analyze the call request; means for the server to send initial scenario data to the generation AI; means for the generation AI to generate conversation text based on the initial scenario data; means for the server to convert the generated conversation text into voice; means for the server to send the converted voice file to the terminal and the terminal to play the voice; means for the user to respond to the terminal, convert the response into text using a voice recognition engine, and resend it to the server; and means for the server to send next scenario data to the generation AI based on the resent text. This enables users to have natural conversations in real time with their selected characters, providing an experience that is particularly familiar to children.

[0085] "User" refers to a person who uses a terminal to select a character and make a call request.

[0086] A "terminal" is a device operated by a user, which transmits a call request including character selection information to a server and plays back audio received from the server.

[0087] "Character selection information" refers to information for identifying a specific character selected by a user.

[0088] "Server" refers to a computer system on a network that receives and analyzes call requests sent from a device and works in conjunction with the generation AI and voice conversion function.

[0089] A "call request" refers to a request to the server to initiate a real-time conversation, including identification information for a user-selected character.

[0090] "Initial scenario data" refers to the data that defines the initial conversation scenario that the server sends to the generation AI.

[0091] "Generative AI" refers to an artificial intelligence model that generates conversational text based on received scenario data.

[0092] "Conversational text" refers to text that describes what a character will say to the user, generated by the generation AI based on scenario data.

[0093] "Convert to speech" refers to converting the conversational text generated by the generative AI into audio data using a text-to-speech engine.

[0094] "Speech recognition engine" refers to software or algorithms for converting a user's spoken responses into text data.

[0095] "Scenario data" refers to data that is passed to the generation AI to generate the content of the conversation.

[0096] This invention is a system that allows users to have natural conversations in real time with selected animal characters using a terminal. The core of this system is composed of a server, a terminal, and a generative AI model working together. Specific embodiments of the system are described below.

[0097] First, the user uses a device such as a smartphone or tablet. The user launches a dedicated app and selects the animal character they want to talk to from the displayed characters. For example, characters such as rabbits and bears can be selected. After selecting the character, the user presses the "Start Call" button. This action causes the device to package a call request in JSON format, including the identification information of the selected character, and send it to the server as an HTTP POST request.

[0098] When the server receives a call request, it analyzes the request and extracts the identification information of the selected character. The server then generates initial conversation scenario data and passes it to a generative AI model (e.g., OpenAI's GPT-4). The initial scenario data includes a prompt such as, "Hello, this is AA. What would you like to talk about today?"

[0099] The generative AI model generates dialogue text based on scenario data, imitating the character's personality and speaking style. The dialogue text is then sent to a server, which converts the received text into audio data using a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech). This audio data is then adjusted to be played in the character's unique voice.

[0100] The server sends the converted audio file to the device, which receives and plays it. The user can then hear the character's voice through the device and begin the conversation.

[0101] When the user responds to the device, the device records the voice and converts it into text data using an internal speech recognition engine (e.g., Google Cloud Speech-to-Text). The converted text data is then sent back to the server.

[0102] The server receives the text data sent by the user and passes it to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[0103] As a concrete example, consider the case where a user asks, "What did you do today?" Below is an example of a prompt sentence to input to the generation AI.

[0104] "Generate a conversation scenario in which the user asks, 'What did you do today?' and the character replies that he was walking in the park."

[0105] Based on this prompt, the generative AI model generates a response text such as, "I was walking in the park today. I could hear lots of birds singing and it felt great." This response text is converted into speech via the server and sent and played back on the device, allowing the user to enjoy a conversation with the character.

[0106] As a result, the system allows users to have natural conversations with selected animal characters in real time, providing an experience that is particularly friendly to children.

[0107] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0108] Step 1:

[0109] The user launches the dedicated app and selects an animal character.

[0110] Input: A device such as a smartphone or tablet operated by the user.

[0111] Specific operation: The user checks the list of available animal characters on the device's app screen and taps the desired character to select it.

[0112] Output: The identity of the selected character.

[0113] Step 2:

[0114] The user presses the "start call" button.

[0115] Input: The user-selected character's identity.

[0116] Specific operation: The user taps the "Start Call" button on the app screen. This action prepares the device to perform the following process.

[0117] Output: Call request trigger.

[0118] Step 3:

[0119] The terminal transmits a call request including character selection information to the server.

[0120] Input: The identity of the selected character, a trigger indicating that the "Start Call" button has been pressed.

[0121] Specific operation: The device creates a call request in JSON format including the identification information of the selected character and sends it to the server as an HTTP POST request.

[0122] Output: The call request data sent to the server.

[0123] Step 4:

[0124] The server receives and analyzes the call request.

[0125] Input: Call request data sent from the device.

[0126] Specific operation: The server receives the call request, parses the JSON data to extract the character's identification information, confirms the selection information, and prepares for the next process.

[0127] Output: Parsed character identities.

[0128] Step 5:

[0129] The server sends the initial scenario data to the generation AI.

[0130] Input: Parsed character identity.

[0131] Specific operation: The server generates the initial conversation scenario data (e.g., "Hello, this is AA. What would you like to talk about today?") and sends it to the generation AI as a prompt.

[0132] Output: Initial scenario data sent to the generation AI.

[0133] Step 6:

[0134] The generation AI generates conversational text based on the initial scenario data.

[0135] Input: Initial scenario data sent to the generating AI.

[0136] How it works: A generative AI model (e.g., OpenAI GPT-4) generates an initial conversation text based on the received prompt, imitating the character's personality and speaking style.

[0137] Output: The generated conversation text.

[0138] Step 7:

[0139] The server converts the generated conversation text into speech.

[0140] Input: The generated conversation text.

[0141] How it works: The server passes the generated conversation text to a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech) and converts it into audio data. The settings are adjusted to create the character's unique voice tone.

[0142] Output: The converted audio data.

[0143] Step 8:

[0144] The server sends the converted audio file to the terminal, which then plays the audio.

[0145] Input: The converted audio data.

[0146] Specific operation: The server transfers the generated audio file to the device, sending the audio data as an HTTP response. The device then plays the received audio file, allowing the user to hear the character's voice.

[0147] Output: The audio the user hears.

[0148] Step 9:

[0149] The user responds.

[0150] Input: The character's voice as heard by the user on their device.

[0151] Specific operation: After listening to the character's voice, the user responds.

[0152] Output: The user's spoken response.

[0153] Step 10:

[0154] The device records the user's response and converts it into text using a voice recognition engine.

[0155] Input: The user's voice reply.

[0156] Specific operation: The device records the user's voice and converts it into text data using an internal voice recognition engine (e.g., Google Cloud Speech-to-Text).

[0157] Output: User responses converted to text data.

[0158] Step 11:

[0159] The terminal transmits the converted text data to the server.

[0160] Input: User responses converted to text data.

[0161] Specific operation: The terminal packages the converted text data in JSON format and sends it to the server as an HTTP POST request.

[0162] Output: The text data of the user response sent to the server.

[0163] Step 12:

[0164] The server sends the next scenario data to the generation AI based on the resent text.

[0165] Input: Text data of the user's response sent from the terminal.

[0166] Specific operation: The server receives the user's response text and passes it to the generation AI as the next conversation scenario. It creates a prompt sentence to input to the generation AI.

[0167] Output: Send the following conversation scenario data to the generation AI.

[0168] Step 13:

[0169] The generation AI will generate the following conversation text again.

[0170] Input: The prompt sent to the generation AI as the following scenario data.

[0171] How it works: The AI ​​generates the next conversation text based on the user's response. This process is repeated until the user ends the conversation.

[0172] Output: The following conversation text.

[0173] Through this series of steps, users can enjoy natural conversations in real time with their chosen animal character.

[0174] (Application example 1)

[0175] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0176] In conventional character conversation systems, conversations between users and characters are often scripted, making it difficult to provide natural conversations in real time. Furthermore, there was a lack of technology to generate appropriate responses to user voice input and play them back in a natural voice, limiting the user experience.

[0177] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0178] In this invention, the server includes: means for a user to select a character and make a call request using an information processing device; means for the server to receive a call request including character selection information from the information processing device and transmit first conversation scenario data to the generation AI model; means for the generation AI model to generate conversation text based on the scenario data; means for the server to convert the generated conversation text into voice data; means for transmitting voice data from the server to the information processing device and for the information processing device to play the voice; means for the user to respond to the information processing device, converting the response into text data and retransmitting it to the server; means for the server to pass next conversation scenario data to the generation AI model based on the retransmitted text data; and means for converting the conversation text of the generation AI model into voice using a text-to-speech engine and providing the conversation data in real time, thereby enabling users to enjoy natural conversations in real time.

[0179] "Information processing device" is a general term for electronic devices used for processing and communicating data, and specifically includes smartphones, tablets, personal computers, etc.

[0180] A "server" is a central computer system that receives requests from multiple terminals on a network and provides data processing and communication services.

[0181] "Character selection information" is data for specifying the character with which the user wishes to converse, and this data is received by the server.

[0182] A "call request" is a request made by a user to start a conversation with a selected character using an information processing device.

[0183] "First conversation scenario data" is conversation script data generated to smoothly start the first conversation between a character and a user.

[0184] A "generative AI model" is an artificial intelligence model used to generate natural-sounding conversational text, using machine learning to generate responses in response to user input.

[0185] "Conversation text" is text data generated by a generative AI model that indicates the content of a conversation between a user and a character.

[0186] "Audio data" refers to an audio file generated from text data using speech synthesis technology, and is played back by an information processing device.

[0187] "Text data" refers to character information to be displayed or processed on an information processing device, and includes user responses and generated conversation content.

[0188] A "text-to-speech engine" is a software component that converts text data into audio data, enabling playback in a natural voice.

[0189] "Conversation scenario data" is data that indicates the next conversation content that the generative AI model will generate based on user input, enabling continuous conversation between the character and the user.

[0190] "Real-time" means that the conversation between the user and the character is immediate, with natural responses provided without delay.

[0191] A specific embodiment of the present invention will be described below: This system allows a user to use an information processing device to enjoy natural conversations with a specific character in real time.

[0192] First, the user launches a dedicated application using an information processing device (e.g., a smartphone, tablet, or head-mounted display (HMD)). Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button, which sends a call request to the server. The call request includes the identification information of the character selected by the user.

[0193] The server receives the call request and analyzes its content. Based on the analysis results, the server passes the initial conversation scenario data to the generative AI model. This initial conversation scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[0194] A generative AI model (e.g., GPT-3) generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing a natural and familiar conversation to the user. This conversational text is sent to a server and converted into audio data using a text-to-speech engine (TTS engine).

[0195] The voice data is sent from the server to the information processing device, which then plays back the voice data. This allows the user to hear the character's voice. The user listens to the conversation and responds. The information processing device records the user's response and converts the voice into text data using a voice recognition engine (e.g., Google Speech API). This text data is then sent back to the server.

[0196] The server receives the text data sent by the user and passes it back to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, which the server converts again into voice data using a TTS engine and sends and plays back to the information processing device. This process is repeated, allowing for a natural conversation between the user and the character to continue.

[0197] As a concrete example, suppose a user asks, "What did you do today?" The server passes this question to a generative AI model, which generates a response such as, "I took a walk in the park today. I was pleasantly surprised to hear the birds singing." This response is converted by a TTS engine, sent to an information processing device, and played back. The user can listen to this and continue the conversation.

[0198] Examples of prompts include:

[0199] User Input:

[0200] User: What did you do today?

[0201] Response from the generative AI model:

[0202] Character: I was walking in the park today. I could hear a lot of birds singing and it felt great.

[0203] The system allows users to engage in natural conversations with selected characters in real time, providing a familiar and engaging experience especially for children.

[0204] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0205] Step 1:

[0206] The user launches a dedicated application using an information processing device. Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button. Identification information for the character selected by the user is then generated and sent to the server as a call request. The input is the user's character selection, and the output is the call request data.

[0207] Step 2:

[0208] The server receives a call request sent from the user's information processing device. The server then analyzes the request content and extracts the character's identification information. The input is the call request data, and the output is the character's identification information.

[0209] Step 3:

[0210] The server passes the first conversation scenario data to the generation AI model based on the character's identification information. This scenario data includes the content of the first conversation based on the character's personality and speaking style. The input is the character's identification information, and the output is the first conversation scenario data.

[0211] Step 4:

[0212] The generative AI model generates conversational text based on the received initial conversation scenario data. For example, it generates the content "Hello, this is AA. What would you like to talk about today?" The input is the initial conversation scenario data, and the output is the conversational text.

[0213] Step 5:

[0214] The server receives the conversational text from the generative AI model and converts it into audio data using a text-to-speech engine (TTS engine), where the character's unique voice is reproduced. The input is the conversational text, and the output is audio data.

[0215] Step 6:

[0216] The server transmits the generated voice data to the information processing device, which receives the voice data and plays it back for the user to hear. The input is the voice data, and the output is the situation in which the user can hear the voice.

[0217] Step 7:

[0218] The user listens to the character's voice and responds. The information processing device records the user's response and converts it into text data using a voice recognition engine. The input is the user's voice, and the output is text data.

[0219] Step 8:

[0220] The information processing device sends the user's response text data to the server. The server receives this text data and passes it to the generative AI model as the next conversation scenario data. The input is the user's response text data, and the output is the next conversation scenario data.

[0221] Step 9:

[0222] The generative AI model generates new conversation text based on the next conversation scenario data. The generated conversation text is sent to the server again. The input is the next conversation scenario data, and the output is the new conversation text.

[0223] Step 10:

[0224] The processes from step 5 to step 9 are repeated to continue a natural conversation between the user and the character, thereby maintaining a natural conversation in real time.

[0225] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0226] This invention is a system that enables a user to use a terminal to have real-time telephone conversations with famous animal characters, and that recognizes the user's emotions and adjusts the content of the conversation.

[0227] Character selection and call request

[0228] First, the user opens the app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. The selection information includes the character specified by the user.

[0229] Receiving a call request and generating a first conversation scenario

[0230] The server analyzes the call request received from the device and extracts the selected character information. Based on this information, the server sends the initial conversation scenario data to the generation AI, instructing it to generate an initial greeting such as, "Hello, this is XX. What would you like to talk about today?"

[0231] Conversational text generation and speech conversion

[0232] The generation AI generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. The server receives this dialogue text and converts it into audio data using a text-to-speech (TTS) engine. The converted audio data is generated in the character's voice.

[0233] Sending and playing audio files

[0234] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user, who then enjoys the conversation while listening to the animal characters' voices.

[0235] Processing user responses and sentiment analysis

[0236] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine. The converted text data is then resent to the server, where the emotion engine extracts and analyzes emotional information from the user's speech and text.

[0237] The process of emotion-based conversation generation

[0238] The server receives the text data sent by the user and the analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI. The generation AI generates conversation text according to the user's emotions. For example, if the user replies in a happy voice, "I had a great time at the picnic in the park today," the generation AI will generate a conversation such as, "That's great! Who were your friends playing with?"

[0239] Sending and playing audio files

[0240] The server then converts the newly generated conversation text into audio data using a text-to-speech engine and sends it to the device. The device then plays the received audio file back to the user. This process is repeated, allowing for a natural conversation between the user and the character to continue, and providing responses that correspond to the user's emotions.

[0241] Specific examples

[0242] For example, suppose a user says aloud to a device, "I was so happy my teacher praised me at school today!" The device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a text-to-speech engine and sent to the device for playback.

[0243] In this way, the system, combined with the emotion engine, enables more natural and friendly conversations based on the user's emotions, making interactions with characters that understand and respond to emotions even more appealing, especially for children.

[0244] The processing flow will be explained below.

[0245] Step 1:

[0246] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[0247] Step 2:

[0248] The server analyzes the call request received from the device and extracts the selected character information. Based on the character information obtained from the analysis, the initial conversation scenario data is provided to the generation AI.

[0249] Step 3:

[0250] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[0251] Step 4:

[0252] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[0253] Step 5:

[0254] The server sends the converted audio file to the device, which then plays the received audio file, allowing the user to hear the character's voice.

[0255] Step 6:

[0256] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[0257] Step 7:

[0258] The device retransmits the text data to the server, and simultaneously transmits the recorded data to the server, where the emotion engine extracts and analyzes emotional information from the user's speech.

[0259] Step 8:

[0260] The server provides the next conversation scenario data to the generation AI based on the received text data and emotional information. The generation AI generates the optimal response text based on the user's emotions analyzed by the emotion engine.

[0261] Step 9:

[0262] The generative AI generates new conversational text based on the user's responses and emotional information. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[0263] Step 10:

[0264] The server converts the generated new conversation text into voice data using a text-to-speech (TTS) engine again and sends it to the terminal.

[0265] Step 11:

[0266] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character, while providing responses that correspond to the user's emotions.

[0267] Step 12:

[0268] As a concrete example, if a user says to a device, "I was so happy when my teacher praised me at school today!", the device records this voice and converts it into text using a speech recognition engine. The server then passes this text data and the emotional information of "happy" analyzed by the emotion engine to the generation AI, which then generates a response such as, "That's wonderful! What did your teacher praise you for?" This response is converted into audio and played on the device.

[0269] Example 2

[0270] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0271] Conventional conversational systems have the problem that the conversation between the user and the character is formulaic and lacks responses that reflect the user's emotions. In particular, there is a need for a system that can recognize the user's emotions and generate natural conversations accordingly.

[0272] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0273] In this invention, the server includes a means for the user to respond to the terminal, converting the response into text and retransmitting it to the server, a means for the server to pass the next scenario data to the generation AI based on the retransmitted text, a means for the terminal to record the user's response and convert it into text data using a voice recognition engine, and a means for the server to extract and analyze the user's emotional information using an emotion analysis engine, thereby enabling a response that corresponds to the user's emotions.

[0274] A "terminal" is an electronic device such as a computer or smartphone that is operated by a user.

[0275] A "server" is a computer system that receives requests and data from terminals, processes them, and gives instructions to the generating AI.

[0276] "Generative AI" is an artificial intelligence model that generates conversational text based on scenario data.

[0277] "Character selection information" is information about the animal character with which the user wants to converse.

[0278] A "call request" is request data that a user sends from a terminal to a server in order to start a call with a character.

[0279] "Scenario data" is the initial set of information that the generation AI uses to generate conversational text.

[0280] "Conversational text" is a sentence generated by the generation AI based on scenario data.

[0281] "Audio data" refers to an audio file converted from text by a text-to-speech engine.

[0282] An "audio file" is a digital file containing audio data that is sent from a server to a terminal.

[0283] A "voice recognition engine" is software that records a user's speech and converts it into text data.

[0284] An "emotion analysis engine" is software that extracts and analyzes emotional information from a user's speech or text.

[0285] A "text-to-speech engine" is software that converts text data into voice data.

[0286] This invention is a system that allows users to use their terminals to have real-time telephone conversations with famous animal characters. Furthermore, it can recognize the user's emotions and adjust the content of the conversation. The detailed implementation method of this system is described below.

[0287] Hardware and software used

[0288] Device: A device operated by a user, such as a smartphone or tablet.

[0289] Server: A computer system that processes user requests and interacts with generative AI models.

[0290] Generative AI model: An artificial intelligence model for generating conversational text.

[0291] Speech recognition engine: Software that records what a user says and converts it into text data (e.g., Google Speech-to-Text).

[0292] Sentiment analysis engine: Software that extracts and analyzes users' emotional information (e.g., IBM Watson Tone Analyzer).

[0293] Text-to-speech engine: Software that converts text data into speech data (e.g., Amazon Polly).

[0294] Specific operation of the system

[0295] 1. Select a character and send a call request:

[0296] Users use their device to open a dedicated app and select the animal character they want to talk to.

[0297] When the user presses the "start call" button, a call request including character selection information is sent from the terminal to the server.

[0298] 2. Generating the first conversation scenario:

[0299] The server analyzes the call request received from the terminal and extracts the selected character information.

[0300] The server sends the initial conversation scenario data to the generation AI and instructs it to generate the initial greeting.

[0301] 3. Conversational text generation and speech conversion:

[0302] The generation AI generates conversational text based on the initial scenario data.

[0303] The server receives the generated text and converts it into audio data using a text-to-speech engine.

[0304] 4. Send and play audio files:

[0305] The server sends the converted audio file to the terminal.

[0306] The terminal receives the audio file and plays it for the user.

[0307] 5. User response processing and sentiment analysis:

[0308] The user listens to what the character says and responds, for example, by asking, "What did you do today?"

[0309] The device records the user's response and converts it into text data using a voice recognition engine.

[0310] The device sends the converted text data to the server, and the emotion analysis engine extracts and analyzes the user's emotional information.

[0311] 6. Emotion-based conversation generation:

[0312] The server receives the text data and emotional information sent by the user, and based on that, sends the next conversation scenario data to the generation AI.

[0313] The AI ​​generates conversational text based on the user's emotions. For example, if the user speaks in a happy tone, the AI ​​generates a positive response.

[0314] 7. Sending and playing audio files:

[0315] The server converts the generated new conversation text into voice data again using the text-to-speech engine.

[0316] The server transmits the converted audio file to the terminal, which receives it and plays it back.

[0317] This process is repeated, allowing for a natural conversation between the user and the character to continue.

[0318] Specific examples

[0319] For example, consider the case where a user says to a device, "I was so happy when my teacher praised me at school today!" The device records this voice and converts it into text using a speech recognition engine. The resulting text is sent to a server, which then uses an emotion analysis engine to determine that the user's emotion is "happy." Based on this determination, the generation AI generates a response such as, "That's wonderful! What did the teacher praise you for?" This response is converted into audio using a text-to-speech engine and sent to and played back on the device.

[0320] Prompt Sentence Examples

[0321] Here are some example prompts you can enter into your generative AI model:

[0322] 1. "If the user says in a cheerful voice, 'I had a great time picnicking in the park today,' generate an appropriate response in the following conversation."

[0323] 2. "If the user sadly says, 'I had a fight with my friend,' generate a comforting response in the following conversation."

[0324] In this way, by combining emotion analysis and generative AI, this system is able to respond according to the user's emotions.

[0325] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0326] Step 1:

[0327] The user opens the app on their device and selects the animal character they want to talk to. When they press the "Start Call" button, a call request including the character selection information is sent from the device to the server. The input here is the user's selection and the call start action, and the output is the call request being sent to the server.

[0328] Step 2:

[0329] The server analyzes the call request received from the device. The request contains character selection information, and extracts this information. The input is the call request sent from the device, and the output is the extracted character selection information. The server generates the initial conversation scenario data based on this information.

[0330] Step 3:

[0331] Based on the extracted character information, the server sends the initial conversation scenario data to the generation AI. An instruction is sent to generate an initial greeting such as "Hello, this is □□. What would you like to talk about today?" The input here is character selection information, and the output is scenario data for the generation AI.

[0332] Step 4:

[0333] The generation AI generates dialogue text based on the initial scenario data. The generated text is adjusted to match the character's personality and speaking style. The input is the initial scenario data, and the output is the generated dialogue text.

[0334] Step 5:

[0335] The server receives the generated text and converts it into audio data using a text-to-speech engine. The converted audio data is generated in the character's voice. Here, the input is the spoken text and the output is audio data.

[0336] Step 6:

[0337] The server sends the converted audio file to the device. The input is audio data, and the output is an audio file sent to the device. The device receives this audio file and plays it for the user. The user can enjoy the conversation while listening to the voices of the animal characters.

[0338] Step 7:

[0339] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input here is the user's response, which the device records. The output is the recorded voice data.

[0340] Step 8:

[0341] The device records the user's response and converts it into text using a speech recognition engine. The input is the voice data and the output is the converted text data, which is then sent to the server.

[0342] Step 9:

[0343] The server receives the converted text data and uses a sentiment analysis engine to extract and analyze the user's sentiment information, where the input is the converted text data and the output is the extracted sentiment information.

[0344] Step 10:

[0345] The server receives the text data and emotional information sent by the user and sends the next conversation scenario data to the generation AI based on that. The input here is the text data and emotional information, and the output is the next scenario data for the generation AI.

[0346] Step 11:

[0347] The generative AI generates conversational text according to the user's emotions. For example, if the user speaks in a cheerful manner, the generative AI generates a positive response. The input here is the following scenario data, and the output is the generated conversational text.

[0348] Step 12:

[0349] The server converts the newly generated conversation text into audio data again using a text-to-speech engine, where the input is the new conversation text and the output is audio data.

[0350] Step 13:

[0351] The server sends the converted audio file to the device, which then receives and plays it. The input here is audio data, and the output is the audio file sent to the device. This allows for a natural conversation between the user and the character to continue.

[0352] (Application example 2)

[0353] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0354] Conventional character dialogue systems have a formulaic user experience and have difficulty providing appropriate responses based on the user's emotions. As a result, particularly in interactive storytelling, users are less likely to feel a sense of realism or familiarity, and the dialogue tends to be uniform. The present invention aims to solve this problem by providing a more natural, interactive conversation experience that responds to the user's emotions.

[0355] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for the user to select a character and make a call request using a terminal, means for the server to receive a call request including character selection information from the terminal and send initial scenario data to the generation AI, and means for the generation AI to generate conversation text based on the scenario data. This makes it possible to analyze the user's emotional information in real time and generate and provide dynamic conversation content accordingly.

[0356] "Terminal" refers to an electronic device through which a user selects a character and makes a call request.

[0357] "Server" refers to the central processing unit that receives information from the terminal and works with the generative AI and other engines to generate and manage conversation scenarios.

[0358] "Character selection information" refers to information about a specific character selected by a user through a terminal.

[0359] A "call request" refers to request information sent from a terminal to a server in order for a user to start a conversation with a character.

[0360] "Initial scenario data" refers to the scenario information at the initial stage of the conversation that the server sends to the generation AI.

[0361] "Generative AI" refers to an artificial intelligence model that generates conversational text based on user input and scenario data.

[0362] "Conversational text" refers to the text data of the dialogue between the character generated by the generation AI and the user.

[0363] "Means for converting to speech" refers to a text-to-speech (TTS) engine for converting the generated conversational text into audio data.

[0364] "Audio file" refers to audio data sent from the server to the terminal and played as the character's voice.

[0365] The "means for converting a response into text" refers to a voice recognition engine that converts a voice response given by a user to a terminal into text data.

[0366] "Emotion analysis engine" refers to a program and device for analyzing emotional information from the content and voice of a user's responses.

[0367] "Emotion information" refers to data that indicates the type and intensity of emotions extracted from the user's responses.

[0368] This invention is a system that allows users to have real-time conversations with famous characters using a terminal, and adjusts the content of the conversation based on the user's emotions. This system is implemented by combining the following steps and hardware / software:

[0369] System Overview

[0370] The user selects a character and makes a call request through the device. The server receives this request information and generates initial scenario data based on the generation AI. The generation AI creates conversation text based on the generated scenario data, and the server converts this text into audio data using a text-to-speech (TTS) engine. This audio data is sent to the device and played back to the user.

[0371] Main processing

[0372] 1. Select a character and send a call request

[0373] The user opens the dedicated app on their device and selects the character they want to talk to. After selecting, the user presses the "Start Call" button, and a call request including the character selection information is sent to the server.

[0374] 2. Receiving a call request and generating a first conversation scenario

[0375] The server receives the call request and sends the initial scenario data based on the character information to the generation AI. The generation AI then generates an initial greeting such as, "Hello, this is □□. What would you like to talk about today?"

[0376] 3. Conversational text generation and speech conversion

[0377] The AI ​​generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. This dialogue text is then converted into voice data using a TTS engine.

[0378] 4. Sending and playing audio files

[0379] The server then sends the converted audio file to the device, which then plays it back, allowing the user to enjoy the conversation while listening to the character's voice.

[0380] Processing user responses and sentiment analysis

[0381] The user listens to what the character says and responds. The device records the user's response and converts it into text data using a voice recognition engine. The converted text data is then retransmitted to the server, which then uses an emotion analysis engine to extract and analyze the user's emotional information.

[0382] The process of emotion-based conversation generation

[0383] The server receives the user's text data and analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI, which then generates conversation text according to the user's emotions.

[0384] Hardware and software used

[0385] Hardware: Smartphones, servers

[0386] Software: Dedicated apps, generative AI models (e.g., GPT-3), text-to-speech (TTS) engines, speech recognition engines, sentiment analysis engines

[0387] Specific examples

[0388] For example, if a user says to a device, "I was so happy my teacher praised me at school today!", the device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion analysis engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a TTS engine and sent to the device for playback.

[0389] Prompt Sentence Examples

[0390] User: I was so happy when my teacher praised me at school today!

[0391] Prompt for the generated AI:

[0392] Generate a follow-up response in a friendly and cheerful tone for a user who just said, "I was so happy that my teacher praised me at school today!" The response should include a question to keep the conversation going.

[0393] This allows for natural, emotionally-responsive interactive conversation between the user and the character.

[0394] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0395] Step 1:

[0396] The user opens the dedicated app on their device. Within the app, the user selects the character they want to talk to and presses the "Start Call" button. This generates a call request including character selection information and sends it to the server. In this process, the input is the character information selected by the user, and the output is the call request.

[0397] Step 2:

[0398] The server analyzes the call request received from the device and extracts character selection information. Based on this information, the server sends the initial scenario data to the generation AI. The input is the call request, and the output is the initial scenario data to be passed to the generation AI.

[0399] Step 3:

[0400] The generation AI generates conversational text based on the initial scenario data. For example, the generation AI generates the initial greeting message, "Hello, this is □□. What would you like to talk about today?" In this case, the input is the initial scenario data, and the output is the conversational text.

[0401] Step 4:

[0402] The server receives the generated conversation text and converts it into audio data using a text-to-speech (TTS) engine. The input is the conversation text and the output is an audio file.

[0403] Step 5:

[0404] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user. The input is the audio file and the output is the played audio.

[0405] Step 6:

[0406] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input is the user's response to the character's utterance, and the output is the response as voice data.

[0407] Step 7:

[0408] The device records the user's response and converts it into text data using a speech recognition engine. The input is the user's voice response, and the output is the text-converted response data.

[0409] Step 8:

[0410] The recorded text data is retransmitted to the server, where the server uses an emotion analysis engine to extract and analyze emotional information from the user's speech and text. The input is the text-converted response data, and the output is the extracted emotional information.

[0411] Step 9:

[0412] The server receives the text data and analyzed emotional information sent by the user, and then sends the next conversation scenario data to the generation AI based on that. The input is the text data and emotional information, and the output is the next scenario data.

[0413] Step 10:

[0414] The generation AI generates conversational text according to the user's emotions. For example, if the user replies, "I had a great time at the picnic in the park today," the generation AI generates the following conversational content: "That's great! Who were your friends playing with?" The input is the next scenario data, and the output is new conversational text.

[0415] Step 11:

[0416] The server converts the generated new conversation text into audio data using the text-to-speech engine again and sends it to the terminal. The input is the new conversation text, and the output is an audio file as audio data.

[0417] Step 12:

[0418] The device then plays the received audio file back to the user, allowing for a natural conversation between the user and the character, and providing responses according to the user's emotions. The input is the audio file, and the output is the played audio.

[0419] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the 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.

[0420] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0421] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0422] [Second embodiment]

[0423] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0424] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0425] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0426] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0427] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0428] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0429] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0430] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0431] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0432] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0433] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0434] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0435] This invention is a system that allows users to use a terminal to have real-time telephone conversations with famous animal characters. Below, we will explain the program processing of this system and specific examples.

[0436] Character selection and call request

[0437] First, the user opens the dedicated app on their device. Within the app, they select the animal character they want to talk to. Next, when the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. This request also includes the identification information of the selected character.

[0438] Receiving a call request and generating a first conversation scenario

[0439] The server receives a call request from the device and analyzes the request. Based on the character information obtained as a result of the analysis, the server passes the initial conversation scenario data to the generation AI. This scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[0440] Conversational text generation

[0441] The AI ​​generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing the user with natural and familiar conversations.

[0442] Conversational text-to-speech

[0443] The server receives the generated conversational text from the AI ​​generator and converts it into audio using a text-to-speech (TTS) engine. The converted audio data is then tuned to play in the character's unique voice.

[0444] Sending and playing audio files

[0445] The server sends the converted audio file to the device, which then receives the file and plays it for the user to listen to. The user can enjoy the conversation while listening to the character's voice.

[0446] Handling the user's response

[0447] The user listens to the conversation and responds. The device records the user's response and converts it into text data using a speech recognition engine. This text data is then sent back to the server.

[0448] Next conversation generation process

[0449] The server receives the text data sent by the user and passes it to the generation AI as the next conversation scenario data. The generation AI generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[0450] Specific examples

[0451] For example, suppose a user asks, "What did you do today?" The server passes this question to the generation AI, which then generates a response such as, "I took a walk in the park today. I could hear lots of birds and it was very pleasant." This response is then converted into audio by the text-to-speech engine and sent to and played back on the device. The user can then listen to this and continue the conversation.

[0452] In this way, the system allows users to have natural conversations in real time with their chosen animal characters, providing an engaging and relatable experience especially for children.

[0453] The processing flow will be explained below.

[0454] Step 1:

[0455] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[0456] Step 2:

[0457] The server analyzes the call request received from the device and extracts the selected character information.The server then sends the initial conversation scenario data to the generation AI, instructing it to generate an opening greeting and introductory sentence.

[0458] Step 3:

[0459] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[0460] Step 4:

[0461] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[0462] Step 5:

[0463] The server sends the converted audio file to the terminal, which then plays the received audio file for the user to listen to.

[0464] Step 6:

[0465] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[0466] Step 7:

[0467] The device resends the converted text data to the server, which analyzes the received user response and passes the next conversation scenario data based on that to the generation AI.

[0468] Step 8:

[0469] The generative AI generates new conversational text based on the user's response. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[0470] Step 9:

[0471] The server converts the generated new conversation text into voice data using the text-to-speech engine again, and transmits the voice data to the terminal.

[0472] Step 10:

[0473] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character.

[0474] Example 1

[0475] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0476] Conventional systems have struggled to enable users to have natural, real-time conversations with their chosen characters, resulting in limited user experience. Furthermore, character voice generation and user response processing often lacked smoothness, resulting in delayed conversations and unnatural responses. These issues pose a significant obstacle, particularly for children, where familiarity and naturalness are crucial.

[0477] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0478] In this invention, the server includes: means for a user to use a terminal to select a character and make a call request; means for the terminal to send a call request including character selection information to the server; means for the server to receive and analyze the call request; means for the server to send initial scenario data to the generation AI; means for the generation AI to generate conversation text based on the initial scenario data; means for the server to convert the generated conversation text into voice; means for the server to send the converted voice file to the terminal and the terminal to play the voice; means for the user to respond to the terminal, convert the response into text using a voice recognition engine, and resend it to the server; and means for the server to send next scenario data to the generation AI based on the resent text. This enables users to have natural conversations in real time with their selected characters, providing an experience that is particularly familiar to children.

[0479] "User" refers to a person who uses a terminal to select a character and make a call request.

[0480] A "terminal" is a device operated by a user, which transmits a call request including character selection information to a server and plays back audio received from the server.

[0481] "Character selection information" refers to information for identifying a specific character selected by a user.

[0482] "Server" refers to a computer system on a network that receives and analyzes call requests sent from a device and works in conjunction with the generation AI and voice conversion function.

[0483] A "call request" refers to a request to the server to initiate a real-time conversation, including identification information for a user-selected character.

[0484] "Initial scenario data" refers to the data that defines the initial conversation scenario that the server sends to the generation AI.

[0485] "Generative AI" refers to an artificial intelligence model that generates conversational text based on received scenario data.

[0486] "Conversational text" refers to text that describes what a character will say to the user, generated by the generation AI based on scenario data.

[0487] "Convert to speech" refers to converting the conversational text generated by the generative AI into audio data using a text-to-speech engine.

[0488] "Speech recognition engine" refers to software or algorithms for converting a user's spoken responses into text data.

[0489] "Scenario data" refers to data that is passed to the generation AI to generate the content of the conversation.

[0490] This invention is a system that allows users to have natural conversations in real time with selected animal characters using a terminal. The core of this system is composed of a server, a terminal, and a generative AI model working together. Specific embodiments of the system are described below.

[0491] First, the user uses a device such as a smartphone or tablet. The user launches a dedicated app and selects the animal character they want to talk to from the displayed characters. For example, characters such as rabbits and bears can be selected. After selecting the character, the user presses the "Start Call" button. This action causes the device to package a call request in JSON format, including the identification information of the selected character, and send it to the server as an HTTP POST request.

[0492] When the server receives a call request, it analyzes the request and extracts the identification information of the selected character. The server then generates initial conversation scenario data and passes it to a generative AI model (e.g., OpenAI's GPT-4). The initial scenario data includes a prompt such as, "Hello, this is AA. What would you like to talk about today?"

[0493] The generative AI model generates dialogue text based on scenario data, imitating the character's personality and speaking style. The dialogue text is then sent to a server, which converts the received text into audio data using a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech). This audio data is then adjusted to be played in the character's unique voice.

[0494] The server sends the converted audio file to the device, which receives and plays it. The user can then hear the character's voice through the device and begin the conversation.

[0495] When the user responds to the device, the device records the voice and converts it into text data using an internal speech recognition engine (e.g., Google Cloud Speech-to-Text). The converted text data is then sent back to the server.

[0496] The server receives the text data sent by the user and passes it to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[0497] As a concrete example, consider the case where a user asks, "What did you do today?" Below is an example of a prompt sentence to input to the generation AI.

[0498] "Generate a conversation scenario in which the user asks, 'What did you do today?' and the character replies that he was walking in the park."

[0499] Based on this prompt, the generative AI model generates a response text such as, "I was walking in the park today. I could hear lots of birds singing and it felt great." This response text is converted into speech via the server and sent and played back on the device, allowing the user to enjoy a conversation with the character.

[0500] As a result, the system allows users to have natural conversations with selected animal characters in real time, providing an experience that is particularly friendly to children.

[0501] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0502] Step 1:

[0503] The user launches the dedicated app and selects an animal character.

[0504] Input: A device such as a smartphone or tablet operated by the user.

[0505] Specific operation: The user checks the list of available animal characters on the device's app screen and taps the desired character to select it.

[0506] Output: The identity of the selected character.

[0507] Step 2:

[0508] The user presses the "start call" button.

[0509] Input: The user-selected character's identity.

[0510] Specific operation: The user taps the "Start Call" button on the app screen. This action prepares the device to perform the following process.

[0511] Output: Call request trigger.

[0512] Step 3:

[0513] The terminal transmits a call request including character selection information to the server.

[0514] Input: The identity of the selected character, a trigger indicating that the "Start Call" button has been pressed.

[0515] Specific operation: The device creates a call request in JSON format including the identification information of the selected character and sends it to the server as an HTTP POST request.

[0516] Output: The call request data sent to the server.

[0517] Step 4:

[0518] The server receives and analyzes the call request.

[0519] Input: Call request data sent from the device.

[0520] Specific operation: The server receives the call request, parses the JSON data to extract the character's identification information, confirms the selection information, and prepares for the next process.

[0521] Output: Parsed character identities.

[0522] Step 5:

[0523] The server sends the initial scenario data to the generation AI.

[0524] Input: Parsed character identity.

[0525] Specific operation: The server generates the initial conversation scenario data (e.g., "Hello, this is AA. What would you like to talk about today?") and sends it to the generation AI as a prompt.

[0526] Output: Initial scenario data sent to the generation AI.

[0527] Step 6:

[0528] The generation AI generates conversational text based on the initial scenario data.

[0529] Input: Initial scenario data sent to the generating AI.

[0530] How it works: A generative AI model (e.g., OpenAI GPT-4) generates an initial conversation text based on the received prompt, imitating the character's personality and speaking style.

[0531] Output: The generated conversation text.

[0532] Step 7:

[0533] The server converts the generated conversation text into speech.

[0534] Input: The generated conversation text.

[0535] How it works: The server passes the generated conversation text to a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech) and converts it into audio data. The settings are adjusted to create the character's unique voice tone.

[0536] Output: The converted audio data.

[0537] Step 8:

[0538] The server sends the converted audio file to the terminal, which then plays the audio.

[0539] Input: The converted audio data.

[0540] Specific operation: The server transfers the generated audio file to the device, sending the audio data as an HTTP response. The device then plays the received audio file, allowing the user to hear the character's voice.

[0541] Output: The audio the user hears.

[0542] Step 9:

[0543] The user responds.

[0544] Input: The character's voice as heard by the user on their device.

[0545] Specific operation: After listening to the character's voice, the user responds.

[0546] Output: The user's spoken response.

[0547] Step 10:

[0548] The device records the user's response and converts it into text using a voice recognition engine.

[0549] Input: The user's voice reply.

[0550] Specific operation: The device records the user's voice and converts it into text data using an internal voice recognition engine (e.g., Google Cloud Speech-to-Text).

[0551] Output: User responses converted to text data.

[0552] Step 11:

[0553] The terminal transmits the converted text data to the server.

[0554] Input: User responses converted to text data.

[0555] Specific operation: The terminal packages the converted text data in JSON format and sends it to the server as an HTTP POST request.

[0556] Output: The text data of the user response sent to the server.

[0557] Step 12:

[0558] The server sends the next scenario data to the generation AI based on the resent text.

[0559] Input: Text data of the user's response sent from the terminal.

[0560] Specific operation: The server receives the user's response text and passes it to the generation AI as the next conversation scenario. It creates a prompt sentence to input to the generation AI.

[0561] Output: Send the following conversation scenario data to the generation AI.

[0562] Step 13:

[0563] The generation AI will generate the following conversation text again.

[0564] Input: The prompt sent to the generation AI as the following scenario data.

[0565] How it works: The AI ​​generates the next conversation text based on the user's response. This process is repeated until the user ends the conversation.

[0566] Output: The following conversation text.

[0567] Through this series of steps, users can enjoy natural conversations in real time with their chosen animal character.

[0568] (Application example 1)

[0569] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0570] In conventional character conversation systems, conversations between users and characters are often scripted, making it difficult to provide natural conversations in real time. Furthermore, there was a lack of technology to generate appropriate responses to user voice input and play them back in a natural voice, limiting the user experience.

[0571] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0572] In this invention, the server includes: means for a user to select a character and make a call request using an information processing device; means for the server to receive a call request including character selection information from the information processing device and transmit first conversation scenario data to the generation AI model; means for the generation AI model to generate conversation text based on the scenario data; means for the server to convert the generated conversation text into voice data; means for transmitting voice data from the server to the information processing device and for the information processing device to play the voice; means for the user to respond to the information processing device, converting the response into text data and retransmitting it to the server; means for the server to pass next conversation scenario data to the generation AI model based on the retransmitted text data; and means for converting the conversation text of the generation AI model into voice using a text-to-speech engine and providing the conversation data in real time, thereby enabling users to enjoy natural conversations in real time.

[0573] "Information processing device" is a general term for electronic devices used for processing and communicating data, and specifically includes smartphones, tablets, personal computers, etc.

[0574] A "server" is a central computer system that receives requests from multiple terminals on a network and provides data processing and communication services.

[0575] "Character selection information" is data for specifying the character with which the user wishes to converse, and this data is received by the server.

[0576] A "call request" is a request made by a user to start a conversation with a selected character using an information processing device.

[0577] "First conversation scenario data" is conversation script data generated to smoothly start the first conversation between a character and a user.

[0578] A "generative AI model" is an artificial intelligence model used to generate natural-sounding conversational text, using machine learning to generate responses in response to user input.

[0579] "Conversation text" is text data generated by a generative AI model that indicates the content of a conversation between a user and a character.

[0580] "Audio data" refers to an audio file generated from text data using speech synthesis technology, and is played back by an information processing device.

[0581] "Text data" refers to character information to be displayed or processed on an information processing device, and includes user responses and generated conversation content.

[0582] A "text-to-speech engine" is a software component that converts text data into audio data, enabling playback in a natural voice.

[0583] "Conversation scenario data" is data that indicates the next conversation content that the generative AI model will generate based on user input, enabling continuous conversation between the character and the user.

[0584] "Real-time" means that the conversation between the user and the character is immediate, with natural responses provided without delay.

[0585] A specific embodiment of the present invention will be described below: This system allows a user to use an information processing device to enjoy natural conversations with a specific character in real time.

[0586] First, the user launches a dedicated application using an information processing device (e.g., a smartphone, tablet, or head-mounted display (HMD)). Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button, which sends a call request to the server. The call request includes the identification information of the character selected by the user.

[0587] The server receives the call request and analyzes its content. Based on the analysis results, the server passes the initial conversation scenario data to the generative AI model. This initial conversation scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[0588] A generative AI model (e.g., GPT-3) generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing a natural and familiar conversation to the user. This conversational text is sent to a server and converted into audio data using a text-to-speech engine (TTS engine).

[0589] The voice data is sent from the server to the information processing device, which then plays back the voice data. This allows the user to hear the character's voice. The user listens to the conversation and responds. The information processing device records the user's response and converts the voice into text data using a voice recognition engine (e.g., Google Speech API). This text data is then sent back to the server.

[0590] The server receives the text data sent by the user and passes it back to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, which the server converts again into voice data using a TTS engine and sends and plays back to the information processing device. This process is repeated, allowing for a natural conversation between the user and the character to continue.

[0591] As a concrete example, suppose a user asks, "What did you do today?" The server passes this question to a generative AI model, which generates a response such as, "I took a walk in the park today. I was pleasantly surprised to hear the birds singing." This response is converted by a TTS engine, sent to an information processing device, and played back. The user can listen to this and continue the conversation.

[0592] Examples of prompts include:

[0593] User Input:

[0594] User: What did you do today?

[0595] Response from the generative AI model:

[0596] Character: I was walking in the park today. I could hear a lot of birds singing and it felt great.

[0597] The system allows users to engage in natural conversations with selected characters in real time, providing a familiar and engaging experience especially for children.

[0598] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0599] Step 1:

[0600] The user launches a dedicated application using an information processing device. Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button. Identification information for the character selected by the user is then generated and sent to the server as a call request. The input is the user's character selection, and the output is the call request data.

[0601] Step 2:

[0602] The server receives a call request sent from the user's information processing device. The server then analyzes the request content and extracts the character's identification information. The input is the call request data, and the output is the character's identification information.

[0603] Step 3:

[0604] The server passes the first conversation scenario data to the generation AI model based on the character's identification information. This scenario data includes the content of the first conversation based on the character's personality and speaking style. The input is the character's identification information, and the output is the first conversation scenario data.

[0605] Step 4:

[0606] The generative AI model generates conversational text based on the received initial conversation scenario data. For example, it generates the content "Hello, this is AA. What would you like to talk about today?" The input is the initial conversation scenario data, and the output is the conversational text.

[0607] Step 5:

[0608] The server receives the conversational text from the generative AI model and converts it into audio data using a text-to-speech engine (TTS engine), where the character's unique voice is reproduced. The input is the conversational text, and the output is audio data.

[0609] Step 6:

[0610] The server transmits the generated voice data to the information processing device, which receives the voice data and plays it back for the user to hear. The input is the voice data, and the output is the situation in which the user can hear the voice.

[0611] Step 7:

[0612] The user listens to the character's voice and responds. The information processing device records the user's response and converts it into text data using a voice recognition engine. The input is the user's voice, and the output is text data.

[0613] Step 8:

[0614] The information processing device sends the user's response text data to the server. The server receives this text data and passes it to the generative AI model as the next conversation scenario data. The input is the user's response text data, and the output is the next conversation scenario data.

[0615] Step 9:

[0616] The generative AI model generates new conversation text based on the next conversation scenario data. The generated conversation text is sent to the server again. The input is the next conversation scenario data, and the output is the new conversation text.

[0617] Step 10:

[0618] The processes from step 5 to step 9 are repeated to continue a natural conversation between the user and the character, thereby maintaining a natural conversation in real time.

[0619] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0620] This invention is a system that enables a user to use a terminal to have real-time telephone conversations with famous animal characters, and that recognizes the user's emotions and adjusts the content of the conversation.

[0621] Character selection and call request

[0622] First, the user opens the app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. The selection information includes the character specified by the user.

[0623] Receiving a call request and generating a first conversation scenario

[0624] The server analyzes the call request received from the device and extracts the selected character information. Based on this information, the server sends the initial conversation scenario data to the generation AI, instructing it to generate an initial greeting such as, "Hello, this is XX. What would you like to talk about today?"

[0625] Conversational text generation and speech conversion

[0626] The generation AI generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. The server receives this dialogue text and converts it into audio data using a text-to-speech (TTS) engine. The converted audio data is generated in the character's voice.

[0627] Sending and playing audio files

[0628] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user, who then enjoys the conversation while listening to the animal characters' voices.

[0629] Processing user responses and sentiment analysis

[0630] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine. The converted text data is then resent to the server, where the emotion engine extracts and analyzes emotional information from the user's speech and text.

[0631] The process of emotion-based conversation generation

[0632] The server receives the text data sent by the user and the analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI. The generation AI generates conversation text according to the user's emotions. For example, if the user replies in a happy voice, "I had a great time at the picnic in the park today," the generation AI will generate a conversation such as, "That's great! Who were your friends playing with?"

[0633] Sending and playing audio files

[0634] The server then converts the newly generated conversation text into audio data using a text-to-speech engine and sends it to the device. The device then plays the received audio file back to the user. This process is repeated, allowing for a natural conversation between the user and the character to continue, and providing responses that correspond to the user's emotions.

[0635] Specific examples

[0636] For example, suppose a user says aloud to a device, "I was so happy my teacher praised me at school today!" The device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a text-to-speech engine and sent to the device for playback.

[0637] In this way, the system, combined with the emotion engine, enables more natural and friendly conversations based on the user's emotions, making interactions with characters that understand and respond to emotions even more appealing, especially for children.

[0638] The processing flow will be explained below.

[0639] Step 1:

[0640] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[0641] Step 2:

[0642] The server analyzes the call request received from the device and extracts the selected character information. Based on the character information obtained from the analysis, the initial conversation scenario data is provided to the generation AI.

[0643] Step 3:

[0644] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[0645] Step 4:

[0646] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[0647] Step 5:

[0648] The server sends the converted audio file to the device, which then plays the received audio file, allowing the user to hear the character's voice.

[0649] Step 6:

[0650] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[0651] Step 7:

[0652] The device retransmits the text data to the server, and simultaneously transmits the recorded data to the server, where the emotion engine extracts and analyzes emotional information from the user's speech.

[0653] Step 8:

[0654] The server provides the next conversation scenario data to the generation AI based on the received text data and emotional information. The generation AI generates the optimal response text based on the user's emotions analyzed by the emotion engine.

[0655] Step 9:

[0656] The generative AI generates new conversational text based on the user's responses and emotional information. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[0657] Step 10:

[0658] The server converts the generated new conversation text into voice data using a text-to-speech (TTS) engine again and sends it to the terminal.

[0659] Step 11:

[0660] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character, while providing responses that correspond to the user's emotions.

[0661] Step 12:

[0662] As a concrete example, if a user says to a device, "I was so happy when my teacher praised me at school today!", the device records this voice and converts it into text using a speech recognition engine. The server then passes this text data and the emotional information of "happy" analyzed by the emotion engine to the generation AI, which then generates a response such as, "That's wonderful! What did your teacher praise you for?" This response is converted into audio and played on the device.

[0663] Example 2

[0664] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0665] Conventional conversational systems have the problem that the conversation between the user and the character is formulaic and lacks responses that reflect the user's emotions. In particular, there is a need for a system that can recognize the user's emotions and generate natural conversations accordingly.

[0666] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0667] In this invention, the server includes a means for the user to respond to the terminal, converting the response into text and retransmitting it to the server, a means for the server to pass the next scenario data to the generation AI based on the retransmitted text, a means for the terminal to record the user's response and convert it into text data using a voice recognition engine, and a means for the server to extract and analyze the user's emotional information using an emotion analysis engine, thereby enabling a response that corresponds to the user's emotions.

[0668] A "terminal" is an electronic device such as a computer or smartphone that is operated by a user.

[0669] A "server" is a computer system that receives requests and data from terminals, processes them, and gives instructions to the generating AI.

[0670] "Generative AI" is an artificial intelligence model that generates conversational text based on scenario data.

[0671] "Character selection information" is information about the animal character with which the user wants to converse.

[0672] A "call request" is request data that a user sends from a terminal to a server in order to start a call with a character.

[0673] "Scenario data" is the initial set of information that the generation AI uses to generate conversational text.

[0674] "Conversational text" is a sentence generated by the generation AI based on scenario data.

[0675] "Audio data" refers to an audio file converted from text by a text-to-speech engine.

[0676] An "audio file" is a digital file containing audio data that is sent from a server to a terminal.

[0677] A "voice recognition engine" is software that records a user's speech and converts it into text data.

[0678] An "emotion analysis engine" is software that extracts and analyzes emotional information from a user's speech or text.

[0679] A "text-to-speech engine" is software that converts text data into voice data.

[0680] This invention is a system that allows users to use their terminals to have real-time telephone conversations with famous animal characters. Furthermore, it can recognize the user's emotions and adjust the content of the conversation. The detailed implementation method of this system is described below.

[0681] Hardware and software used

[0682] Device: A device operated by a user, such as a smartphone or tablet.

[0683] Server: A computer system that processes user requests and interacts with generative AI models.

[0684] Generative AI model: An artificial intelligence model for generating conversational text.

[0685] Speech recognition engine: Software that records what a user says and converts it into text data (e.g., Google Speech-to-Text).

[0686] Sentiment analysis engine: Software that extracts and analyzes users' emotional information (e.g., IBM Watson Tone Analyzer).

[0687] Text-to-speech engine: Software that converts text data into speech data (e.g., Amazon Polly).

[0688] Specific operation of the system

[0689] 1. Select a character and send a call request:

[0690] Users use their device to open a dedicated app and select the animal character they want to talk to.

[0691] When the user presses the "start call" button, a call request including character selection information is sent from the terminal to the server.

[0692] 2. Generating the first conversation scenario:

[0693] The server analyzes the call request received from the terminal and extracts the selected character information.

[0694] The server sends the initial conversation scenario data to the generation AI and instructs it to generate the initial greeting.

[0695] 3. Conversational text generation and speech conversion:

[0696] The generation AI generates conversational text based on the initial scenario data.

[0697] The server receives the generated text and converts it into audio data using a text-to-speech engine.

[0698] 4. Send and play audio files:

[0699] The server sends the converted audio file to the terminal.

[0700] The terminal receives the audio file and plays it for the user.

[0701] 5. User response processing and sentiment analysis:

[0702] The user listens to what the character says and responds, for example, by asking, "What did you do today?"

[0703] The device records the user's response and converts it into text data using a voice recognition engine.

[0704] The device sends the converted text data to the server, and the emotion analysis engine extracts and analyzes the user's emotional information.

[0705] 6. Emotion-based conversation generation:

[0706] The server receives the text data and emotional information sent by the user, and based on that, sends the next conversation scenario data to the generation AI.

[0707] The AI ​​generates conversational text based on the user's emotions. For example, if the user speaks in a happy tone, the AI ​​generates a positive response.

[0708] 7. Sending and playing audio files:

[0709] The server converts the generated new conversation text into voice data again using the text-to-speech engine.

[0710] The server transmits the converted audio file to the terminal, which receives it and plays it back.

[0711] This process is repeated, allowing for a natural conversation between the user and the character to continue.

[0712] Specific examples

[0713] For example, consider the case where a user says to a device, "I was so happy when my teacher praised me at school today!" The device records this voice and converts it into text using a speech recognition engine. The resulting text is sent to a server, which then uses an emotion analysis engine to determine that the user's emotion is "happy." Based on this determination, the generation AI generates a response such as, "That's wonderful! What did the teacher praise you for?" This response is converted into audio using a text-to-speech engine and sent to and played back on the device.

[0714] Prompt Sentence Examples

[0715] Here are some example prompts you can enter into your generative AI model:

[0716] 1. "If the user says in a cheerful voice, 'I had a great time picnicking in the park today,' generate an appropriate response in the following conversation."

[0717] 2. "If the user sadly says, 'I had a fight with my friend,' generate a comforting response in the following conversation."

[0718] In this way, by combining emotion analysis and generative AI, this system is able to respond according to the user's emotions.

[0719] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0720] Step 1:

[0721] The user opens the app on their device and selects the animal character they want to talk to. When they press the "Start Call" button, a call request including the character selection information is sent from the device to the server. The input here is the user's selection and the call start action, and the output is the call request being sent to the server.

[0722] Step 2:

[0723] The server analyzes the call request received from the device. The request contains character selection information, and extracts this information. The input is the call request sent from the device, and the output is the extracted character selection information. The server generates the initial conversation scenario data based on this information.

[0724] Step 3:

[0725] Based on the extracted character information, the server sends the initial conversation scenario data to the generation AI. An instruction is sent to generate an initial greeting such as "Hello, this is □□. What would you like to talk about today?" The input here is character selection information, and the output is scenario data for the generation AI.

[0726] Step 4:

[0727] The generation AI generates dialogue text based on the initial scenario data. The generated text is adjusted to match the character's personality and speaking style. The input is the initial scenario data, and the output is the generated dialogue text.

[0728] Step 5:

[0729] The server receives the generated text and converts it into audio data using a text-to-speech engine. The converted audio data is generated in the character's voice. Here, the input is the spoken text and the output is audio data.

[0730] Step 6:

[0731] The server sends the converted audio file to the device. The input is audio data, and the output is an audio file sent to the device. The device receives this audio file and plays it for the user. The user can enjoy the conversation while listening to the voices of the animal characters.

[0732] Step 7:

[0733] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input here is the user's response, which the device records. The output is the recorded voice data.

[0734] Step 8:

[0735] The device records the user's response and converts it into text using a speech recognition engine. The input is the voice data and the output is the converted text data, which is then sent to the server.

[0736] Step 9:

[0737] The server receives the converted text data and uses a sentiment analysis engine to extract and analyze the user's sentiment information, where the input is the converted text data and the output is the extracted sentiment information.

[0738] Step 10:

[0739] The server receives the text data and emotional information sent by the user and sends the next conversation scenario data to the generation AI based on that. The input here is the text data and emotional information, and the output is the next scenario data for the generation AI.

[0740] Step 11:

[0741] The generative AI generates conversational text according to the user's emotions. For example, if the user speaks in a cheerful manner, the generative AI generates a positive response. The input here is the following scenario data, and the output is the generated conversational text.

[0742] Step 12:

[0743] The server converts the newly generated conversation text into audio data again using a text-to-speech engine, where the input is the new conversation text and the output is audio data.

[0744] Step 13:

[0745] The server sends the converted audio file to the device, which then receives and plays it. The input here is audio data, and the output is the audio file sent to the device. This allows for a natural conversation between the user and the character to continue.

[0746] (Application example 2)

[0747] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0748] Conventional character dialogue systems have a formulaic user experience and have difficulty providing appropriate responses based on the user's emotions. As a result, particularly in interactive storytelling, users are less likely to feel a sense of realism or familiarity, and the dialogue tends to be uniform. The present invention aims to solve this problem by providing a more natural, interactive conversation experience that responds to the user's emotions.

[0749] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for the user to select a character and make a call request using a terminal, means for the server to receive a call request including character selection information from the terminal and send initial scenario data to the generation AI, and means for the generation AI to generate conversation text based on the scenario data. This makes it possible to analyze the user's emotional information in real time and generate and provide dynamic conversation content accordingly.

[0750] "Terminal" refers to an electronic device through which a user selects a character and makes a call request.

[0751] "Server" refers to the central processing unit that receives information from the terminal and works with the generative AI and other engines to generate and manage conversation scenarios.

[0752] "Character selection information" refers to information about a specific character selected by a user through a terminal.

[0753] A "call request" refers to request information sent from a terminal to a server in order for a user to start a conversation with a character.

[0754] "Initial scenario data" refers to the scenario information at the initial stage of the conversation that the server sends to the generation AI.

[0755] "Generative AI" refers to an artificial intelligence model that generates conversational text based on user input and scenario data.

[0756] "Conversational text" refers to the text data of the dialogue between the character generated by the generation AI and the user.

[0757] "Means for converting to speech" refers to a text-to-speech (TTS) engine for converting the generated conversational text into audio data.

[0758] "Audio file" refers to audio data sent from the server to the terminal and played as the character's voice.

[0759] The "means for converting a response into text" refers to a voice recognition engine that converts a voice response given by a user to a terminal into text data.

[0760] "Emotion analysis engine" refers to a program and device for analyzing emotional information from the content and voice of a user's responses.

[0761] "Emotion information" refers to data that indicates the type and intensity of emotions extracted from the user's responses.

[0762] This invention is a system that allows users to have real-time conversations with famous characters using a terminal, and adjusts the content of the conversation based on the user's emotions. This system is implemented by combining the following steps and hardware / software:

[0763] System Overview

[0764] The user selects a character and makes a call request through the device. The server receives this request information and generates initial scenario data based on the generation AI. The generation AI creates conversation text based on the generated scenario data, and the server converts this text into audio data using a text-to-speech (TTS) engine. This audio data is sent to the device and played back to the user.

[0765] Main processing

[0766] 1. Select a character and send a call request

[0767] The user opens the dedicated app on their device and selects the character they want to talk to. After selecting, the user presses the "Start Call" button, and a call request including the character selection information is sent to the server.

[0768] 2. Receiving a call request and generating a first conversation scenario

[0769] The server receives the call request and sends the initial scenario data based on the character information to the generation AI. The generation AI then generates an initial greeting such as, "Hello, this is □□. What would you like to talk about today?"

[0770] 3. Conversational text generation and speech conversion

[0771] The AI ​​generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. This dialogue text is then converted into voice data using a TTS engine.

[0772] 4. Sending and playing audio files

[0773] The server then sends the converted audio file to the device, which then plays it back, allowing the user to enjoy the conversation while listening to the character's voice.

[0774] Processing user responses and sentiment analysis

[0775] The user listens to what the character says and responds. The device records the user's response and converts it into text data using a voice recognition engine. The converted text data is then retransmitted to the server, which then uses an emotion analysis engine to extract and analyze the user's emotional information.

[0776] The process of emotion-based conversation generation

[0777] The server receives the user's text data and analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI, which then generates conversation text according to the user's emotions.

[0778] Hardware and software used

[0779] Hardware: Smartphones, servers

[0780] Software: Dedicated apps, generative AI models (e.g., GPT-3), text-to-speech (TTS) engines, speech recognition engines, sentiment analysis engines

[0781] Specific examples

[0782] For example, if a user says to a device, "I was so happy my teacher praised me at school today!", the device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion analysis engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a TTS engine and sent to the device for playback.

[0783] Prompt Sentence Examples

[0784] User: I was so happy when my teacher praised me at school today!

[0785] Prompt for the generated AI:

[0786] Generate a follow-up response in a friendly and cheerful tone for a user who just said, "I was so happy that my teacher praised me at school today!" The response should include a question to keep the conversation going.

[0787] This allows for natural, emotionally-responsive interactive conversation between the user and the character.

[0788] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0789] Step 1:

[0790] The user opens the dedicated app on their device. Within the app, the user selects the character they want to talk to and presses the "Start Call" button. This generates a call request including character selection information and sends it to the server. In this process, the input is the character information selected by the user, and the output is the call request.

[0791] Step 2:

[0792] The server analyzes the call request received from the device and extracts character selection information. Based on this information, the server sends the initial scenario data to the generation AI. The input is the call request, and the output is the initial scenario data to be passed to the generation AI.

[0793] Step 3:

[0794] The generation AI generates conversational text based on the initial scenario data. For example, the generation AI generates the initial greeting message, "Hello, this is □□. What would you like to talk about today?" In this case, the input is the initial scenario data, and the output is the conversational text.

[0795] Step 4:

[0796] The server receives the generated conversation text and converts it into audio data using a text-to-speech (TTS) engine. The input is the conversation text and the output is an audio file.

[0797] Step 5:

[0798] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user. The input is the audio file and the output is the played audio.

[0799] Step 6:

[0800] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input is the user's response to the character's utterance, and the output is the response as voice data.

[0801] Step 7:

[0802] The device records the user's response and converts it into text data using a speech recognition engine. The input is the user's voice response, and the output is the text-converted response data.

[0803] Step 8:

[0804] The recorded text data is retransmitted to the server, where the server uses an emotion analysis engine to extract and analyze emotional information from the user's speech and text. The input is the text-converted response data, and the output is the extracted emotional information.

[0805] Step 9:

[0806] The server receives the text data and analyzed emotional information sent by the user, and then sends the next conversation scenario data to the generation AI based on that. The input is the text data and emotional information, and the output is the next scenario data.

[0807] Step 10:

[0808] The generation AI generates conversational text according to the user's emotions. For example, if the user replies, "I had a great time at the picnic in the park today," the generation AI generates the following conversational content: "That's great! Who were your friends playing with?" The input is the next scenario data, and the output is new conversational text.

[0809] Step 11:

[0810] The server converts the generated new conversation text into audio data using the text-to-speech engine again and sends it to the terminal. The input is the new conversation text, and the output is an audio file as audio data.

[0811] Step 12:

[0812] The device then plays the received audio file back to the user, allowing for a natural conversation between the user and the character, and providing responses according to the user's emotions. The input is the audio file, and the output is the played audio.

[0813] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0814] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0815] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0816] [Third embodiment]

[0817] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0818] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0819] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0820] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0821] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0822] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0823] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0824] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0825] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0826] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0827] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0828] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0829] This invention is a system that allows users to use a terminal to have real-time telephone conversations with famous animal characters. Below, we will explain the program processing of this system and specific examples.

[0830] Character selection and call request

[0831] First, the user opens the dedicated app on their device. Within the app, they select the animal character they want to talk to. Next, when the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. This request also includes the identification information of the selected character.

[0832] Receiving a call request and generating a first conversation scenario

[0833] The server receives a call request from the device and analyzes the request. Based on the character information obtained as a result of the analysis, the server passes the initial conversation scenario data to the generation AI. This scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[0834] Conversational text generation

[0835] The AI ​​generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing the user with natural and familiar conversations.

[0836] Conversational text-to-speech

[0837] The server receives the generated conversational text from the AI ​​generator and converts it into audio using a text-to-speech (TTS) engine. The converted audio data is then tuned to play in the character's unique voice.

[0838] Sending and playing audio files

[0839] The server sends the converted audio file to the device, which then receives the file and plays it for the user to listen to. The user can enjoy the conversation while listening to the character's voice.

[0840] Handling the user's response

[0841] The user listens to the conversation and responds. The device records the user's response and converts it into text data using a speech recognition engine. This text data is then sent back to the server.

[0842] Next conversation generation process

[0843] The server receives the text data sent by the user and passes it to the generation AI as the next conversation scenario data. The generation AI generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[0844] Specific examples

[0845] For example, suppose a user asks, "What did you do today?" The server passes this question to the generation AI, which then generates a response such as, "I took a walk in the park today. I could hear lots of birds and it was very pleasant." This response is then converted into audio by the text-to-speech engine and sent to and played back on the device. The user can then listen to this and continue the conversation.

[0846] In this way, the system allows users to have natural conversations in real time with their chosen animal characters, providing an engaging and relatable experience especially for children.

[0847] The processing flow will be explained below.

[0848] Step 1:

[0849] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[0850] Step 2:

[0851] The server analyzes the call request received from the device and extracts the selected character information.The server then sends the initial conversation scenario data to the generation AI, instructing it to generate an opening greeting and introductory sentence.

[0852] Step 3:

[0853] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[0854] Step 4:

[0855] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[0856] Step 5:

[0857] The server sends the converted audio file to the terminal, which then plays the received audio file for the user to listen to.

[0858] Step 6:

[0859] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[0860] Step 7:

[0861] The device resends the converted text data to the server, which analyzes the received user response and passes the next conversation scenario data based on that to the generation AI.

[0862] Step 8:

[0863] The generative AI generates new conversational text based on the user's response. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[0864] Step 9:

[0865] The server converts the generated new conversation text into voice data using the text-to-speech engine again, and transmits the voice data to the terminal.

[0866] Step 10:

[0867] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character.

[0868] Example 1

[0869] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0870] Conventional systems have struggled to enable users to have natural, real-time conversations with their chosen characters, resulting in limited user experience. Furthermore, character voice generation and user response processing often lacked smoothness, resulting in delayed conversations and unnatural responses. These issues pose a significant obstacle, particularly for children, where familiarity and naturalness are crucial.

[0871] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0872] In this invention, the server includes: means for a user to use a terminal to select a character and make a call request; means for the terminal to send a call request including character selection information to the server; means for the server to receive and analyze the call request; means for the server to send initial scenario data to the generation AI; means for the generation AI to generate conversation text based on the initial scenario data; means for the server to convert the generated conversation text into voice; means for the server to send the converted voice file to the terminal and the terminal to play the voice; means for the user to respond to the terminal, convert the response into text using a voice recognition engine, and resend it to the server; and means for the server to send next scenario data to the generation AI based on the resent text. This enables users to have natural conversations in real time with their selected characters, providing an experience that is particularly familiar to children.

[0873] "User" refers to a person who uses a terminal to select a character and make a call request.

[0874] A "terminal" is a device operated by a user, which transmits a call request including character selection information to a server and plays back audio received from the server.

[0875] "Character selection information" refers to information for identifying a specific character selected by a user.

[0876] "Server" refers to a computer system on a network that receives and analyzes call requests sent from a device and works in conjunction with the generation AI and voice conversion function.

[0877] A "call request" refers to a request to the server to initiate a real-time conversation, including identification information for a user-selected character.

[0878] "Initial scenario data" refers to the data that defines the initial conversation scenario that the server sends to the generation AI.

[0879] "Generative AI" refers to an artificial intelligence model that generates conversational text based on received scenario data.

[0880] "Conversational text" refers to text that describes what a character will say to the user, generated by the generation AI based on scenario data.

[0881] "Convert to speech" refers to converting the conversational text generated by the generative AI into audio data using a text-to-speech engine.

[0882] "Speech recognition engine" refers to software or algorithms for converting a user's spoken responses into text data.

[0883] "Scenario data" refers to data that is passed to the generation AI to generate the content of the conversation.

[0884] This invention is a system that allows users to have natural conversations in real time with selected animal characters using a terminal. The core of this system is composed of a server, a terminal, and a generative AI model working together. Specific embodiments of the system are described below.

[0885] First, the user uses a device such as a smartphone or tablet. The user launches a dedicated app and selects the animal character they want to talk to from the displayed characters. For example, characters such as rabbits and bears can be selected. After selecting the character, the user presses the "Start Call" button. This action causes the device to package a call request in JSON format, including the identification information of the selected character, and send it to the server as an HTTP POST request.

[0886] When the server receives a call request, it analyzes the request and extracts the identification information of the selected character. The server then generates initial conversation scenario data and passes it to a generative AI model (e.g., OpenAI's GPT-4). The initial scenario data includes a prompt such as, "Hello, this is AA. What would you like to talk about today?"

[0887] The generative AI model generates dialogue text based on scenario data, imitating the character's personality and speaking style. The dialogue text is then sent to a server, which converts the received text into audio data using a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech). This audio data is then adjusted to be played in the character's unique voice.

[0888] The server sends the converted audio file to the device, which receives and plays it. The user can then hear the character's voice through the device and begin the conversation.

[0889] When the user responds to the device, the device records the voice and converts it into text data using an internal speech recognition engine (e.g., Google Cloud Speech-to-Text). The converted text data is then sent back to the server.

[0890] The server receives the text data sent by the user and passes it to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[0891] As a concrete example, consider the case where a user asks, "What did you do today?" Below is an example of a prompt sentence to input to the generation AI.

[0892] "Generate a conversation scenario in which the user asks, 'What did you do today?' and the character replies that he was walking in the park."

[0893] Based on this prompt, the generative AI model generates a response text such as, "I was walking in the park today. I could hear lots of birds singing and it felt great." This response text is converted into speech via the server and sent and played back on the device, allowing the user to enjoy a conversation with the character.

[0894] As a result, the system allows users to have natural conversations with selected animal characters in real time, providing an experience that is particularly friendly to children.

[0895] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0896] Step 1:

[0897] The user launches the dedicated app and selects an animal character.

[0898] Input: A device such as a smartphone or tablet operated by the user.

[0899] Specific operation: The user checks the list of available animal characters on the device's app screen and taps the desired character to select it.

[0900] Output: The identity of the selected character.

[0901] Step 2:

[0902] The user presses the "start call" button.

[0903] Input: The user-selected character's identity.

[0904] Specific operation: The user taps the "Start Call" button on the app screen. This action prepares the device to perform the following process.

[0905] Output: Call request trigger.

[0906] Step 3:

[0907] The terminal transmits a call request including character selection information to the server.

[0908] Input: The identity of the selected character, a trigger indicating that the "Start Call" button has been pressed.

[0909] Specific operation: The device creates a call request in JSON format including the identification information of the selected character and sends it to the server as an HTTP POST request.

[0910] Output: The call request data sent to the server.

[0911] Step 4:

[0912] The server receives and analyzes the call request.

[0913] Input: Call request data sent from the device.

[0914] Specific operation: The server receives the call request, parses the JSON data to extract the character's identification information, confirms the selection information, and prepares for the next process.

[0915] Output: Parsed character identities.

[0916] Step 5:

[0917] The server sends the initial scenario data to the generation AI.

[0918] Input: Parsed character identity.

[0919] Specific operation: The server generates the initial conversation scenario data (e.g., "Hello, this is AA. What would you like to talk about today?") and sends it to the generation AI as a prompt.

[0920] Output: Initial scenario data sent to the generation AI.

[0921] Step 6:

[0922] The generation AI generates conversational text based on the initial scenario data.

[0923] Input: Initial scenario data sent to the generating AI.

[0924] How it works: A generative AI model (e.g., OpenAI GPT-4) generates an initial conversation text based on the received prompt, imitating the character's personality and speaking style.

[0925] Output: The generated conversation text.

[0926] Step 7:

[0927] The server converts the generated conversation text into speech.

[0928] Input: The generated conversation text.

[0929] How it works: The server passes the generated conversation text to a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech) and converts it into audio data. The settings are adjusted to create the character's unique voice tone.

[0930] Output: The converted audio data.

[0931] Step 8:

[0932] The server sends the converted audio file to the terminal, which then plays the audio.

[0933] Input: The converted audio data.

[0934] Specific operation: The server transfers the generated audio file to the device, sending the audio data as an HTTP response. The device then plays the received audio file, allowing the user to hear the character's voice.

[0935] Output: The audio the user hears.

[0936] Step 9:

[0937] The user responds.

[0938] Input: The character's voice as heard by the user on their device.

[0939] Specific operation: After listening to the character's voice, the user responds.

[0940] Output: The user's spoken response.

[0941] Step 10:

[0942] The device records the user's response and converts it into text using a voice recognition engine.

[0943] Input: The user's voice reply.

[0944] Specific operation: The device records the user's voice and converts it into text data using an internal voice recognition engine (e.g., Google Cloud Speech-to-Text).

[0945] Output: User responses converted to text data.

[0946] Step 11:

[0947] The terminal transmits the converted text data to the server.

[0948] Input: User responses converted to text data.

[0949] Specific operation: The terminal packages the converted text data in JSON format and sends it to the server as an HTTP POST request.

[0950] Output: The text data of the user response sent to the server.

[0951] Step 12:

[0952] The server sends the next scenario data to the generation AI based on the resent text.

[0953] Input: Text data of the user's response sent from the terminal.

[0954] Specific operation: The server receives the user's response text and passes it to the generation AI as the next conversation scenario. It creates a prompt sentence to input to the generation AI.

[0955] Output: Send the following conversation scenario data to the generation AI.

[0956] Step 13:

[0957] The generation AI will generate the following conversation text again.

[0958] Input: The prompt sent to the generation AI as the following scenario data.

[0959] How it works: The AI ​​generates the next conversation text based on the user's response. This process is repeated until the user ends the conversation.

[0960] Output: The following conversation text.

[0961] Through this series of steps, users can enjoy natural conversations in real time with their chosen animal character.

[0962] (Application example 1)

[0963] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0964] In conventional character conversation systems, conversations between users and characters are often scripted, making it difficult to provide natural conversations in real time. Furthermore, there was a lack of technology to generate appropriate responses to user voice input and play them back in a natural voice, limiting the user experience.

[0965] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0966] In this invention, the server includes: means for a user to select a character and make a call request using an information processing device; means for the server to receive a call request including character selection information from the information processing device and transmit first conversation scenario data to the generation AI model; means for the generation AI model to generate conversation text based on the scenario data; means for the server to convert the generated conversation text into voice data; means for transmitting voice data from the server to the information processing device and for the information processing device to play the voice; means for the user to respond to the information processing device, converting the response into text data and retransmitting it to the server; means for the server to pass next conversation scenario data to the generation AI model based on the retransmitted text data; and means for converting the conversation text of the generation AI model into voice using a text-to-speech engine and providing the conversation data in real time, thereby enabling users to enjoy natural conversations in real time.

[0967] "Information processing device" is a general term for electronic devices used for processing and communicating data, and specifically includes smartphones, tablets, personal computers, etc.

[0968] A "server" is a central computer system that receives requests from multiple terminals on a network and provides data processing and communication services.

[0969] "Character selection information" is data for specifying the character with which the user wishes to converse, and this data is received by the server.

[0970] A "call request" is a request made by a user to start a conversation with a selected character using an information processing device.

[0971] "First conversation scenario data" is conversation script data generated to smoothly start the first conversation between a character and a user.

[0972] A "generative AI model" is an artificial intelligence model used to generate natural-sounding conversational text, using machine learning to generate responses in response to user input.

[0973] "Conversation text" is text data generated by a generative AI model that indicates the content of a conversation between a user and a character.

[0974] "Audio data" refers to an audio file generated from text data using speech synthesis technology, and is played back by an information processing device.

[0975] "Text data" refers to character information to be displayed or processed on an information processing device, and includes user responses and generated conversation content.

[0976] A "text-to-speech engine" is a software component that converts text data into audio data, enabling playback in a natural voice.

[0977] "Conversation scenario data" is data that indicates the next conversation content that the generative AI model will generate based on user input, enabling continuous conversation between the character and the user.

[0978] "Real-time" means that the conversation between the user and the character is immediate, with natural responses provided without delay.

[0979] A specific embodiment of the present invention will be described below: This system allows a user to use an information processing device to enjoy natural conversations with a specific character in real time.

[0980] First, the user launches a dedicated application using an information processing device (e.g., a smartphone, tablet, or head-mounted display (HMD)). Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button, which sends a call request to the server. The call request includes the identification information of the character selected by the user.

[0981] The server receives the call request and analyzes its content. Based on the analysis results, the server passes the initial conversation scenario data to the generative AI model. This initial conversation scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[0982] A generative AI model (e.g., GPT-3) generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing a natural and familiar conversation to the user. This conversational text is sent to a server and converted into audio data using a text-to-speech engine (TTS engine).

[0983] The voice data is sent from the server to the information processing device, which then plays back the voice data. This allows the user to hear the character's voice. The user listens to the conversation and responds. The information processing device records the user's response and converts the voice into text data using a voice recognition engine (e.g., Google Speech API). This text data is then sent back to the server.

[0984] The server receives the text data sent by the user and passes it back to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, which the server converts again into voice data using a TTS engine and sends and plays back to the information processing device. This process is repeated, allowing for a natural conversation between the user and the character to continue.

[0985] As a concrete example, suppose a user asks, "What did you do today?" The server passes this question to a generative AI model, which generates a response such as, "I took a walk in the park today. I was pleasantly surprised to hear the birds singing." This response is converted by a TTS engine, sent to an information processing device, and played back. The user can listen to this and continue the conversation.

[0986] Examples of prompts include:

[0987] User Input:

[0988] User: What did you do today?

[0989] Response from the generative AI model:

[0990] Character: I was walking in the park today. I could hear a lot of birds singing and it felt great.

[0991] The system allows users to engage in natural conversations with selected characters in real time, providing a familiar and engaging experience especially for children.

[0992] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0993] Step 1:

[0994] The user launches a dedicated application using an information processing device. Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button. Identification information for the character selected by the user is then generated and sent to the server as a call request. The input is the user's character selection, and the output is the call request data.

[0995] Step 2:

[0996] The server receives a call request sent from the user's information processing device. The server then analyzes the request content and extracts the character's identification information. The input is the call request data, and the output is the character's identification information.

[0997] Step 3:

[0998] The server passes the first conversation scenario data to the generation AI model based on the character's identification information. This scenario data includes the content of the first conversation based on the character's personality and speaking style. The input is the character's identification information, and the output is the first conversation scenario data.

[0999] Step 4:

[1000] The generative AI model generates conversational text based on the received initial conversation scenario data. For example, it generates the content "Hello, this is AA. What would you like to talk about today?" The input is the initial conversation scenario data, and the output is the conversational text.

[1001] Step 5:

[1002] The server receives the conversational text from the generative AI model and converts it into audio data using a text-to-speech engine (TTS engine), where the character's unique voice is reproduced. The input is the conversational text, and the output is audio data.

[1003] Step 6:

[1004] The server transmits the generated voice data to the information processing device, which receives the voice data and plays it back for the user to hear. The input is the voice data, and the output is the situation in which the user can hear the voice.

[1005] Step 7:

[1006] The user listens to the character's voice and responds. The information processing device records the user's response and converts it into text data using a voice recognition engine. The input is the user's voice, and the output is text data.

[1007] Step 8:

[1008] The information processing device sends the user's response text data to the server. The server receives this text data and passes it to the generative AI model as the next conversation scenario data. The input is the user's response text data, and the output is the next conversation scenario data.

[1009] Step 9:

[1010] The generative AI model generates new conversation text based on the next conversation scenario data. The generated conversation text is sent to the server again. The input is the next conversation scenario data, and the output is the new conversation text.

[1011] Step 10:

[1012] The processes from step 5 to step 9 are repeated to continue a natural conversation between the user and the character, thereby maintaining a natural conversation in real time.

[1013] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1014] This invention is a system that enables a user to use a terminal to have real-time telephone conversations with famous animal characters, and that recognizes the user's emotions and adjusts the content of the conversation.

[1015] Character selection and call request

[1016] First, the user opens the app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. The selection information includes the character specified by the user.

[1017] Receiving a call request and generating a first conversation scenario

[1018] The server analyzes the call request received from the device and extracts the selected character information. Based on this information, the server sends the initial conversation scenario data to the generation AI, instructing it to generate an initial greeting such as, "Hello, this is XX. What would you like to talk about today?"

[1019] Conversational text generation and speech conversion

[1020] The generation AI generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. The server receives this dialogue text and converts it into audio data using a text-to-speech (TTS) engine. The converted audio data is generated in the character's voice.

[1021] Sending and playing audio files

[1022] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user, who then enjoys the conversation while listening to the animal characters' voices.

[1023] Processing user responses and sentiment analysis

[1024] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine. The converted text data is then resent to the server, where the emotion engine extracts and analyzes emotional information from the user's speech and text.

[1025] The process of emotion-based conversation generation

[1026] The server receives the text data sent by the user and the analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI. The generation AI generates conversation text according to the user's emotions. For example, if the user replies in a happy voice, "I had a great time at the picnic in the park today," the generation AI will generate a conversation such as, "That's great! Who were your friends playing with?"

[1027] Sending and playing audio files

[1028] The server then converts the newly generated conversation text into audio data using a text-to-speech engine and sends it to the device. The device then plays the received audio file back to the user. This process is repeated, allowing for a natural conversation between the user and the character to continue, and providing responses that correspond to the user's emotions.

[1029] Specific examples

[1030] For example, suppose a user says aloud to a device, "I was so happy my teacher praised me at school today!" The device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a text-to-speech engine and sent to the device for playback.

[1031] In this way, the system, combined with the emotion engine, enables more natural and friendly conversations based on the user's emotions, making interactions with characters that understand and respond to emotions even more appealing, especially for children.

[1032] The processing flow will be explained below.

[1033] Step 1:

[1034] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[1035] Step 2:

[1036] The server analyzes the call request received from the device and extracts the selected character information. Based on the character information obtained from the analysis, the initial conversation scenario data is provided to the generation AI.

[1037] Step 3:

[1038] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[1039] Step 4:

[1040] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[1041] Step 5:

[1042] The server sends the converted audio file to the device, which then plays the received audio file, allowing the user to hear the character's voice.

[1043] Step 6:

[1044] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[1045] Step 7:

[1046] The device retransmits the text data to the server, and simultaneously transmits the recorded data to the server, where the emotion engine extracts and analyzes emotional information from the user's speech.

[1047] Step 8:

[1048] The server provides the next conversation scenario data to the generation AI based on the received text data and emotional information. The generation AI generates the optimal response text based on the user's emotions analyzed by the emotion engine.

[1049] Step 9:

[1050] The generative AI generates new conversational text based on the user's responses and emotional information. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[1051] Step 10:

[1052] The server converts the generated new conversation text into voice data using a text-to-speech (TTS) engine again and sends it to the terminal.

[1053] Step 11:

[1054] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character, while providing responses that correspond to the user's emotions.

[1055] Step 12:

[1056] As a concrete example, if a user says to a device, "I was so happy when my teacher praised me at school today!", the device records this voice and converts it into text using a speech recognition engine. The server then passes this text data and the emotional information of "happy" analyzed by the emotion engine to the generation AI, which then generates a response such as, "That's wonderful! What did your teacher praise you for?" This response is converted into audio and played on the device.

[1057] Example 2

[1058] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1059] Conventional conversational systems have the problem that the conversation between the user and the character is formulaic and lacks responses that reflect the user's emotions. In particular, there is a need for a system that can recognize the user's emotions and generate natural conversations accordingly.

[1060] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1061] In this invention, the server includes a means for the user to respond to the terminal, converting the response into text and retransmitting it to the server, a means for the server to pass the next scenario data to the generation AI based on the retransmitted text, a means for the terminal to record the user's response and convert it into text data using a voice recognition engine, and a means for the server to extract and analyze the user's emotional information using an emotion analysis engine, thereby enabling a response that corresponds to the user's emotions.

[1062] A "terminal" is an electronic device such as a computer or smartphone that is operated by a user.

[1063] A "server" is a computer system that receives requests and data from terminals, processes them, and gives instructions to the generating AI.

[1064] "Generative AI" is an artificial intelligence model that generates conversational text based on scenario data.

[1065] "Character selection information" is information about the animal character with which the user wants to converse.

[1066] A "call request" is request data that a user sends from a terminal to a server in order to start a call with a character.

[1067] "Scenario data" is the initial set of information that the generation AI uses to generate conversational text.

[1068] "Conversational text" is a sentence generated by the generation AI based on scenario data.

[1069] "Audio data" refers to an audio file converted from text by a text-to-speech engine.

[1070] An "audio file" is a digital file containing audio data that is sent from a server to a terminal.

[1071] A "voice recognition engine" is software that records a user's speech and converts it into text data.

[1072] An "emotion analysis engine" is software that extracts and analyzes emotional information from a user's speech or text.

[1073] A "text-to-speech engine" is software that converts text data into voice data.

[1074] This invention is a system that allows users to use their terminals to have real-time telephone conversations with famous animal characters. Furthermore, it can recognize the user's emotions and adjust the content of the conversation. The detailed implementation method of this system is described below.

[1075] Hardware and software used

[1076] Device: A device operated by a user, such as a smartphone or tablet.

[1077] Server: A computer system that processes user requests and interacts with generative AI models.

[1078] Generative AI model: An artificial intelligence model for generating conversational text.

[1079] Speech recognition engine: Software that records what a user says and converts it into text data (e.g., Google Speech-to-Text).

[1080] Sentiment analysis engine: Software that extracts and analyzes users' emotional information (e.g., IBM Watson Tone Analyzer).

[1081] Text-to-speech engine: Software that converts text data into speech data (e.g., Amazon Polly).

[1082] Specific operation of the system

[1083] 1. Select a character and send a call request:

[1084] Users use their device to open a dedicated app and select the animal character they want to talk to.

[1085] When the user presses the "start call" button, a call request including character selection information is sent from the terminal to the server.

[1086] 2. Generating the first conversation scenario:

[1087] The server analyzes the call request received from the terminal and extracts the selected character information.

[1088] The server sends the initial conversation scenario data to the generation AI and instructs it to generate the initial greeting.

[1089] 3. Conversational text generation and speech conversion:

[1090] The generation AI generates conversational text based on the initial scenario data.

[1091] The server receives the generated text and converts it into audio data using a text-to-speech engine.

[1092] 4. Send and play audio files:

[1093] The server sends the converted audio file to the terminal.

[1094] The terminal receives the audio file and plays it for the user.

[1095] 5. User response processing and sentiment analysis:

[1096] The user listens to what the character says and responds, for example, by asking, "What did you do today?"

[1097] The device records the user's response and converts it into text data using a voice recognition engine.

[1098] The device sends the converted text data to the server, and the emotion analysis engine extracts and analyzes the user's emotional information.

[1099] 6. Emotion-based conversation generation:

[1100] The server receives the text data and emotional information sent by the user, and based on that, sends the next conversation scenario data to the generation AI.

[1101] The AI ​​generates conversational text based on the user's emotions. For example, if the user speaks in a happy tone, the AI ​​generates a positive response.

[1102] 7. Sending and playing audio files:

[1103] The server converts the generated new conversation text into voice data again using the text-to-speech engine.

[1104] The server transmits the converted audio file to the terminal, which receives it and plays it back.

[1105] This process is repeated, allowing for a natural conversation between the user and the character to continue.

[1106] Specific examples

[1107] For example, consider the case where a user says to a device, "I was so happy when my teacher praised me at school today!" The device records this voice and converts it into text using a speech recognition engine. The resulting text is sent to a server, which then uses an emotion analysis engine to determine that the user's emotion is "happy." Based on this determination, the generation AI generates a response such as, "That's wonderful! What did the teacher praise you for?" This response is converted into audio using a text-to-speech engine and sent to and played back on the device.

[1108] Prompt Sentence Examples

[1109] Here are some example prompts you can enter into your generative AI model:

[1110] 1. "If the user says in a cheerful voice, 'I had a great time picnicking in the park today,' generate an appropriate response in the following conversation."

[1111] 2. "If the user sadly says, 'I had a fight with my friend,' generate a comforting response in the following conversation."

[1112] In this way, by combining emotion analysis and generative AI, this system is able to respond according to the user's emotions.

[1113] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1114] Step 1:

[1115] The user opens the app on their device and selects the animal character they want to talk to. When they press the "Start Call" button, a call request including the character selection information is sent from the device to the server. The input here is the user's selection and the call start action, and the output is the call request being sent to the server.

[1116] Step 2:

[1117] The server analyzes the call request received from the device. The request contains character selection information, and extracts this information. The input is the call request sent from the device, and the output is the extracted character selection information. The server generates the initial conversation scenario data based on this information.

[1118] Step 3:

[1119] Based on the extracted character information, the server sends the initial conversation scenario data to the generation AI. An instruction is sent to generate an initial greeting such as "Hello, this is □□. What would you like to talk about today?" The input here is character selection information, and the output is scenario data for the generation AI.

[1120] Step 4:

[1121] The generation AI generates dialogue text based on the initial scenario data. The generated text is adjusted to match the character's personality and speaking style. The input is the initial scenario data, and the output is the generated dialogue text.

[1122] Step 5:

[1123] The server receives the generated text and converts it into audio data using a text-to-speech engine. The converted audio data is generated in the character's voice. Here, the input is the spoken text and the output is audio data.

[1124] Step 6:

[1125] The server sends the converted audio file to the device. The input is audio data, and the output is an audio file sent to the device. The device receives this audio file and plays it for the user. The user can enjoy the conversation while listening to the voices of the animal characters.

[1126] Step 7:

[1127] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input here is the user's response, which the device records. The output is the recorded voice data.

[1128] Step 8:

[1129] The device records the user's response and converts it into text using a speech recognition engine. The input is the voice data and the output is the converted text data, which is then sent to the server.

[1130] Step 9:

[1131] The server receives the converted text data and uses a sentiment analysis engine to extract and analyze the user's sentiment information, where the input is the converted text data and the output is the extracted sentiment information.

[1132] Step 10:

[1133] The server receives the text data and emotional information sent by the user and sends the next conversation scenario data to the generation AI based on that. The input here is the text data and emotional information, and the output is the next scenario data for the generation AI.

[1134] Step 11:

[1135] The generative AI generates conversational text according to the user's emotions. For example, if the user speaks in a cheerful manner, the generative AI generates a positive response. The input here is the following scenario data, and the output is the generated conversational text.

[1136] Step 12:

[1137] The server converts the newly generated conversation text into audio data again using a text-to-speech engine, where the input is the new conversation text and the output is audio data.

[1138] Step 13:

[1139] The server sends the converted audio file to the device, which then receives and plays it. The input here is audio data, and the output is the audio file sent to the device. This allows for a natural conversation between the user and the character to continue.

[1140] (Application example 2)

[1141] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1142] Conventional character dialogue systems have a formulaic user experience and have difficulty providing appropriate responses based on the user's emotions. As a result, particularly in interactive storytelling, users are less likely to feel a sense of realism or familiarity, and the dialogue tends to be uniform. The present invention aims to solve this problem by providing a more natural, interactive conversation experience that responds to the user's emotions.

[1143] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for the user to select a character and make a call request using a terminal, means for the server to receive a call request including character selection information from the terminal and send initial scenario data to the generation AI, and means for the generation AI to generate conversation text based on the scenario data. This makes it possible to analyze the user's emotional information in real time and generate and provide dynamic conversation content accordingly.

[1144] "Terminal" refers to an electronic device through which a user selects a character and makes a call request.

[1145] "Server" refers to the central processing unit that receives information from the terminal and works with the generative AI and other engines to generate and manage conversation scenarios.

[1146] "Character selection information" refers to information about a specific character selected by a user through a terminal.

[1147] A "call request" refers to request information sent from a terminal to a server in order for a user to start a conversation with a character.

[1148] "Initial scenario data" refers to the scenario information at the initial stage of the conversation that the server sends to the generation AI.

[1149] "Generative AI" refers to an artificial intelligence model that generates conversational text based on user input and scenario data.

[1150] "Conversational text" refers to the text data of the dialogue between the character generated by the generation AI and the user.

[1151] "Means for converting to speech" refers to a text-to-speech (TTS) engine for converting the generated conversational text into audio data.

[1152] "Audio file" refers to audio data sent from the server to the terminal and played as the character's voice.

[1153] The "means for converting a response into text" refers to a voice recognition engine that converts a voice response given by a user to a terminal into text data.

[1154] "Emotion analysis engine" refers to a program and device for analyzing emotional information from the content and voice of a user's responses.

[1155] "Emotion information" refers to data that indicates the type and intensity of emotions extracted from the user's responses.

[1156] This invention is a system that allows users to have real-time conversations with famous characters using a terminal, and adjusts the content of the conversation based on the user's emotions. This system is implemented by combining the following steps and hardware / software:

[1157] System Overview

[1158] The user selects a character and makes a call request through the device. The server receives this request information and generates initial scenario data based on the generation AI. The generation AI creates conversation text based on the generated scenario data, and the server converts this text into audio data using a text-to-speech (TTS) engine. This audio data is sent to the device and played back to the user.

[1159] Main processing

[1160] 1. Select a character and send a call request

[1161] The user opens the dedicated app on their device and selects the character they want to talk to. After selecting, the user presses the "Start Call" button, and a call request including the character selection information is sent to the server.

[1162] 2. Receiving a call request and generating a first conversation scenario

[1163] The server receives the call request and sends the initial scenario data based on the character information to the generation AI. The generation AI then generates an initial greeting such as, "Hello, this is □□. What would you like to talk about today?"

[1164] 3. Conversational text generation and speech conversion

[1165] The AI ​​generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. This dialogue text is then converted into voice data using a TTS engine.

[1166] 4. Sending and playing audio files

[1167] The server then sends the converted audio file to the device, which then plays it back, allowing the user to enjoy the conversation while listening to the character's voice.

[1168] Processing user responses and sentiment analysis

[1169] The user listens to what the character says and responds. The device records the user's response and converts it into text data using a voice recognition engine. The converted text data is then retransmitted to the server, which then uses an emotion analysis engine to extract and analyze the user's emotional information.

[1170] The process of emotion-based conversation generation

[1171] The server receives the user's text data and analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI, which then generates conversation text according to the user's emotions.

[1172] Hardware and software used

[1173] Hardware: Smartphones, servers

[1174] Software: Dedicated apps, generative AI models (e.g., GPT-3), text-to-speech (TTS) engines, speech recognition engines, sentiment analysis engines

[1175] Specific examples

[1176] For example, if a user says to a device, "I was so happy my teacher praised me at school today!", the device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion analysis engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a TTS engine and sent to the device for playback.

[1177] Prompt Sentence Examples

[1178] User: I was so happy when my teacher praised me at school today!

[1179] Prompt for the generated AI:

[1180] Generate a follow-up response in a friendly and cheerful tone for a user who just said, "I was so happy that my teacher praised me at school today!" The response should include a question to keep the conversation going.

[1181] This allows for natural, emotionally-responsive interactive conversation between the user and the character.

[1182] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1183] Step 1:

[1184] The user opens the dedicated app on their device. Within the app, the user selects the character they want to talk to and presses the "Start Call" button. This generates a call request including character selection information and sends it to the server. In this process, the input is the character information selected by the user, and the output is the call request.

[1185] Step 2:

[1186] The server analyzes the call request received from the device and extracts character selection information. Based on this information, the server sends the initial scenario data to the generation AI. The input is the call request, and the output is the initial scenario data to be passed to the generation AI.

[1187] Step 3:

[1188] The generation AI generates conversational text based on the initial scenario data. For example, the generation AI generates the initial greeting message, "Hello, this is □□. What would you like to talk about today?" In this case, the input is the initial scenario data, and the output is the conversational text.

[1189] Step 4:

[1190] The server receives the generated conversation text and converts it into audio data using a text-to-speech (TTS) engine. The input is the conversation text and the output is an audio file.

[1191] Step 5:

[1192] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user. The input is the audio file and the output is the played audio.

[1193] Step 6:

[1194] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input is the user's response to the character's utterance, and the output is the response as voice data.

[1195] Step 7:

[1196] The device records the user's response and converts it into text data using a speech recognition engine. The input is the user's voice response, and the output is the text-converted response data.

[1197] Step 8:

[1198] The recorded text data is retransmitted to the server, where the server uses an emotion analysis engine to extract and analyze emotional information from the user's speech and text. The input is the text-converted response data, and the output is the extracted emotional information.

[1199] Step 9:

[1200] The server receives the text data and analyzed emotional information sent by the user, and then sends the next conversation scenario data to the generation AI based on that. The input is the text data and emotional information, and the output is the next scenario data.

[1201] Step 10:

[1202] The generation AI generates conversational text according to the user's emotions. For example, if the user replies, "I had a great time at the picnic in the park today," the generation AI generates the following conversational content: "That's great! Who were your friends playing with?" The input is the next scenario data, and the output is new conversational text.

[1203] Step 11:

[1204] The server converts the generated new conversation text into audio data using the text-to-speech engine again and sends it to the terminal. The input is the new conversation text, and the output is an audio file as audio data.

[1205] Step 12:

[1206] The device then plays the received audio file back to the user, allowing for a natural conversation between the user and the character, and providing responses according to the user's emotions. The input is the audio file, and the output is the played audio.

[1207] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1208] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1209] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1210] [Fourth embodiment]

[1211] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1212] 7, a 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.

[1213] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1214] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1215] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1216] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1217] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1218] The control object 443 includes a display device, LEDs in the eyes, and motors for driving 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1219] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1220] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1221] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1222] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1223] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1224] This invention is a system that allows users to use a terminal to have real-time telephone conversations with famous animal characters. Below, we will explain the program processing of this system and specific examples.

[1225] Character selection and call request

[1226] First, the user opens the dedicated app on their device. Within the app, they select the animal character they want to talk to. Next, when the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. This request also includes the identification information of the selected character.

[1227] Receiving a call request and generating a first conversation scenario

[1228] The server receives a call request from the device and analyzes the request. Based on the character information obtained as a result of the analysis, the server passes the initial conversation scenario data to the generation AI. This scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[1229] Conversational text generation

[1230] The AI ​​generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing the user with natural and familiar conversations.

[1231] Conversational text-to-speech

[1232] The server receives the generated conversational text from the AI ​​generator and converts it into audio using a text-to-speech (TTS) engine. The converted audio data is then tuned to play in the character's unique voice.

[1233] Sending and playing audio files

[1234] The server sends the converted audio file to the device, which then receives the file and plays it for the user to listen to. The user can enjoy the conversation while listening to the character's voice.

[1235] Handling the user's response

[1236] The user listens to the conversation and responds. The device records the user's response and converts it into text data using a speech recognition engine. This text data is then sent back to the server.

[1237] Next conversation generation process

[1238] The server receives the text data sent by the user and passes it to the generation AI as the next conversation scenario data. The generation AI generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[1239] Specific examples

[1240] For example, suppose a user asks, "What did you do today?" The server passes this question to the generation AI, which then generates a response such as, "I took a walk in the park today. I could hear lots of birds and it was very pleasant." This response is then converted into audio by the text-to-speech engine and sent to and played back on the device. The user can then listen to this and continue the conversation.

[1241] In this way, the system allows users to have natural conversations in real time with their chosen animal characters, providing an engaging and relatable experience especially for children.

[1242] The processing flow will be explained below.

[1243] Step 1:

[1244] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[1245] Step 2:

[1246] The server analyzes the call request received from the device and extracts the selected character information.The server then sends the initial conversation scenario data to the generation AI, instructing it to generate an opening greeting and introductory sentence.

[1247] Step 3:

[1248] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[1249] Step 4:

[1250] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[1251] Step 5:

[1252] The server sends the converted audio file to the terminal, which then plays the received audio file for the user to listen to.

[1253] Step 6:

[1254] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[1255] Step 7:

[1256] The device resends the converted text data to the server, which analyzes the received user response and passes the next conversation scenario data based on that to the generation AI.

[1257] Step 8:

[1258] The generative AI generates new conversational text based on the user's response. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[1259] Step 9:

[1260] The server converts the generated new conversation text into voice data using the text-to-speech engine again, and transmits the voice data to the terminal.

[1261] Step 10:

[1262] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character.

[1263] Example 1

[1264] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1265] Conventional systems have struggled to enable users to have natural, real-time conversations with their chosen characters, resulting in limited user experience. Furthermore, character voice generation and user response processing often lacked smoothness, resulting in delayed conversations and unnatural responses. These issues pose a significant obstacle, particularly for children, where familiarity and naturalness are crucial.

[1266] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1267] In this invention, the server includes: means for a user to use a terminal to select a character and make a call request; means for the terminal to send a call request including character selection information to the server; means for the server to receive and analyze the call request; means for the server to send initial scenario data to the generation AI; means for the generation AI to generate conversation text based on the initial scenario data; means for the server to convert the generated conversation text into voice; means for the server to send the converted voice file to the terminal and the terminal to play the voice; means for the user to respond to the terminal, convert the response into text using a voice recognition engine, and resend it to the server; and means for the server to send next scenario data to the generation AI based on the resent text. This enables users to have natural conversations in real time with their selected characters, providing an experience that is particularly familiar to children.

[1268] "User" refers to a person who uses a terminal to select a character and make a call request.

[1269] A "terminal" is a device operated by a user, which transmits a call request including character selection information to a server and plays back audio received from the server.

[1270] "Character selection information" refers to information for identifying a specific character selected by a user.

[1271] "Server" refers to a computer system on a network that receives and analyzes call requests sent from a device and works in conjunction with the generation AI and voice conversion function.

[1272] A "call request" refers to a request to the server to initiate a real-time conversation, including identification information for a user-selected character.

[1273] "Initial scenario data" refers to the data that defines the initial conversation scenario that the server sends to the generation AI.

[1274] "Generative AI" refers to an artificial intelligence model that generates conversational text based on received scenario data.

[1275] "Conversational text" refers to text that describes what a character will say to the user, generated by the generation AI based on scenario data.

[1276] "Convert to speech" refers to converting the conversational text generated by the generative AI into audio data using a text-to-speech engine.

[1277] "Speech recognition engine" refers to software or algorithms for converting a user's spoken responses into text data.

[1278] "Scenario data" refers to data that is passed to the generation AI to generate the content of the conversation.

[1279] This invention is a system that allows users to have natural conversations in real time with selected animal characters using a terminal. The core of this system is composed of a server, a terminal, and a generative AI model working together. Specific embodiments of the system are described below.

[1280] First, the user uses a device such as a smartphone or tablet. The user launches a dedicated app and selects the animal character they want to talk to from the displayed characters. For example, characters such as rabbits and bears can be selected. After selecting the character, the user presses the "Start Call" button. This action causes the device to package a call request in JSON format, including the identification information of the selected character, and send it to the server as an HTTP POST request.

[1281] When the server receives a call request, it analyzes the request and extracts the identification information of the selected character. The server then generates initial conversation scenario data and passes it to a generative AI model (e.g., OpenAI's GPT-4). The initial scenario data includes a prompt such as, "Hello, this is AA. What would you like to talk about today?"

[1282] The generative AI model generates dialogue text based on scenario data, imitating the character's personality and speaking style. The dialogue text is then sent to a server, which converts the received text into audio data using a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech). This audio data is then adjusted to be played in the character's unique voice.

[1283] The server sends the converted audio file to the device, which receives and plays it. The user can then hear the character's voice through the device and begin the conversation.

[1284] When the user responds to the device, the device records the voice and converts it into text data using an internal speech recognition engine (e.g., Google Cloud Speech-to-Text). The converted text data is then sent back to the server.

[1285] The server receives the text data sent by the user and passes it to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, and this process is repeated to continue the natural conversation between the user and the character.

[1286] As a concrete example, consider the case where a user asks, "What did you do today?" Below is an example of a prompt sentence to input to the generation AI.

[1287] "Generate a conversation scenario in which the user asks, 'What did you do today?' and the character replies that he was walking in the park."

[1288] Based on this prompt, the generative AI model generates a response text such as, "I was walking in the park today. I could hear lots of birds singing and it felt great." This response text is converted into speech via the server and sent and played back on the device, allowing the user to enjoy a conversation with the character.

[1289] As a result, the system allows users to have natural conversations with selected animal characters in real time, providing an experience that is particularly friendly to children.

[1290] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1291] Step 1:

[1292] The user launches the dedicated app and selects an animal character.

[1293] Input: A device such as a smartphone or tablet operated by the user.

[1294] Specific operation: The user checks the list of available animal characters on the device's app screen and taps the desired character to select it.

[1295] Output: The identity of the selected character.

[1296] Step 2:

[1297] The user presses the "start call" button.

[1298] Input: The user-selected character's identity.

[1299] Specific operation: The user taps the "Start Call" button on the app screen. This action prepares the device to perform the following process.

[1300] Output: Call request trigger.

[1301] Step 3:

[1302] The terminal transmits a call request including character selection information to the server.

[1303] Input: The identity of the selected character, a trigger indicating that the "Start Call" button has been pressed.

[1304] Specific operation: The device creates a call request in JSON format including the identification information of the selected character and sends it to the server as an HTTP POST request.

[1305] Output: The call request data sent to the server.

[1306] Step 4:

[1307] The server receives and analyzes the call request.

[1308] Input: Call request data sent from the device.

[1309] Specific operation: The server receives the call request, parses the JSON data to extract the character's identification information, confirms the selection information, and prepares for the next process.

[1310] Output: Parsed character identities.

[1311] Step 5:

[1312] The server sends the initial scenario data to the generation AI.

[1313] Input: Parsed character identity.

[1314] Specific operation: The server generates the initial conversation scenario data (e.g., "Hello, this is AA. What would you like to talk about today?") and sends it to the generation AI as a prompt.

[1315] Output: Initial scenario data sent to the generation AI.

[1316] Step 6:

[1317] The generation AI generates conversational text based on the initial scenario data.

[1318] Input: Initial scenario data sent to the generating AI.

[1319] How it works: A generative AI model (e.g., OpenAI GPT-4) generates an initial conversation text based on the received prompt, imitating the character's personality and speaking style.

[1320] Output: The generated conversation text.

[1321] Step 7:

[1322] The server converts the generated conversation text into speech.

[1323] Input: The generated conversation text.

[1324] How it works: The server passes the generated conversation text to a text-to-speech (TTS) engine (e.g., Google Cloud Text-to-Speech) and converts it into audio data. The settings are adjusted to create the character's unique voice tone.

[1325] Output: The converted audio data.

[1326] Step 8:

[1327] The server sends the converted audio file to the terminal, which then plays the audio.

[1328] Input: The converted audio data.

[1329] Specific operation: The server transfers the generated audio file to the device, sending the audio data as an HTTP response. The device then plays the received audio file, allowing the user to hear the character's voice.

[1330] Output: The audio the user hears.

[1331] Step 9:

[1332] The user responds.

[1333] Input: The character's voice as heard by the user on their device.

[1334] Specific operation: After listening to the character's voice, the user responds.

[1335] Output: The user's spoken response.

[1336] Step 10:

[1337] The device records the user's response and converts it into text using a voice recognition engine.

[1338] Input: The user's voice reply.

[1339] Specific operation: The device records the user's voice and converts it into text data using an internal voice recognition engine (e.g., Google Cloud Speech-to-Text).

[1340] Output: User responses converted to text data.

[1341] Step 11:

[1342] The terminal transmits the converted text data to the server.

[1343] Input: User responses converted to text data.

[1344] Specific operation: The terminal packages the converted text data in JSON format and sends it to the server as an HTTP POST request.

[1345] Output: The text data of the user response sent to the server.

[1346] Step 12:

[1347] The server sends the next scenario data to the generation AI based on the resent text.

[1348] Input: Text data of the user's response sent from the terminal.

[1349] Specific operation: The server receives the user's response text and passes it to the generation AI as the next conversation scenario. It creates a prompt sentence to input to the generation AI.

[1350] Output: Send the following conversation scenario data to the generation AI.

[1351] Step 13:

[1352] The generation AI will generate the following conversation text again.

[1353] Input: The prompt sent to the generation AI as the following scenario data.

[1354] How it works: The AI ​​generates the next conversation text based on the user's response. This process is repeated until the user ends the conversation.

[1355] Output: The following conversation text.

[1356] Through this series of steps, users can enjoy natural conversations in real time with their chosen animal character.

[1357] (Application example 1)

[1358] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1359] In conventional character conversation systems, conversations between users and characters are often scripted, making it difficult to provide natural conversations in real time. Furthermore, there was a lack of technology to generate appropriate responses to user voice input and play them back in a natural voice, limiting the user experience.

[1360] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1361] In this invention, the server includes: means for a user to select a character and make a call request using an information processing device; means for the server to receive a call request including character selection information from the information processing device and transmit first conversation scenario data to the generation AI model; means for the generation AI model to generate conversation text based on the scenario data; means for the server to convert the generated conversation text into voice data; means for transmitting voice data from the server to the information processing device and for the information processing device to play the voice; means for the user to respond to the information processing device, converting the response into text data and retransmitting it to the server; means for the server to pass next conversation scenario data to the generation AI model based on the retransmitted text data; and means for converting the conversation text of the generation AI model into voice using a text-to-speech engine and providing the conversation data in real time, thereby enabling users to enjoy natural conversations in real time.

[1362] "Information processing device" is a general term for electronic devices used for processing and communicating data, and specifically includes smartphones, tablets, personal computers, etc.

[1363] A "server" is a central computer system that receives requests from multiple terminals on a network and provides data processing and communication services.

[1364] "Character selection information" is data for specifying the character with which the user wishes to converse, and this data is received by the server.

[1365] A "call request" is a request made by a user to start a conversation with a selected character using an information processing device.

[1366] "First conversation scenario data" is conversation script data generated to smoothly start the first conversation between a character and a user.

[1367] A "generative AI model" is an artificial intelligence model used to generate natural-sounding conversational text, using machine learning to generate responses in response to user input.

[1368] "Conversation text" is text data generated by a generative AI model that indicates the content of a conversation between a user and a character.

[1369] "Audio data" refers to an audio file generated from text data using speech synthesis technology, and is played back by an information processing device.

[1370] "Text data" refers to character information to be displayed or processed on an information processing device, and includes user responses and generated conversation content.

[1371] A "text-to-speech engine" is a software component that converts text data into audio data, enabling playback in a natural voice.

[1372] "Conversation scenario data" is data that indicates the next conversation content that the generative AI model will generate based on user input, enabling continuous conversation between the character and the user.

[1373] "Real-time" means that the conversation between the user and the character is immediate, with natural responses provided without delay.

[1374] A specific embodiment of the present invention will be described below: This system allows a user to use an information processing device to enjoy natural conversations with a specific character in real time.

[1375] First, the user launches a dedicated application using an information processing device (e.g., a smartphone, tablet, or head-mounted display (HMD)). Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button, which sends a call request to the server. The call request includes the identification information of the character selected by the user.

[1376] The server receives the call request and analyzes its content. Based on the analysis results, the server passes the initial conversation scenario data to the generative AI model. This initial conversation scenario data includes content such as, "Hello, this is AA. What would you like to talk about today?"

[1377] A generative AI model (e.g., GPT-3) generates conversational text based on the received initial scenario data. This generated text mimics the personality and speaking style of the selected character, providing a natural and familiar conversation to the user. This conversational text is sent to a server and converted into audio data using a text-to-speech engine (TTS engine).

[1378] The voice data is sent from the server to the information processing device, which then plays back the voice data. This allows the user to hear the character's voice. The user listens to the conversation and responds. The information processing device records the user's response and converts the voice into text data using a voice recognition engine (e.g., Google Speech API). This text data is then sent back to the server.

[1379] The server receives the text data sent by the user and passes it back to the generative AI model as the next conversation scenario data. The generative AI model generates new conversation text, which the server converts again into voice data using a TTS engine and sends and plays back to the information processing device. This process is repeated, allowing for a natural conversation between the user and the character to continue.

[1380] As a concrete example, suppose a user asks, "What did you do today?" The server passes this question to a generative AI model, which generates a response such as, "I took a walk in the park today. I was pleasantly surprised to hear the birds singing." This response is converted by a TTS engine, sent to an information processing device, and played back. The user can listen to this and continue the conversation.

[1381] Examples of prompts include:

[1382] User Input:

[1383] User: What did you do today?

[1384] Response from the generative AI model:

[1385] Character: I was walking in the park today. I could hear a lot of birds singing and it felt great.

[1386] The system allows users to engage in natural conversations with selected characters in real time, providing a familiar and engaging experience especially for children.

[1387] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1388] Step 1:

[1389] The user launches a dedicated application using an information processing device. Within the application, the user selects the character with whom they want to talk and presses the "Start Call" button. Identification information for the character selected by the user is then generated and sent to the server as a call request. The input is the user's character selection, and the output is the call request data.

[1390] Step 2:

[1391] The server receives a call request sent from the user's information processing device. The server then analyzes the request content and extracts the character's identification information. The input is the call request data, and the output is the character's identification information.

[1392] Step 3:

[1393] The server passes the first conversation scenario data to the generation AI model based on the character's identification information. This scenario data includes the content of the first conversation based on the character's personality and speaking style. The input is the character's identification information, and the output is the first conversation scenario data.

[1394] Step 4:

[1395] The generative AI model generates conversational text based on the received initial conversation scenario data. For example, it generates the content "Hello, this is AA. What would you like to talk about today?" The input is the initial conversation scenario data, and the output is the conversational text.

[1396] Step 5:

[1397] The server receives the conversational text from the generative AI model and converts it into audio data using a text-to-speech engine (TTS engine), where the character's unique voice is reproduced. The input is the conversational text, and the output is audio data.

[1398] Step 6:

[1399] The server transmits the generated voice data to the information processing device, which receives the voice data and plays it back for the user to hear. The input is the voice data, and the output is the situation in which the user can hear the voice.

[1400] Step 7:

[1401] The user listens to the character's voice and responds. The information processing device records the user's response and converts it into text data using a voice recognition engine. The input is the user's voice, and the output is text data.

[1402] Step 8:

[1403] The information processing device sends the user's response text data to the server. The server receives this text data and passes it to the generative AI model as the next conversation scenario data. The input is the user's response text data, and the output is the next conversation scenario data.

[1404] Step 9:

[1405] The generative AI model generates new conversation text based on the next conversation scenario data. The generated conversation text is sent to the server again. The input is the next conversation scenario data, and the output is the new conversation text.

[1406] Step 10:

[1407] The processes from step 5 to step 9 are repeated to continue a natural conversation between the user and the character, thereby maintaining a natural conversation in real time.

[1408] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1409] This invention is a system that enables a user to use a terminal to have real-time telephone conversations with famous animal characters, and that recognizes the user's emotions and adjusts the content of the conversation.

[1410] Character selection and call request

[1411] First, the user opens the app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request to the server, including character selection information. The selection information includes the character specified by the user.

[1412] Receiving a call request and generating a first conversation scenario

[1413] The server analyzes the call request received from the device and extracts the selected character information. Based on this information, the server sends the initial conversation scenario data to the generation AI, instructing it to generate an initial greeting such as, "Hello, this is XX. What would you like to talk about today?"

[1414] Conversational text generation and speech conversion

[1415] The generation AI generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. The server receives this dialogue text and converts it into audio data using a text-to-speech (TTS) engine. The converted audio data is generated in the character's voice.

[1416] Sending and playing audio files

[1417] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user, who then enjoys the conversation while listening to the animal characters' voices.

[1418] Processing user responses and sentiment analysis

[1419] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine. The converted text data is then resent to the server, where the emotion engine extracts and analyzes emotional information from the user's speech and text.

[1420] The process of emotion-based conversation generation

[1421] The server receives the text data sent by the user and the analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI. The generation AI generates conversation text according to the user's emotions. For example, if the user replies in a happy voice, "I had a great time at the picnic in the park today," the generation AI will generate a conversation such as, "That's great! Who were your friends playing with?"

[1422] Sending and playing audio files

[1423] The server then converts the newly generated conversation text into audio data using a text-to-speech engine and sends it to the device. The device then plays the received audio file back to the user. This process is repeated, allowing for a natural conversation between the user and the character to continue, and providing responses that correspond to the user's emotions.

[1424] Specific examples

[1425] For example, suppose a user says aloud to a device, "I was so happy my teacher praised me at school today!" The device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a text-to-speech engine and sent to the device for playback.

[1426] In this way, the system, combined with the emotion engine, enables more natural and friendly conversations based on the user's emotions, making interactions with characters that understand and respond to emotions even more appealing, especially for children.

[1427] The processing flow will be explained below.

[1428] Step 1:

[1429] The user opens the dedicated app on their device and selects the animal character they want to talk to. When the user presses the "Start Call" button, the device sends a call request including the character selection information to the server.

[1430] Step 2:

[1431] The server analyzes the call request received from the device and extracts the selected character information. Based on the character information obtained from the analysis, the initial conversation scenario data is provided to the generation AI.

[1432] Step 3:

[1433] The generation AI generates conversational text based on the initial scenario data. For example, it generates text such as "Hello, this is XX. What would you like to talk about today?"

[1434] Step 4:

[1435] The server converts the conversation text received from the AI ​​into voice data using a text-to-speech (TTS) engine. The converted voice data is generated in the character's distinctive voice.

[1436] Step 5:

[1437] The server sends the converted audio file to the device, which then plays the received audio file, allowing the user to hear the character's voice.

[1438] Step 6:

[1439] The user listens to what the character says and responds, for example, by asking, "What did you do today?" The device records the user's response and converts it into text data using a speech recognition engine.

[1440] Step 7:

[1441] The device retransmits the text data to the server, and simultaneously transmits the recorded data to the server, where the emotion engine extracts and analyzes emotional information from the user's speech.

[1442] Step 8:

[1443] The server provides the next conversation scenario data to the generation AI based on the received text data and emotional information. The generation AI generates the optimal response text based on the user's emotions analyzed by the emotion engine.

[1444] Step 9:

[1445] The generative AI generates new conversational text based on the user's responses and emotional information. For example, it might create text that says, "I was walking in the park today. I could hear lots of birds singing and it felt really nice."

[1446] Step 10:

[1447] The server converts the generated new conversation text into voice data using a text-to-speech (TTS) engine again and sends it to the terminal.

[1448] Step 11:

[1449] The device then plays the received audio file back to the user, and this process is repeated to continue the natural conversation between the user and the character, while providing responses that correspond to the user's emotions.

[1450] Step 12:

[1451] As a concrete example, if a user says to a device, "I was so happy when my teacher praised me at school today!", the device records this voice and converts it into text using a speech recognition engine. The server then passes this text data and the emotional information of "happy" analyzed by the emotion engine to the generation AI, which then generates a response such as, "That's wonderful! What did your teacher praise you for?" This response is converted into audio and played on the device.

[1452] Example 2

[1453] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1454] Conventional conversational systems have the problem that the conversation between the user and the character is formulaic and lacks responses that reflect the user's emotions. In particular, there is a need for a system that can recognize the user's emotions and generate natural conversations accordingly.

[1455] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1456] In this invention, the server includes a means for the user to respond to the terminal, converting the response into text and retransmitting it to the server, a means for the server to pass the next scenario data to the generation AI based on the retransmitted text, a means for the terminal to record the user's response and convert it into text data using a voice recognition engine, and a means for the server to extract and analyze the user's emotional information using an emotion analysis engine, thereby enabling a response that corresponds to the user's emotions.

[1457] A "terminal" is an electronic device such as a computer or smartphone that is operated by a user.

[1458] A "server" is a computer system that receives requests and data from terminals, processes them, and gives instructions to the generating AI.

[1459] "Generative AI" is an artificial intelligence model that generates conversational text based on scenario data.

[1460] "Character selection information" is information about the animal character with which the user wants to converse.

[1461] A "call request" is request data that a user sends from a terminal to a server in order to start a call with a character.

[1462] "Scenario data" is the initial set of information that the generation AI uses to generate conversational text.

[1463] "Conversational text" is a sentence generated by the generation AI based on scenario data.

[1464] "Audio data" refers to an audio file converted from text by a text-to-speech engine.

[1465] An "audio file" is a digital file containing audio data that is sent from a server to a terminal.

[1466] A "voice recognition engine" is software that records a user's speech and converts it into text data.

[1467] An "emotion analysis engine" is software that extracts and analyzes emotional information from a user's speech or text.

[1468] A "text-to-speech engine" is software that converts text data into voice data.

[1469] This invention is a system that allows users to use their terminals to have real-time telephone conversations with famous animal characters. Furthermore, it can recognize the user's emotions and adjust the content of the conversation. The detailed implementation method of this system is described below.

[1470] Hardware and software used

[1471] Device: A device operated by a user, such as a smartphone or tablet.

[1472] Server: A computer system that processes user requests and interacts with generative AI models.

[1473] Generative AI model: An artificial intelligence model for generating conversational text.

[1474] Speech recognition engine: Software that records what a user says and converts it into text data (e.g., Google Speech-to-Text).

[1475] Sentiment analysis engine: Software that extracts and analyzes users' emotional information (e.g., IBM Watson Tone Analyzer).

[1476] Text-to-speech engine: Software that converts text data into speech data (e.g., Amazon Polly).

[1477] Specific operation of the system

[1478] 1. Select a character and send a call request:

[1479] Users use their device to open a dedicated app and select the animal character they want to talk to.

[1480] When the user presses the "start call" button, a call request including character selection information is sent from the terminal to the server.

[1481] 2. Generating the first conversation scenario:

[1482] The server analyzes the call request received from the terminal and extracts the selected character information.

[1483] The server sends the initial conversation scenario data to the generation AI and instructs it to generate the initial greeting.

[1484] 3. Conversational text generation and speech conversion:

[1485] The generation AI generates conversational text based on the initial scenario data.

[1486] The server receives the generated text and converts it into audio data using a text-to-speech engine.

[1487] 4. Send and play audio files:

[1488] The server sends the converted audio file to the terminal.

[1489] The terminal receives the audio file and plays it for the user.

[1490] 5. User response processing and sentiment analysis:

[1491] The user listens to what the character says and responds, for example, by asking, "What did you do today?"

[1492] The device records the user's response and converts it into text data using a voice recognition engine.

[1493] The device sends the converted text data to the server, and the emotion analysis engine extracts and analyzes the user's emotional information.

[1494] 6. Emotion-based conversation generation:

[1495] The server receives the text data and emotional information sent by the user, and based on that, sends the next conversation scenario data to the generation AI.

[1496] The AI ​​generates conversational text based on the user's emotions. For example, if the user speaks in a happy tone, the AI ​​generates a positive response.

[1497] 7. Sending and playing audio files:

[1498] The server converts the generated new conversation text into voice data again using the text-to-speech engine.

[1499] The server transmits the converted audio file to the terminal, which receives it and plays it back.

[1500] This process is repeated, allowing for a natural conversation between the user and the character to continue.

[1501] Specific examples

[1502] For example, consider the case where a user says to a device, "I was so happy when my teacher praised me at school today!" The device records this voice and converts it into text using a speech recognition engine. The resulting text is sent to a server, which then uses an emotion analysis engine to determine that the user's emotion is "happy." Based on this determination, the generation AI generates a response such as, "That's wonderful! What did the teacher praise you for?" This response is converted into audio using a text-to-speech engine and sent to and played back on the device.

[1503] Prompt Sentence Examples

[1504] Here are some example prompts you can enter into your generative AI model:

[1505] 1. "If the user says in a cheerful voice, 'I had a great time picnicking in the park today,' generate an appropriate response in the following conversation."

[1506] 2. "If the user sadly says, 'I had a fight with my friend,' generate a comforting response in the following conversation."

[1507] In this way, by combining emotion analysis and generative AI, this system is able to respond according to the user's emotions.

[1508] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1509] Step 1:

[1510] The user opens the app on their device and selects the animal character they want to talk to. When they press the "Start Call" button, a call request including the character selection information is sent from the device to the server. The input here is the user's selection and the call start action, and the output is the call request being sent to the server.

[1511] Step 2:

[1512] The server analyzes the call request received from the device. The request contains character selection information, and extracts this information. The input is the call request sent from the device, and the output is the extracted character selection information. The server generates the initial conversation scenario data based on this information.

[1513] Step 3:

[1514] Based on the extracted character information, the server sends the initial conversation scenario data to the generation AI. An instruction is sent to generate an initial greeting such as "Hello, this is □□. What would you like to talk about today?" The input here is character selection information, and the output is scenario data for the generation AI.

[1515] Step 4:

[1516] The generation AI generates dialogue text based on the initial scenario data. The generated text is adjusted to match the character's personality and speaking style. The input is the initial scenario data, and the output is the generated dialogue text.

[1517] Step 5:

[1518] The server receives the generated text and converts it into audio data using a text-to-speech engine. The converted audio data is generated in the character's voice. Here, the input is the spoken text and the output is audio data.

[1519] Step 6:

[1520] The server sends the converted audio file to the device. The input is audio data, and the output is an audio file sent to the device. The device receives this audio file and plays it for the user. The user can enjoy the conversation while listening to the voices of the animal characters.

[1521] Step 7:

[1522] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input here is the user's response, which the device records. The output is the recorded voice data.

[1523] Step 8:

[1524] The device records the user's response and converts it into text using a speech recognition engine. The input is the voice data and the output is the converted text data, which is then sent to the server.

[1525] Step 9:

[1526] The server receives the converted text data and uses a sentiment analysis engine to extract and analyze the user's sentiment information, where the input is the converted text data and the output is the extracted sentiment information.

[1527] Step 10:

[1528] The server receives the text data and emotional information sent by the user and sends the next conversation scenario data to the generation AI based on that. The input here is the text data and emotional information, and the output is the next scenario data for the generation AI.

[1529] Step 11:

[1530] The generative AI generates conversational text according to the user's emotions. For example, if the user speaks in a cheerful manner, the generative AI generates a positive response. The input here is the following scenario data, and the output is the generated conversational text.

[1531] Step 12:

[1532] The server converts the newly generated conversation text into audio data again using a text-to-speech engine, where the input is the new conversation text and the output is audio data.

[1533] Step 13:

[1534] The server sends the converted audio file to the device, which then receives and plays it. The input here is audio data, and the output is the audio file sent to the device. This allows for a natural conversation between the user and the character to continue.

[1535] (Application example 2)

[1536] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1537] Conventional character dialogue systems have a formulaic user experience and have difficulty providing appropriate responses based on the user's emotions. As a result, particularly in interactive storytelling, users are less likely to feel a sense of realism or familiarity, and the dialogue tends to be uniform. The present invention aims to solve this problem by providing a more natural, interactive conversation experience that responds to the user's emotions.

[1538] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for the user to select a character and make a call request using a terminal, means for the server to receive a call request including character selection information from the terminal and send initial scenario data to the generation AI, and means for the generation AI to generate conversation text based on the scenario data. This makes it possible to analyze the user's emotional information in real time and generate and provide dynamic conversation content accordingly.

[1539] "Terminal" refers to an electronic device through which a user selects a character and makes a call request.

[1540] "Server" refers to the central processing unit that receives information from the terminal and works with the generative AI and other engines to generate and manage conversation scenarios.

[1541] "Character selection information" refers to information about a specific character selected by a user through a terminal.

[1542] A "call request" refers to request information sent from a terminal to a server in order for a user to start a conversation with a character.

[1543] "Initial scenario data" refers to the scenario information at the initial stage of the conversation that the server sends to the generation AI.

[1544] "Generative AI" refers to an artificial intelligence model that generates conversational text based on user input and scenario data.

[1545] "Conversational text" refers to the text data of the dialogue between the character generated by the generation AI and the user.

[1546] "Means for converting to speech" refers to a text-to-speech (TTS) engine for converting the generated conversational text into audio data.

[1547] "Audio file" refers to audio data sent from the server to the terminal and played as the character's voice.

[1548] The "means for converting a response into text" refers to a voice recognition engine that converts a voice response given by a user to a terminal into text data.

[1549] "Emotion analysis engine" refers to a program and device for analyzing emotional information from the content and voice of a user's responses.

[1550] "Emotion information" refers to data that indicates the type and intensity of emotions extracted from the user's responses.

[1551] This invention is a system that allows users to have real-time conversations with famous characters using a terminal, and adjusts the content of the conversation based on the user's emotions. This system is implemented by combining the following steps and hardware / software:

[1552] System Overview

[1553] The user selects a character and makes a call request through the device. The server receives this request information and generates initial scenario data based on the generation AI. The generation AI creates conversation text based on the generated scenario data, and the server converts this text into audio data using a text-to-speech (TTS) engine. This audio data is sent to the device and played back to the user.

[1554] Main processing

[1555] 1. Select a character and send a call request

[1556] The user opens the dedicated app on their device and selects the character they want to talk to. After selecting, the user presses the "Start Call" button, and a call request including the character selection information is sent to the server.

[1557] 2. Receiving a call request and generating a first conversation scenario

[1558] The server receives the call request and sends the initial scenario data based on the character information to the generation AI. The generation AI then generates an initial greeting such as, "Hello, this is □□. What would you like to talk about today?"

[1559] 3. Conversational text generation and speech conversion

[1560] The AI ​​generates dialogue text based on the initial scenario data. The generated text is tailored to the character's personality and speaking style. This dialogue text is then converted into voice data using a TTS engine.

[1561] 4. Sending and playing audio files

[1562] The server then sends the converted audio file to the device, which then plays it back, allowing the user to enjoy the conversation while listening to the character's voice.

[1563] Processing user responses and sentiment analysis

[1564] The user listens to what the character says and responds. The device records the user's response and converts it into text data using a voice recognition engine. The converted text data is then retransmitted to the server, which then uses an emotion analysis engine to extract and analyze the user's emotional information.

[1565] The process of emotion-based conversation generation

[1566] The server receives the user's text data and analyzed emotional information, and based on that, sends the next conversation scenario data to the generation AI, which then generates conversation text according to the user's emotions.

[1567] Hardware and software used

[1568] Hardware: Smartphones, servers

[1569] Software: Dedicated apps, generative AI models (e.g., GPT-3), text-to-speech (TTS) engines, speech recognition engines, sentiment analysis engines

[1570] Specific examples

[1571] For example, if a user says to a device, "I was so happy my teacher praised me at school today!", the device records the voice and converts it into text using a speech recognition engine. The text data recognized as "I was so happy my teacher praised me at school today!" is sent to the server, which then uses an emotion analysis engine to determine the user's emotion as "happy." Based on this determination, the generation AI generates a response such as "That's wonderful! What did the teacher praise you for?" This response is converted into audio by a TTS engine and sent to the device for playback.

[1572] Prompt Sentence Examples

[1573] User: I was so happy when my teacher praised me at school today!

[1574] Prompt for the generated AI:

[1575] Generate a follow-up response in a friendly and cheerful tone for a user who just said, "I was so happy that my teacher praised me at school today!" The response should include a question to keep the conversation going.

[1576] This allows for natural, emotionally-responsive interactive conversation between the user and the character.

[1577] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1578] Step 1:

[1579] The user opens the dedicated app on their device. Within the app, the user selects the character they want to talk to and presses the "Start Call" button. This generates a call request including character selection information and sends it to the server. In this process, the input is the character information selected by the user, and the output is the call request.

[1580] Step 2:

[1581] The server analyzes the call request received from the device and extracts character selection information. Based on this information, the server sends the initial scenario data to the generation AI. The input is the call request, and the output is the initial scenario data to be passed to the generation AI.

[1582] Step 3:

[1583] The generation AI generates conversational text based on the initial scenario data. For example, the generation AI generates the initial greeting message, "Hello, this is □□. What would you like to talk about today?" In this case, the input is the initial scenario data, and the output is the conversational text.

[1584] Step 4:

[1585] The server receives the generated conversation text and converts it into audio data using a text-to-speech (TTS) engine. The input is the conversation text and the output is an audio file.

[1586] Step 5:

[1587] The server sends the converted audio file to the terminal, which receives the audio file and plays it for the user. The input is the audio file and the output is the played audio.

[1588] Step 6:

[1589] The user listens to what the character says and responds. For example, they might ask, "What did you do today?" The input is the user's response to the character's utterance, and the output is the response as voice data.

[1590] Step 7:

[1591] The device records the user's response and converts it into text data using a speech recognition engine. The input is the user's voice response, and the output is the text-converted response data.

[1592] Step 8:

[1593] The recorded text data is retransmitted to the server, where the server uses an emotion analysis engine to extract and analyze emotional information from the user's speech and text. The input is the text-converted response data, and the output is the extracted emotional information.

[1594] Step 9:

[1595] The server receives the text data and analyzed emotional information sent by the user, and then sends the next conversation scenario data to the generation AI based on that. The input is the text data and emotional information, and the output is the next scenario data.

[1596] Step 10:

[1597] The generation AI generates conversational text according to the user's emotions. For example, if the user replies, "I had a great time at the picnic in the park today," the generation AI generates the following conversational content: "That's great! Who were your friends playing with?" The input is the next scenario data, and the output is new conversational text.

[1598] Step 11:

[1599] The server converts the generated new conversation text into audio data using the text-to-speech engine again and sends it to the terminal. The input is the new conversation text, and the output is an audio file as audio data.

[1600] Step 12:

[1601] The device then plays the received audio file back to the user, allowing for a natural conversation between the user and the character, and providing responses according to the user's emotions. The input is the audio file, and the output is the played audio.

[1602] 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 control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1603] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1604] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1605] The emotion identification model 59 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 an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1606] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1607] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1608] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1609] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1610] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs 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 a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1611] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1612] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1613] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1614] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1615] 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.

[1616] It is not necessary to store all 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 all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1617] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1618] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with 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). Also, the hardware resource that executes the specific processing may be a single processor.

[1619] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1620] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1621] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1622] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1623] The following is further disclosed regarding the above embodiment.

[1624] (Claim 1)

[1625] A means for a user to select a character and make a call request using a terminal;

[1626] A means for the server to receive a call request including character selection information from the terminal and transmit initial scenario data to the generation AI;

[1627] A means for the generation AI to generate conversational text based on the scenario data;

[1628] a means for converting the generated conversation text into speech by the server;

[1629] A means for transmitting an audio file from the server to the terminal and for the terminal to play the audio;

[1630] A means for the user to reply to the terminal, convert the reply into text, and retransmit the text to the server;

[1631] The server will then pass the next scenario data to the AI ​​based on the resent text.

[1632] A system including:

[1633] (Claim 2)

[1634] 10. The system of claim 1, wherein the character generates a dialogue responsive to the user's responses.

[1635] (Claim 3)

[1636] The system of claim 1, wherein the generation AI generates conversation content in real time based on the user's responses.

[1637] "Example 1"

[1638] (Claim 1)

[1639] A means for a user to select a character and make a call request using a terminal;

[1640] A means for the terminal to transmit a call request including character selection information to a server;

[1641] a means for the server to receive and analyze the call request;

[1642] A means for the server to send initial scenario data to the generation AI;

[1643] A means for the generation AI to generate conversation text based on the initial scenario data;

[1644] a means for converting the generated conversation text into speech by the server;

[1645] A means for the server to transmit the converted audio file to the terminal and for the terminal to play the audio;

[1646] a means for the user to reply to the terminal, converting the reply into text using a speech recognition engine, and retransmitting the text to the server;

[1647] A means for the server to send the next scenario data to the generation AI based on the resent text;

[1648] A system including:

[1649] (Claim 2)

[1650] The system of claim 1, wherein the character generates a conversation in response to the user's responses.

[1651] (Claim 3)

[1652] The system of claim 1, wherein the generation AI generates conversation content in real time based on the user's responses.

[1653] "Application Example 1"

[1654] (Claim 1)

[1655] A means for a user to select a character and make a call request using an information processing device;

[1656] A server receives a call request including character selection information from an information processing device and transmits first conversation scenario data to a generation AI model;

[1657] A means for the generative AI model to generate conversational text based on the scenario data;

[1658] A means for converting the generated conversation text into voice data by the server;

[1659] means for transmitting audio data from the server to the information processing device and for the information processing device to play back the audio;

[1660] a means for receiving a reply from the user via the information processing device, converting the reply into text data, and retransmitting the text data to the server;

[1661] A means for the server to pass the next conversation scenario data to the generation AI model based on the retransmitted text data;

[1662] A means for converting the conversational text of the generative AI model into speech using a text-to-speech engine and providing the conversational data in real time;

[1663] A system including:

[1664] (Claim 2)

[1665] 10. The system of claim 1, wherein the character generates a dialogue responsive to the user's responses.

[1666] (Claim 3)

[1667] 10. The system of claim 1, wherein the generative AI model generates conversational content in real time based on user responses.

[1668] "Example 2: Combining Emotion Engines"

[1669] (Claim 1)

[1670] A means for a user to select a character and make a call request using a terminal;

[1671] a means for the server to receive a call request including character selection information from the terminal and generate initial scenario data;

[1672] A means for the generation AI to generate conversational text based on the scenario data;

[1673] a means for converting the generated conversation text into speech by the server;

[1674] A means for transmitting an audio file from the server to the terminal and for the terminal to play the audio;

[1675] A means for the user to reply to the terminal, convert the reply into text, and retransmit the text to the server;

[1676] The server will then pass the next scenario data to the AI ​​based on the resent text.

[1677] a means for the terminal to record the user's response and convert it into text data using a voice recognition engine;

[1678] A server extracts and analyzes user's emotion information using an emotion analysis engine;

[1679] A means for the generation AI to generate conversational text according to the user's emotions;

[1680] A means for converting the generated conversation text into voice data using a text-to-speech conversion engine by the server and transmitting the voice data to the terminal;

[1681] A system including:

[1682] (Claim 2)

[1683] 10. The system of claim 1, wherein the character generates dialogue according to the user's emotions.

[1684] (Claim 3)

[1685] The system of claim 1, wherein the generation AI generates conversation content in real time based on the user's emotional information.

[1686] "Application example 2 when combining emotion engines"

[1687] (Claim 1)

[1688] A means for a user to select a character and make a call request using a terminal;

[1689] A means for the server to receive a call request including character selection information from the terminal and transmit initial scenario data to the generation AI;

[1690] A means for the generation AI to generate conversational text based on the scenario data;

[1691] a means for converting the generated conversation text into speech by the server;

[1692] A means for transmitting an audio file from the server to the terminal and for the terminal to play the audio;

[1693] A means for the user to reply to the terminal, convert the reply into text, and retransmit the text to the server;

[1694] The server will then pass the next scenario data to the AI ​​based on the resent text.

[1695] means including an emotion analysis engine for analyzing the user's emotion information and adjusting the conversation content based on the analysis;

[1696] A system including:

[1697] (Claim 2)

[1698] 10. The system of claim 1, wherein the character generates dialogue according to the user's responses and emotional information.

[1699] (Claim 3)

[1700] The system of claim 1, wherein the generation AI generates conversation content in real time based on user responses and emotional information. [Explanation of symbols]

[1701] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for a user to select a character and make a call request using a terminal; A means for the server to receive a call request including character selection information from the terminal and transmit initial scenario data to the generation AI; A means for the generation AI to generate conversational text based on the scenario data; a means for converting the generated conversation text into speech by the server; A means for transmitting an audio file from the server to the terminal and for the terminal to play the audio; A means for the user to reply to the terminal, convert the reply into text, and retransmit the text to the server; The server will then pass the next scenario data to the AI ​​based on the resent text. A system including:

2. The system of claim 1 , wherein the character generates a dialogue in response to the user's responses.

3. The system of claim 1, wherein the generation AI generates conversation content in real time based on the user's responses.

Images