System

Abstract

A system is provided.SOLUTION: A system comprising: means for inputting user information; means for storing user information; means for a user to initiate a conversation; means for generating a response in response to a user utterance by a generative artificial intelligence; means for converting the generated response to speech in a selected voice; means for playing the speech to the user; means for storing and analyzing a conversation log; means for training the generative artificial intelligence; means for receiving and analyzing feedback from the user; and means for improving the generative artificial intelligence based on the feedback.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] In modern society, many people suffer from stress at work and in their daily lives, but there is a lack of communication methods that can be used to effectively relieve stress while maintaining privacy. Existing interactive applications have problems such as being difficult to use due to the risk of revealing real names, and it is difficult to provide personalized conversations tailored to each individual user. Therefore, there is a need to provide a reliable, low-cost, and flexible communication method that allows users to relieve stress using a voice that they prefer. [Means for solving the problem]

[0005] The present invention solves the above problems by the following means. First, it provides a means for inputting user information and a means for saving it. Next, it provides a means for the user to start a conversation and a means for generating a response based on the user's words using a generative artificial intelligence. It further includes a means for converting the generated response into speech in a selected voice and a means for playing that speech to the user. It also provides a means for saving and analyzing a conversation log and a means for training the generative artificial intelligence. Finally, it provides a means for receiving and analyzing feedback from the user and a means for improving the generative artificial intelligence based on that feedback. These means provide a system that allows users to relieve stress through personal conversations while maintaining their privacy.

[0006] "User Information" refers to information about you personally, such as your name, email address, and preferred voice type.

[0007] "Generative AI" refers to an AI model that generates appropriate responses based on user input.

[0008] "Conversation log" refers to data that records the content of a conversation session with a user.

[0009] "Speech synthesis engine" refers to a system that converts text data into speech based on a specified voice profile.

[0010] "Feedback" refers to opinions such as satisfaction and areas for improvement that users enter after the conversation ends.

[0011] "Response" refers to a response message generated by a generative artificial intelligence in response to a user's input.

[0012] "Voice profile" refers to a voice setting with specific vocal characteristics selected by a user.

[0013] "Training data" refers to a dataset used to train generative artificial intelligence. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

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

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

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

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form 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.

[0029] 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.

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

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] System Overview

[0036] The present invention is a system that inputs and stores user information and uses artificial intelligence to provide virtual conversations with users. The system plays back responses based on the voice profile selected by the user, realizing personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[0037] User Registration

[0038] Step 1. Enter and save your user information

[0039] The user launches the application and first enters information such as their name, email address, and preferred voice type.

[0040] The terminal receives the input user information and transmits it to the server.

[0041] The server stores the received user information in a database and completes the initial setup.

[0042] Dialogue generation and voice selection

[0043] Step 2. Initiating a conversation and generating a response

[0044] The user presses the start conversation button in the application to begin the conversation.

[0045] The terminal converts the user's voice input into text using a voice recognition engine and transmits the text data to the server in real time.

[0046] The server passes the received text data to a generative artificial intelligence, which generates an appropriate response along with the user's past conversation history.

[0047] Step 3. Synthesize and play the response

[0048] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[0049] The terminal receives the audio file from the server and plays it for the user.

[0050] Iterative learning and feedback

[0051] Step 4. Saving and analyzing conversation logs

[0052] The server stores the contents of each conversation session as a log and analyzes it periodically.

[0053] The generative AI uses the newly saved log data as training data to continue learning about the user's language usage and stress points.

[0054] Step 5. Receive and analyze feedback

[0055] After the conversation, the user enters their satisfaction level and feedback.

[0056] The terminal transmits the feedback information to the server.

[0057] The server analyzes the received feedback information and reflects it in the training data of the generative artificial intelligence, thereby improving the quality of the next conversation.

[0058] Specific examples

[0059] Example 1: Stress-relieving conversations

[0060] The user inputs "I'm very tired today."

[0061] The server responds to the input by generating the message, "You've worked hard today. Is there anything you'd like to say?" and then uses a speech synthesis engine to convert it into the "gentle female voice" selected by the user.

[0062] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[0063] Example 2: Everyday conversation

[0064] The user types, "What should I have for dinner tonight?"

[0065] The server responds to the input by generating "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[0066] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[0067] As described above, this system effectively reduces user stress and provides a highly satisfying communication experience.

[0068] The processing flow will be explained below.

[0069] Step 1:

[0070] The user launches the application and is presented with a welcome screen where they enter initial information such as their name, email address, and preferred voice type.

[0071] Step 2:

[0072] The device receives the entered user information and sends it to the server, including information such as name, email address, and voice type.

[0073] Step 3:

[0074] The server stores the received user information in a database, which is used to tailor responses to the user's preferences in subsequent conversation sessions.

[0075] Step 4:

[0076] The user is ready to start the conversation by pressing the Start Conversation button, which starts the conversation session.

[0077] Step 5:

[0078] The device passes the user's voice input to a speech recognition engine, which converts it into text. For example, a speech phrase like "I'm tired today" is generated as text data.

[0079] Step 6:

[0080] The terminal transmits the converted text data to the server in real time.

[0081] Step 7:

[0082] The server inputs text data into a generative artificial intelligence (GPT model), which generates an appropriate response based on that data, such as, "You seem tired today. Is there anything you'd like to talk about?"

[0083] Step 8:

[0084] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[0085] Step 9:

[0086] The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[0087] Step 10:

[0088] The server stores the content of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[0089] Step 11:

[0090] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized to the user's preferences.

[0091] Step 12:

[0092] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[0093] Step 13:

[0094] The terminal sends feedback information to the server.

[0095] Step 14:

[0096] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[0097] The above is the specific processing flow of this system. Through these detailed steps, users can effectively relieve stress through individually personalized dialogue.

[0098] Example 1

[0099] 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."

[0100] Conventional user information input and dialogue systems lack the ability to accurately convert user voice input into text, generate appropriate responses, and play back the generated responses in a voice that suits the user's preferences. Furthermore, log management and feedback to analyze conversation history and improve the performance of generative AI are not fully utilized. This makes it difficult to improve user satisfaction.

[0101] 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.

[0102] In this invention, the server includes a means for storing user information, a means for generating a response based on the user's words using a generative artificial intelligence, and a means for converting the generated response into speech using a selected voice. This allows the server to convert the user's voice input into text, have the generative artificial intelligence generate an appropriate response based on the text, and then play back the response as speech based on a voice profile selected by the user. Furthermore, by including a means for storing and analyzing a conversation log and a means for improving the generative artificial intelligence based on feedback, the server can reflect the user's past conversation history and feedback, enabling more natural and effective dialogue.

[0103] "Means for inputting user information" refers to an interface that allows a user to input their own information into the system.

[0104] The "means for saving user information" is a function for temporarily or permanently storing input user information in a storage such as a database.

[0105] The "means for a user to start a conversation" refers to an operating means or interface that allows a user to start a conversation with the system.

[0106] "Means for generating responses based on user words using generative artificial intelligence" refers to algorithms or software for generating appropriate responses based on input data obtained from the user.

[0107] The "means for converting the generated response into speech in a selected voice" is a technique for converting the generated textual response into speech data based on a voice profile selected by the user.

[0108] The "means for reproducing audio to the user" refers to a device or function for allowing the user to listen to the generated audio data.

[0109] "Means for saving and analyzing conversation logs" refers to a mechanism for recording conversation history data and analyzing it to improve the system and understand user trends.

[0110] "Means for training a generative artificial intelligence" refers to a learning process that uses stored data and feedback to improve the accuracy and performance of a generative artificial intelligence.

[0111] "Means for receiving and analyzing feedback from users" is a function for receiving and analyzing opinions and satisfaction levels provided by users after using the system.

[0112] "Means for improving generative artificial intelligence based on feedback" refers to means for optimizing the behavior and responses of generative artificial intelligence by utilizing received feedback information.

[0113] The "means for converting speech input to text" is software or hardware for analyzing speech input from a user and converting it into text format data.

[0114] "Means for transmitting converted text data to the generative artificial intelligence" refers to a communication function that passes the text data converted from speech to the generative artificial intelligence to generate a response.

[0115] The "means for passing the generated response text to the speech synthesis engine" is a process for sending the text response generated by the generative artificial intelligence to the speech synthesis engine.

[0116] The "means for transmitting a synthesized voice file from a server to a terminal" is a communication technology for transferring a voice file generated on a server to a terminal and allowing the user to play it.

[0117] The present invention is a system that inputs and stores user information and uses generative artificial intelligence to provide virtual conversations with the user. The system converts the user's voice input into text, and the generative artificial intelligence generates appropriate responses based on that text. The system then converts the generated responses into speech using a voice profile selected by the user and plays them back, enabling personalized communication. The system also stores and analyzes conversation logs, receives and analyzes feedback, and incorporates it into the generative artificial intelligence's training data to continuously improve the system's performance.

[0118] The hardware and software required to implement this system are as follows:

[0119] Device: The smartphone or personal computer used by the user.

[0120] Server: Cloud server (e.g. AWS (registered trademark), Google (registered trademark) Cloud).

[0121] Database: A relational database such as MySQL (registered trademark).

[0122] Speech recognition engine: Google Speech-to-Text, etc.

[0123] Generative artificial intelligence: OpenAI (registered trademark), GPT-3 (registered trademark), etc.

[0124] Speech synthesis engine: Google Text-to-Speech, Amazon Polly, etc.

[0125] Entering and saving user information

[0126] The user starts the application and enters user information, such as name, email address, and preferred voice type, into the registration form and presses the submit button. The device then sends this information to the server, which then stores it in a database.

[0127] Conversation generation and response

[0128] The user presses a button to start a conversation within the application. The device converts the user's voice input into text using a speech recognition engine and sends the generated text data to the server. The server then passes this text data to a generative artificial intelligence (AI) system, which generates an appropriate response along with the user's past conversation history. The generated response text is then passed to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The device then receives this audio file and plays it back to the user.

[0129] Iterative learning and feedback

[0130] The server saves the content of each conversation session as a log and periodically analyzes it. The generative AI then uses the newly saved log data as training data to continue learning about the user's vocabulary and stress points. After the conversation ends, the user enters their satisfaction level and feedback. The device sends this feedback information to the server, which analyzes the received information and reflects it in the generative AI's training data.

[0131] Specific examples

[0132] Example 1: Stress-relieving conversations

[0133] The user speaks, "I'm very tired today."

[0134] The terminal converts this speech into text and sends it to the server.

[0135] The server uses an AI model to generate a response such as, "You've worked hard today. Is there anything you'd like to say?"

[0136] The server passes this text to a speech synthesis engine set to a "gentle female voice" and generates an audio file.

[0137] The terminal plays this sound to the user.

[0138] Example 2: Everyday conversation

[0139] The user speaks, "What should I have for dinner tonight?"

[0140] The terminal converts this speech into text and sends it to the server.

[0141] The server uses an AI model to generate a response like, "How about pasta? It's easy and delicious."

[0142] The server passes this text to a speech synthesis engine set to a "male, low-medium voice" and generates an audio file.

[0143] The terminal plays this sound to the user.

[0144] Prompt Sentence Examples

[0145] Example prompt to generate a response to the input "I'm very tired today":

[0146] User: I'm very tired today.

[0147] AI: You've done well today. Is there anything you'd like to talk about?

[0148] An example prompt to generate a response to the input "What should I have for dinner tonight?" would be:

[0149] User: What should I have for dinner tonight?

[0150] AI: How about pasta? It's easy and delicious.

[0151] As described above, this system converts the user's voice input into text, and then uses generative AI to generate appropriate responses based on that text, and plays back the responses aloud based on the voice profile selected by the user. By managing conversation logs and utilizing feedback, the system's performance can be continuously improved.

[0152] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0153] Step 1: Enter and save user information

[0154] A user launches the application and enters information such as their name, email address, and preferred voice type.

[0155] Input: The type of personal information or voice input the user makes.

[0156] Output: Completed user profile data.

[0157] How it works: The device receives the information the user enters into the input form and sends it as an HTTP POST request to the server, which stores the information in a database and generates profile data.

[0158] Step 2: Start a conversation

[0159] The user presses the "Start Conversation" button in the application to begin the conversation.

[0160] Input: A user action to start a conversation.

[0161] Output: Start flag of conversation session.

[0162] Operation: The terminal detects the user's "start conversation" operation and sends a request to the server to start a conversation session.

[0163] Step 3: Convert voice to text

[0164] The user provides voice input during the conversation.

[0165] Input: Audio data from the user.

[0166] Output: The audio data converted to text.

[0167] How it works: The device captures the user's voice with a microphone and sends it to a speech recognition engine (e.g., Google Speech-to-Text). The speech recognition engine converts the voice data into text, and the device sends the text data to a server.

[0168] Step 4: Generate a response

[0169] The server passes the received text data to a generative artificial intelligence (AI model), which generates an appropriate response along with the user's past conversation history.

[0170] Input: User utterance data in text format.

[0171] Output: The generated response text data.

[0172] How it works: The server inputs text data in the form of a prompt sentence into a generative AI (e.g., OpenAI GPT-3) and obtains the generated response text.

[0173] Step 5: Text-to-speech response

[0174] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[0175] Input: The generated response text and the user's selected voice profile.

[0176] Output: The generated audio file.

[0177] What it does: The server sends the text data to a speech synthesis engine (e.g., Google Text-to-Speech) and generates an audio file using the specified voice profile.

[0178] Step 6: Play the audio file

[0179] The terminal plays the audio file received from the server to the user.

[0180] Input: The audio file sent from the server.

[0181] Output: The audio played to the user.

[0182] How it works: The device downloads the audio file from the server, plays it in an audio player, and allows the user to listen.

[0183] Step 7: Saving and analyzing conversation logs

[0184] The server stores the contents of each conversation session as a log and analyzes it periodically.

[0185] Input: All text data from the conversation session.

[0186] Output: Saved conversation log.

[0187] How it works: The server stores textual conversation data in a "conversation log" table in the database, and then periodically analyzes it using an analysis algorithm.

[0188] Step 8: Receive and analyze feedback

[0189] After the conversation, the user enters their satisfaction level and feedback.

[0190] Input: Feedback information from the user.

[0191] Output: Feedback data stored on the server.

[0192] How it works: The device sends the information entered into the feedback form to the server, which stores it in an analysis database.

[0193] Step 9: Training the generative AI

[0194] The server trains the generative artificial intelligence based on the stored log data and feedback information.

[0195] Input: Conversation logs and feedback data.

[0196] Output: An improved generative artificial intelligence model.

[0197] How it works: The server, through a training system, includes this data in the training set of the generative artificial intelligence, improving the AI's accuracy and quality of responses.

[0198] (Application example 1)

[0199] 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."

[0200] In today's world, there is a demand for improved personal security awareness and for rapid and effective security responses. However, it is difficult for individual users to effectively manage and use complex and diverse security information. In particular, it is often difficult to take appropriate action in emergencies that require a rapid response. For this reason, an interface that allows users to easily obtain security-related support is needed.

[0201] 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.

[0202] In this invention, the server includes means for inputting user information, means for saving user information, means for a user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, means for inputting security questions or problems by voice, means for responding to the user with security-related information, means for providing emergency response instructions, and means for functioning as a security assistant. This allows the user to easily receive security-related assistance and instructions via voice, enabling efficient and effective security responses.

[0203] "Means for inputting user information" refers to devices or software that allow users to input personal data such as their name and voice profile.

[0204] The "means for storing user information" refers to a database or storage device for retaining input user information for a long period of time.

[0205] A "means for a user to initiate a conversation" is an interface such as a button or voice command that a user uses to initiate a conversation with the system.

[0206] "Means for generating a response based on the user's words using generative artificial intelligence" is a system that uses an artificial intelligence algorithm to automatically generate an appropriate response based on the content of the user's statements.

[0207] The "means for converting the generated response into speech in a selected voice" is a speech synthesis engine for converting the response text into a speech file based on a voice profile selected by the user.

[0208] The "means for reproducing audio to the user" is a device for delivering the synthesized audio to the user via a speaker or earphones.

[0209] "Means for storing and analyzing conversation logs" refers to data storage and analysis algorithms for recording the content of conversations between users and the system and subsequently analyzing the data.

[0210] "Means for training generative artificial intelligence" refers to machine learning algorithms and computing resources that improve the performance of generative artificial intelligence based on new data.

[0211] The "means for receiving and analyzing feedback from users" refers to an interface and analysis system for collecting and analyzing feedback data such as satisfaction levels and requests for improvement provided by users.

[0212] "Means for improving generative artificial intelligence based on feedback" refers to a process for using feedback data to improve generative artificial intelligence algorithms and databases.

[0213] The "means for inputting security questions or problems by voice" is a voice recognition device that allows a user to input security questions or problems in a voice format.

[0214] The "means for responding to the user with security-related information" refers to generative artificial intelligence and response generation algorithms that provide appropriate responses to security-related questions or problems entered by the user.

[0215] The "means for providing emergency response instructions" is a system that provides instructions and advice necessary for users to take appropriate action in an emergency.

[0216] "Means to function as a security assistant" refers to services and interfaces that comprehensively provide security-related information and emergency response support.

[0217] MODE FOR CARRYING OUT THE INVENTION

[0218] System Overview

[0219] This invention is a system that inputs and stores user information and uses artificial intelligence to provide the user with virtual security-related conversations. The system plays back responses based on the voice profile selected by the user, enabling personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[0220] User Registration

[0221] Step 1. Enter and save your user information

[0222] A user starts an application and inputs information such as their name and voice type. The terminal receives the input user information and sends it to the server. The server stores the received user information in a database.

[0223] Dialogue generation and voice selection

[0224] Step 2. Initiating a conversation and generating a response

[0225] The user presses the "Start conversation" button in the application to begin the conversation. The device converts the user's voice input into text using a speech recognition engine and sends the text data to the server in real time. The server then passes the received text data and conversation history to a generative AI system, which generates an appropriate response.

[0226] Step 3. Synthesize and play the response

[0227] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The terminal receives the audio file from the server and plays it back to the user.

[0228] Iterative learning and feedback

[0229] Step 4. Saving and analyzing conversation logs

[0230] The server stores the content of each conversation session as a log and periodically analyzes it. The generative AI uses the newly saved log data as training data to improve its performance.

[0231] Step 5. Receive and analyze feedback

[0232] After the conversation ends, the user inputs their satisfaction level and feedback. The device then sends the feedback information to the server. The server analyzes the received feedback information and reflects it in the training data for the generative AI.

[0233] Specific examples of security measures

[0234] Specific examples of security assistant AI

[0235] When a user says, "I think a burglar might have broken in," the server generates a response, "Should I immediately contact the police or call a nearby security guard?" and converts it into the voice selected by the user using a speech synthesis engine. The device then plays back this generated voice and gives instructions to the user.

[0236] Example prompt sentence:

[0237] What is the course of action if a burglar breaks in?

[0238] In this embodiment of the invention, the hardware required is a microphone and speaker, and the software uses the speech_recognition library and the gtts library. The generative AI uses the GPT-2 model and the transformers library. A database and various analysis algorithms are used for data storage and analysis. This allows users to easily receive security-related assistance and instructions via voice.

[0239] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0240] Step 1:

[0241] A user starts an application and inputs user information such as name, voice type, etc. This inputs the user information.

[0242] Step 2:

[0243] The terminal receives the input user information and transmits it to the server, where the input data includes the user name and voice profile.

[0244] Step 3:

[0245] The server stores the received user information in a database, which is then used for conversation generation and voice synthesis.

[0246] Step 4:

[0247] The user presses the start speech button in the application to start the conversation, which starts voice input.

[0248] Step 5:

[0249] The device converts the user's voice input into text using a speech recognition engine (e.g., Google Speech Recognition API) and sends the text data to the server in real time. The voice data is converted into text data and sent to the server.

[0250] Step 6:

[0251] The server passes the received text data to a generative AI (e.g., GPT-2), which generates an appropriate response along with the user's past conversation history. The prompt sentence is used as input data, and the generative AI model generates text data.

[0252] Step 7:

[0253] The server passes the generated response text to a speech synthesis engine (e.g., GTTS), which generates an audio file based on the voice profile selected by the user. The text data and the voice profile are used as input data to generate the audio file.

[0254] Step 8:

[0255] The terminal receives the audio file from the server and plays it for the user, so that the audio file is played and the user can listen to it.

[0256] Step 9:

[0257] The server stores the content of each conversation session as a log and periodically analyzes it. By storing and analyzing the conversation log, the accuracy of the generative AI model is improved.

[0258] Step 10:

[0259] After the conversation ends, the user inputs their satisfaction level and feedback, which generates feedback data.

[0260] Step 11:

[0261] The terminal sends feedback information to the server. The feedback data is sent to the server.

[0262] Step 12:

[0263] The server analyzes the received feedback information and reflects it in the training data of the generative AI. The feedback data is analyzed and used to improve the generative AI model.

[0264] 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.

[0265] System Overview

[0266] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[0267] User Registration

[0268] Step 1. Enter and save your user information

[0269] A user launches the application and enters information such as their name, email address, and preferred voice type.

[0270] The terminal receives the input user information and transmits it to the server.

[0271] The server stores the received user information in a database and completes the initial setup.

[0272] Dialogue generation and voice selection

[0273] Step 2. Initiating a conversation and recognizing emotions

[0274] The user presses the start conversation button in the application and is ready to start the conversation.

[0275] The device passes the user's voice input to a speech recognition engine and converts it into text. For example, a speech such as "I'm tired today" is generated as text data.

[0276] The terminal transmits the converted text data to the server, and also passes the voice data to an emotion engine for emotion analysis.

[0277] Step 3. Emotion-based response generation

[0278] The server inputs the emotional data and text data obtained from the emotion engine into a generative AI system to generate an optimal response, such as, "You seem tired today. Would you like to know how to relax?"

[0279] Step 4. Synthesize and play the response

[0280] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[0281] The terminal receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[0282] Iterative learning and feedback

[0283] Step 5. Saving and analyzing conversation logs

[0284] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[0285] The emotion engine analyzes the user's emotional tendencies from log data and reflects this in training the generative artificial intelligence.

[0286] Step 6. Receive and analyze feedback

[0287] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[0288] The terminal transmits the feedback information to the server.

[0289] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[0290] Specific examples

[0291] Example 1: Stress-relieving conversations

[0292] The user inputs "I'm very tired today."

[0293] Based on the input and emotional data, the server generates a response such as, "You seem tired today. Shall I think of a way to help you relax?" and converts this into the "gentle female voice" selected by the user using a speech synthesis engine.

[0294] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[0295] Example 2: Everyday conversation

[0296] The user types, "What should I have for dinner tonight?"

[0297] Based on the input and emotional data, the server generates the response, "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[0298] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[0299] As described above, this system can recognize the user's emotions and generate responses based on them, providing a more personalized communication experience and effectively reducing stress.

[0300] The processing flow will be explained below.

[0301] Step 1:

[0302] The user launches the application and is presented with a welcome screen where the user enters initial information such as name, email address, and preferred voice type.

[0303] Step 2:

[0304] The device receives the entered user information and transmits the data to the server, including the user's name, email address, and selected voice profile.

[0305] Step 3:

[0306] The server stores the received user information in a database, which serves as the basis for responding to the user's preferences in subsequent conversation sessions.

[0307] Step 4:

[0308] The user clicks the Start Conversation button and is ready to begin the conversation, which causes the system to start a conversation session.

[0309] Step 5:

[0310] The device passes the user's voice input to a voice recognition engine, which converts the voice into text data. For example, the voice saying "I'm tired today" is converted into text "I'm tired today."

[0311] Step 6:

[0312] The device sends the converted text data to the server in real time, and simultaneously sends the voice data to the emotion engine.

[0313] Step 7:

[0314] The server receives the emotion data obtained from the emotion engine and the text data from the voice recognition engine, and inputs them into the generative AI. The emotion data includes information such as "fatigue."

[0315] Step 8:

[0316] Generative AI generates appropriate responses based on emotional and text data, such as "You seem tired today. Take a break and refresh yourself."

[0317] Step 9:

[0318] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. If the user selects a "soft female voice," the response will be spoken in that voice.

[0319] Step 10:

[0320] The terminal plays the audio file received from the server, allowing the user to experience a conversation with a virtual conversation partner.

[0321] Step 11:

[0322] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and emotional tendencies.

[0323] Step 12:

[0324] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized for the user's preferences and emotions.

[0325] Step 13:

[0326] After the conversation, the user inputs their satisfaction level and feedback, such as "I would like more specific advice."

[0327] Step 14:

[0328] The terminal transmits feedback information from the user to the server.

[0329] Step 15:

[0330] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[0331] The above is the specific processing flow of this system, which combines an emotion engine. Through these detailed steps, users can achieve more effective stress relief through personalized dialogue tailored to their emotions.

[0332] Example 2

[0333] 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."

[0334] In today's world, many people are increasingly experiencing stress and loneliness. In these circumstances, there is a growing need for dialogue systems that can understand emotions and respond appropriately based on those emotions. However, conventional dialogue systems lack emotion recognition capabilities, making it difficult to provide the personalized communication users desire. Furthermore, they lack natural-sounding voice synthesis based on voice profile selection and a feedback system for improving the quality of training data. This leads to issues such as reduced user satisfaction and difficulty in improving dialogue quality.

[0335] 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.

[0336] In this invention, the server includes a means for saving input information, a means for generating a response based on the user's words using generative artificial intelligence, and an emotion recognition means for analyzing emotion data and generating a response. This makes it possible to understand emotions based on the information entered by the user and generate an optimal response. Furthermore, a means for voice synthesis of the generated response based on a voice profile selected by the user is provided, making it possible to provide natural and personalized voice responses. Furthermore, by including a means for saving a conversation log as training data for the generative artificial intelligence and a means for receiving and analyzing feedback from the user, the quality of the dialogue system can be continuously improved.

[0337] 1. "Input information" refers to information such as name, email address, and voice profile that a user enters into the system.

[0338] 2. "Storage means" refers to a method or device for storing input information or conversation logs in a database or memory.

[0339] 3. "Means for starting a conversation" refers to the buttons or interface that a user operates to start a conversation.

[0340] 4. "Generative AI" refers to machine learning models and algorithms that generate optimal responses based on user input and emotional data.

[0341] 5. "Voice Profile" means a voice configuration with specific voice characteristics (e.g., pitch, gender, timbre, etc.) selected by a user.

[0342] 6. "Speech synthesis means" means a technology or device for converting text data into speech based on a specified speech profile.

[0343] 7. "Playback means" refers to an output device such as a speaker or earphones that allows the user to hear the generated audio.

[0344] 8. "Conversation log" means a record of the content of a conversation between a user and a system, including emotional data.

[0345] 9. "Emotion recognition means" means technologies or algorithms for analyzing emotions from a user's voice or text data.

[0346] 10. "Feedback" refers to the satisfaction rating and suggestions for improvement provided by the user to the system after the conversation has ended.

[0347] MODE FOR CARRYING OUT THE INVENTION

[0348] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[0349] Hardware and Software Configuration

[0350] Server: The server stores user information, runs generative AI, analyzes emotion data, generates responses, synthesizes responses, stores and analyzes conversation logs, and analyzes feedback. Specifically, it uses a database management system (e.g., MySQL or MongoDB), generative AI (e.g., OpenAI's GPT-4 (registered trademark)), an emotion engine (e.g., IBM Watson (registered trademark) Tone Analyzer), and a speech synthesis engine (e.g., Amazon Polly).

[0351] Terminal: The terminal inputs user information, receives and converts voice input to text, analyzes emotions in the voice data, plays back replies, and inputs and sends feedback information. Specifically, it uses a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[0352] User: The user enters information or speaks through the application and receives spoken responses provided by the system.

[0353] Specific processing of the system

[0354] 1. Entering and saving user information: The user launches the application and enters information such as their name, email address, preferred voice type, etc. The device sends the entered information to the server, which stores it in a database.

[0355] 2. Starting a conversation and recognizing emotions: The user presses the conversation start button to prepare for the conversation. The device passes the user's voice input to the speech recognition engine, converts it into text, and sends it to the server. At the same time, the voice data is passed to the emotion engine for emotion analysis.

[0356] 3. Emotion-based response generation: The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[0357] For example, a prompt such as "The user says he is tired. Please generate the best response" is sent to the generative artificial intelligence.

[0358] 4. Response voice synthesis and playback: The server passes the generated response text to a voice synthesis engine, which generates an audio file based on the voice profile selected by the user. The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[0359] 5. Saving and analyzing conversation logs, receiving and analyzing feedback: The server saves the content and emotional data of each conversation session as a log and uses it as training data for the generative AI. After the conversation ends, the user enters feedback and sends it to the server via their device. The server analyzes the feedback information and reflects it in the training data for the generative AI to improve the quality of the next conversation.

[0360] Specific examples

[0361] Stress relief conversations:

[0362] The user speaks "I'm very tired today."

[0363] The device converts the speech into text using a voice recognition engine and sends the text data, such as "I'm tired today," to the server.

[0364] The server detects the "tired" emotion using an emotion engine and sends a prompt to the generative AI model.

[0365] The server generates a text response such as "You seem tired today. Want to think of a way to relax?"

[0366] The server uses a voice synthesis engine to convert the voice into a gentle female voice.

[0367] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[0368] Through the above process, the system can recognize the user's emotions and generate optimal responses based on those emotions, thereby providing a personalized communication experience and increasing user satisfaction.

[0369] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0370] Step 1: Enter and save user information

[0371] A user launches the application and enters information such as their name, email address, and preferred voice type.

[0372] Input: User information such as name, email address, and voice profile.

[0373] The terminal receives the entered user information and transmits this information to the server.

[0374] Output: The user information sent to the server.

[0375] The server stores the received user information in a database, typically using a database management system such as MySQL or MongoDB.

[0376] Output: User information stored in the database.

[0377] Step 2: Initiating a conversation and recognizing emotions

[0378] The user presses the start conversation button in the application and is ready to start the conversation.

[0379] The terminal receives the user's voice input and converts it into text using a speech recognition engine.

[0380] Input: User's voice input (e.g. "I'm tired today").

[0381] Output: The audio data that is passed to the speech recognition engine.

[0382] The device sends the generated text data to the server and simultaneously passes the voice data to the emotion engine for emotion analysis. For example, the device generates text data such as "I'm tired today" using the Google Cloud Speech-to-Text API.

[0383] Input: Text data and audio data.

[0384] Output: The audio data passed to the emotion engine.

[0385] The server analyzes emotions from the received voice data using an emotion engine, for example, IBM Watson Tone Analyzer.

[0386] Output: Parsed emotion data.

[0387] Step 3: Emotion-based response generation

[0388] The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[0389] Input: Emotion data and text data.

[0390] Output: The prompt input to the generative AI model (e.g., "The user says they are tired. Please generate the best response.").

[0391] The server receives responses from a generative AI model, such as OpenAI's GPT-4, to generate responses like, "You seem tired today. Want to know how to relax?"

[0392] Output: The generated response text.

[0393] Step 4: Synthesize and play the response

[0394] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[0395] Input: Response text and voice profile.

[0396] Output: The data that is passed to the speech synthesis engine.

[0397] The server sends the generated audio file to the terminal using a speech synthesis engine (e.g., Amazon Polly).

[0398] Output: The generated audio file.

[0399] The terminal receives the generated audio file and plays it for the user.

[0400] Output: The audio that is played to the user.

[0401] Step 5: Save and analyze conversation logs, receive and analyze feedback

[0402] The server stores the content and emotional data of each conversation session as a log, for example, in MongoDB.

[0403] Input: Conversational content and emotion data.

[0404] Output: Saved conversation logs.

[0405] The server uses the stored log data to train the generative artificial intelligence.

[0406] Input: Saved conversation logs.

[0407] Output: Conversation logs used as training data.

[0408] After the conversation, the user enters their satisfaction level and feedback.

[0409] Input: Satisfaction rating and suggestions for improvement.

[0410] The terminal transmits the feedback information input by the user to the server.

[0411] Output: Feedback information sent to the server.

[0412] The server analyzes the received feedback information and reflects it in the training data for the generative artificial intelligence.

[0413] Input: Feedback information.

[0414] Output: An improved generative artificial intelligence model.

[0415] Through these processing steps, the system can recognize the user's emotions and generate optimal responses, providing a personalized communication experience. It can also utilize feedback to improve the system's performance and continuously increase user satisfaction.

[0416] (Application example 2)

[0417] 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."

[0418] Ensuring personal safety and security is becoming increasingly important in modern society. However, traditional security measures are generally implemented by administrators or experts, and lack personalized, interactive security measures that reflect individual needs and feelings. Therefore, a system is needed that allows users to feel reassured in specific situations and receive appropriate advice on implementing specific security measures.

[0419] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0420] In this invention, the server includes means for inputting user information, means for saving the user information, means for the user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, a visual display device worn by the user, a speech recognition device for acquiring the user's speech, means for converting the acquired speech into text data, means for analyzing the text data using an emotion recognition engine, means for generating an optimal response based on the analyzed emotion data, and means for providing audio security information to the user. This allows the user to receive a sense of security and appropriate advice for implementing specific security measures through a dialogue that reflects their individual emotional state and security concerns in real time.

[0421] "User Information" refers to data about a user, such as their name, email address, and preferred voice type.

[0422] "Means for inputting user information" refers to a function that allows a user to input their own information through a smart device.

[0423] "Means for saving user information" refers to a function for storing input user data in a storage device such as a database.

[0424] "Means for a user to start a conversation" refers to an interface that allows a user to start a conversation with the system using voice or gestures.

[0425] "Generative AI" refers to an AI system that generates optimal responses based on user input data.

[0426] "Means for converting the generated response to speech in a selected voice" refers to functionality for converting the generated text response to speech using a voice profile selected by the user.

[0427] "Means for playing audio to the user" refers to a function for playing the generated audio through a speaker or the like so that the user can hear it.

[0428] "Means for saving and analyzing conversation logs" refers to a function for recording the content of conversations with users and analyzing that data.

[0429] "Means for training generative AI" refers to a learning function that uses collected data to improve the performance of generative AI.

[0430] "Means for receiving and analyzing feedback from users" refers to a function for collecting opinions and evaluations provided by users to the system and analyzing them.

[0431] "Means for improving generative AI based on feedback" refers to the function of improving the algorithms and response quality of generative AI based on analyzed feedback data.

[0432] "A visual display device worn by a user" refers to a display device such as smart glasses or a head-mounted display used by a user.

[0433] The "voice recognition device for acquiring the user's voice" refers to a microphone or voice input device for picking up the user's voice.

[0434] The "means for converting acquired voice into text data" refers to a function for converting voice data into text data using a voice conversion engine.

[0435] "Means for analyzing text data based on an emotion recognition engine" refers to a function that evaluates and analyzes the emotional state of converted text data using an emotion analysis engine.

[0436] "Means for generating an optimal response based on analyzed emotion data" refers to a function that generates an optimal response taking into account the results of emotion recognition.

[0437] "Means for providing audio security information to the user" refers to a function for providing the generated response or security information to the user as audio.

[0438] The specific system for realizing this invention is an interactive security assistant system designed to provide users with safety and security. This system uses the following hardware and software to handle everything from inputting user information to generating responses and playing voice.

[0439] Hardware and software used

[0440] 1. Hardware

[0441] Smart glasses: When worn by the user, they function as a display and provide audio output and input.

[0442] Microphone: Built into the smart glasses and used to capture the user's voice.

[0443] Speaker: Built into the smart glasses, used to play generated responses to the user.

[0444] 2. Software

[0445] Speech recognition engine: For example, using the Google Speech-to-Text API, converts the user's voice into text data.

[0446] Emotion recognition engine: For example, using Azure® Cognitive Services Emotion API, analyzes emotions based on text data.

[0447] Generative AI: For example, OpenAI GPT-4 is used to generate optimal responses based on user input and sentiment analysis results.

[0448] A speech synthesis engine: For example, using Amazon Polly to convert the generated responses into speech.

[0449] Processing Details

[0450] 1. Enter and save user information

[0451] The user puts on the smart glasses, launches the application, and enters their name, email address, preferred voice type, etc. This information is sent to the server through the smart glasses interface and stored in a database.

[0452] 2. Conversation initiation and emotion recognition

[0453] The user initiates a conversation using voice or touch gestures. The smart glasses' microphone picks up the user's voice and converts it into text using the Google Speech-to-Text API. This text is then sent to a server where an emotion recognition engine (such as the Azure Cognitive Services Emotion API) performs emotion analysis.

[0454] 3. Generating and playing optimal responses

[0455] Based on the emotional data and text data obtained from the emotion recognition engine, a generative artificial intelligence (e.g., OpenAI GPT-4) generates the optimal response. The generated response is passed to a speech synthesis engine (e.g., Amazon Polly), which converts it into an audio file using the voice profile selected by the user. The generated audio is played back from the smartglasses' speakers, allowing the user to experience the dialogue.

[0456] 4. Saving and analyzing conversation logs

[0457] The conversation content and emotional data are stored on the server and used as a dataset to learn the user's tendencies, which improves the quality of the next conversation. Users also provide feedback after the conversation, which the server analyzes and reflects in training the generative AI.

[0458] Specific prompt examples

[0459] For example, if a user says, "I'm at home now, but I'm a little worried about my safety," the system will generate the following optimal response:

[0460] "If a user voice-inputs a security-related question like, 'I'm at home now, but I'm a little worried about my safety,' how would you provide a sense of security? Considering the emotional data, what response should your generative AI model give? Think of five specific example responses."

[0461] In this way, it is possible to provide users with a sense of security and appropriate advice for implementing specific security measures through dialogue that reflects their individual emotional state and security concerns in real time.

[0462] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0463] Step 1:

[0464] The user puts on the smart glasses and launches the application.

[0465] Specific operation: The application is launched in the smart glasses, and first an interface is displayed for the user to enter user information such as name, email address, preferred voice type, etc. The user manually enters this information and presses the send button.

[0466] Input: User information such as name, email address, and preferred voice type.

[0467] Output: The entered user information is sent to the server and stored in the database.

[0468] Step 2:

[0469] The user presses the start conversation button to prepare for the conversation. The microphone in the smart glasses picks up the user's voice and passes it to the speech recognition engine.

[0470] What it does: A user says, "I'm at home now, but I'm a little worried about my safety." The microphone captures this speech, and the Google Speech-to-Text API converts it into text data, which is then sent to the server.

[0471] Input: Speech input from the user.

[0472] Output: Text data is generated by the Google Speech-to-Text API and sent to the server.

[0473] Step 3:

[0474] The server passes the generated text data to an emotion recognition engine to analyze the emotion.

[0475] How it works: The server passes text data to the Azure Cognitive Services Emotion API and requests emotion recognition. The emotion recognition engine analyzes the user's emotions based on the text data and returns the results to the server.

[0476] Input: Text data converted from audio.

[0477] Output: Emotion data as emotion recognition results.

[0478] Step 4:

[0479] The server uses generative artificial intelligence to generate the optimal response based on emotional data and text data.

[0480] How it works: The server instructs OpenAI GPT-4 to generate the optimal response based on emotional and text data. The generative AI analyzes the data and generates an appropriate response text. For example, "Please double-check the area around your house. Are the front door and windows securely closed?"

[0481] Input: Emotion data and text data.

[0482] Output: The generated response text data.

[0483] Step 5:

[0484] The server passes the generated response text to a speech synthesis engine, which generates an audio file using the voice profile selected by the user.

[0485] What happens: The server asks Amazon Polly to convert the response text data into an audio file. Amazon Polly generates the audio file using the voice profile selected by the user and returns the file to the server.

[0486] Input: Generated response text data.

[0487] Output: The generated audio file based on the voice profile selected by the user.

[0488] Step 6:

[0489] The server sends the generated audio file to the smart glasses for playback.

[0490] What happens: The server sends an audio file to the smart glasses, and the smart glasses' speaker plays the audio. The user hears a message saying, "Please double-check your surroundings. Are the front door and windows securely closed?"

[0491] Input: The generated audio file.

[0492] Output: The sound is played and heard by the user.

[0493] Step 7:

[0494] The server stores and analyzes conversation logs, receives user feedback, and uses it to improve the generative AI.

[0495] How it works: The server records the conversation and emotional data and stores it in a database. After the conversation ends, the user is presented with an interface to input their satisfaction and feedback. The input feedback is sent to the server, analyzed, and used as training data for the generative AI.

[0496] Input: Conversation logs, user feedback.

[0497] Output: Improved performance of generative artificial intelligence as a result of analysis.

[0498] 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.

[0499] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0500] 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.

[0501] [Second embodiment]

[0502] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0503] 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.

[0504] 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).

[0505] 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.

[0506] 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.

[0507] 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).

[0508] 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.

[0509] 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.

[0510] 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.

[0511] 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.

[0512] In the smart glasses 214, 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.

[0513] 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."

[0514] System Overview

[0515] The present invention is a system that inputs and stores user information and uses artificial intelligence to provide virtual conversations with users. The system plays back responses based on the voice profile selected by the user, realizing personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[0516] User Registration

[0517] Step 1. Enter and save your user information

[0518] The user launches the application and first enters information such as their name, email address, and preferred voice type.

[0519] The terminal receives the input user information and transmits it to the server.

[0520] The server stores the received user information in a database and completes the initial setup.

[0521] Dialogue generation and voice selection

[0522] Step 2. Initiating a conversation and generating a response

[0523] The user presses the start conversation button in the application to begin the conversation.

[0524] The terminal converts the user's voice input into text using a voice recognition engine and transmits the text data to the server in real time.

[0525] The server passes the received text data to a generative artificial intelligence, which generates an appropriate response along with the user's past conversation history.

[0526] Step 3. Synthesize and play the response

[0527] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[0528] The terminal receives the audio file from the server and plays it for the user.

[0529] Iterative learning and feedback

[0530] Step 4. Saving and analyzing conversation logs

[0531] The server stores the contents of each conversation session as a log and analyzes it periodically.

[0532] The generative AI uses the newly saved log data as training data to continue learning about the user's language usage and stress points.

[0533] Step 5. Receive and analyze feedback

[0534] After the conversation, the user enters their satisfaction level and feedback.

[0535] The terminal transmits the feedback information to the server.

[0536] The server analyzes the received feedback information and reflects it in the training data of the generative artificial intelligence, thereby improving the quality of the next conversation.

[0537] Specific examples

[0538] Example 1: Stress-relieving conversations

[0539] The user inputs "I'm very tired today."

[0540] The server responds to the input by generating the message, "You've worked hard today. Is there anything you'd like to say?" and then uses a speech synthesis engine to convert it into the "gentle female voice" selected by the user.

[0541] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[0542] Example 2: Everyday conversation

[0543] The user types, "What should I have for dinner tonight?"

[0544] The server responds to the input by generating "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[0545] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[0546] As described above, this system effectively reduces user stress and provides a highly satisfying communication experience.

[0547] The processing flow will be explained below.

[0548] Step 1:

[0549] The user launches the application and is presented with a welcome screen where they enter initial information such as their name, email address, and preferred voice type.

[0550] Step 2:

[0551] The device receives the entered user information and sends it to the server, including information such as name, email address, and voice type.

[0552] Step 3:

[0553] The server stores the received user information in a database, which is used to tailor responses to the user's preferences in subsequent conversation sessions.

[0554] Step 4:

[0555] The user is ready to start the conversation by pressing the Start Conversation button, which starts the conversation session.

[0556] Step 5:

[0557] The device passes the user's voice input to a speech recognition engine, which converts it into text. For example, a speech phrase like "I'm tired today" is generated as text data.

[0558] Step 6:

[0559] The terminal transmits the converted text data to the server in real time.

[0560] Step 7:

[0561] The server inputs text data into a generative artificial intelligence (GPT model), which generates an appropriate response based on that data, such as, "You seem tired today. Is there anything you'd like to talk about?"

[0562] Step 8:

[0563] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[0564] Step 9:

[0565] The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[0566] Step 10:

[0567] The server stores the content of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[0568] Step 11:

[0569] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized to the user's preferences.

[0570] Step 12:

[0571] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[0572] Step 13:

[0573] The terminal sends feedback information to the server.

[0574] Step 14:

[0575] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[0576] The above is the specific processing flow of this system. Through these detailed steps, users can effectively relieve stress through individually personalized dialogue.

[0577] Example 1

[0578] 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."

[0579] Conventional user information input and dialogue systems lack the ability to accurately convert user voice input into text, generate appropriate responses, and play back the generated responses in a voice that suits the user's preferences. Furthermore, log management and feedback to analyze conversation history and improve the performance of generative AI are not fully utilized. This makes it difficult to improve user satisfaction.

[0580] 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.

[0581] In this invention, the server includes a means for storing user information, a means for generating a response based on the user's words using a generative artificial intelligence, and a means for converting the generated response into speech using a selected voice. This allows the server to convert the user's voice input into text, have the generative artificial intelligence generate an appropriate response based on the text, and then play back the response as speech based on a voice profile selected by the user. Furthermore, by including a means for storing and analyzing a conversation log and a means for improving the generative artificial intelligence based on feedback, the server can reflect the user's past conversation history and feedback, enabling more natural and effective dialogue.

[0582] "Means for inputting user information" refers to an interface that allows a user to input their own information into the system.

[0583] The "means for saving user information" is a function for temporarily or permanently storing input user information in a storage such as a database.

[0584] The "means for a user to start a conversation" refers to an operating means or interface that allows a user to start a conversation with the system.

[0585] "Means for generating responses based on user words using generative artificial intelligence" refers to algorithms or software for generating appropriate responses based on input data obtained from the user.

[0586] The "means for converting the generated response into speech in a selected voice" is a technique for converting the generated textual response into speech data based on a voice profile selected by the user.

[0587] The "means for reproducing audio to the user" refers to a device or function for allowing the user to listen to the generated audio data.

[0588] "Means for saving and analyzing conversation logs" refers to a mechanism for recording conversation history data and analyzing it to improve the system and understand user trends.

[0589] "Means for training a generative artificial intelligence" refers to a learning process that uses stored data and feedback to improve the accuracy and performance of a generative artificial intelligence.

[0590] "Means for receiving and analyzing feedback from users" is a function for receiving and analyzing opinions and satisfaction levels provided by users after using the system.

[0591] "Means for improving generative artificial intelligence based on feedback" refers to means for optimizing the behavior and responses of generative artificial intelligence by utilizing received feedback information.

[0592] The "means for converting speech input to text" is software or hardware for analyzing speech input from a user and converting it into text format data.

[0593] "Means for transmitting converted text data to the generative artificial intelligence" refers to a communication function that passes the text data converted from speech to the generative artificial intelligence to generate a response.

[0594] The "means for passing the generated response text to the speech synthesis engine" is a process for sending the text response generated by the generative artificial intelligence to the speech synthesis engine.

[0595] The "means for transmitting a synthesized voice file from a server to a terminal" is a communication technology for transferring a voice file generated on a server to a terminal and allowing the user to play it.

[0596] The present invention is a system that inputs and stores user information and uses generative artificial intelligence to provide virtual conversations with the user. The system converts the user's voice input into text, and the generative artificial intelligence generates appropriate responses based on that text. The system then converts the generated responses into speech using a voice profile selected by the user and plays them back, enabling personalized communication. The system also stores and analyzes conversation logs, receives and analyzes feedback, and incorporates it into the generative artificial intelligence's training data to continuously improve the system's performance.

[0597] The hardware and software required to implement this system are as follows:

[0598] Device: The smartphone or personal computer used by the user.

[0599] Server: Cloud server (e.g. AWS, Google Cloud).

[0600] Database: A relational database such as MySQL.

[0601] Speech recognition engine: Google Speech-to-Text, etc.

[0602] Generative artificial intelligence: OpenAI GPT-3, etc.

[0603] Speech synthesis engine: Google Text-to-Speech, Amazon Polly, etc.

[0604] Entering and saving user information

[0605] The user starts the application and enters user information, such as name, email address, and preferred voice type, into the registration form and presses the submit button. The device then sends this information to the server, which then stores it in a database.

[0606] Conversation generation and response

[0607] The user presses a button to start a conversation within the application. The device converts the user's voice input into text using a speech recognition engine and sends the generated text data to the server. The server then passes this text data to a generative artificial intelligence (AI) system, which generates an appropriate response along with the user's past conversation history. The generated response text is then passed to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The device then receives this audio file and plays it back to the user.

[0608] Iterative learning and feedback

[0609] The server saves the content of each conversation session as a log and periodically analyzes it. The generative AI then uses the newly saved log data as training data to continue learning about the user's vocabulary and stress points. After the conversation ends, the user enters their satisfaction level and feedback. The device sends this feedback information to the server, which analyzes the received information and reflects it in the generative AI's training data.

[0610] Specific examples

[0611] Example 1: Stress-relieving conversations

[0612] The user speaks, "I'm very tired today."

[0613] The terminal converts this speech into text and sends it to the server.

[0614] The server uses an AI model to generate a response such as, "You've worked hard today. Is there anything you'd like to say?"

[0615] The server passes this text to a speech synthesis engine set to a "gentle female voice" and generates an audio file.

[0616] The terminal plays this sound to the user.

[0617] Example 2: Everyday conversation

[0618] The user speaks, "What should I have for dinner tonight?"

[0619] The terminal converts this speech into text and sends it to the server.

[0620] The server uses an AI model to generate a response like, "How about pasta? It's easy and delicious."

[0621] The server passes this text to a speech synthesis engine set to a "male, low-medium voice" and generates an audio file.

[0622] The terminal plays this sound to the user.

[0623] Prompt Sentence Examples

[0624] Example prompt to generate a response to the input "I'm very tired today":

[0625] User: I'm very tired today.

[0626] AI: You've done well today. Is there anything you'd like to talk about?

[0627] An example prompt to generate a response to the input "What should I have for dinner tonight?" would be:

[0628] User: What should I have for dinner tonight?

[0629] AI: How about pasta? It's easy and delicious.

[0630] As described above, this system converts the user's voice input into text, and then uses generative AI to generate appropriate responses based on that text, and plays back the responses aloud based on the voice profile selected by the user. By managing conversation logs and utilizing feedback, the system's performance can be continuously improved.

[0631] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0632] Step 1: Enter and save user information

[0633] A user launches the application and enters information such as their name, email address, and preferred voice type.

[0634] Input: The type of personal information or voice input the user makes.

[0635] Output: Completed user profile data.

[0636] How it works: The device receives the information the user enters into the input form and sends it as an HTTP POST request to the server, which stores the information in a database and generates profile data.

[0637] Step 2: Start a conversation

[0638] The user presses the "Start Conversation" button in the application to begin the conversation.

[0639] Input: A user action to start a conversation.

[0640] Output: Start flag of conversation session.

[0641] Operation: The terminal detects the user's "start conversation" operation and sends a request to the server to start a conversation session.

[0642] Step 3: Convert voice to text

[0643] The user provides voice input during the conversation.

[0644] Input: Audio data from the user.

[0645] Output: The audio data converted to text.

[0646] How it works: The device captures the user's voice with a microphone and sends it to a speech recognition engine (e.g., Google Speech-to-Text). The speech recognition engine converts the voice data into text, and the device sends the text data to a server.

[0647] Step 4: Generate a response

[0648] The server passes the received text data to a generative artificial intelligence (AI model), which generates an appropriate response along with the user's past conversation history.

[0649] Input: User utterance data in text format.

[0650] Output: The generated response text data.

[0651] How it works: The server inputs text data in the form of a prompt sentence into a generative AI (e.g., OpenAI GPT-3) and obtains the generated response text.

[0652] Step 5: Text-to-speech response

[0653] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[0654] Input: The generated response text and the user's selected voice profile.

[0655] Output: The generated audio file.

[0656] What it does: The server sends the text data to a speech synthesis engine (e.g., Google Text-to-Speech) and generates an audio file using the specified voice profile.

[0657] Step 6: Play the audio file

[0658] The terminal plays the audio file received from the server to the user.

[0659] Input: The audio file sent from the server.

[0660] Output: The audio played to the user.

[0661] How it works: The device downloads the audio file from the server, plays it in an audio player, and allows the user to listen.

[0662] Step 7: Saving and analyzing conversation logs

[0663] The server stores the contents of each conversation session as a log and analyzes it periodically.

[0664] Input: All text data from the conversation session.

[0665] Output: Saved conversation log.

[0666] How it works: The server stores textual conversation data in a "conversation log" table in the database, and then periodically analyzes it using an analysis algorithm.

[0667] Step 8: Receive and analyze feedback

[0668] After the conversation, the user enters their satisfaction level and feedback.

[0669] Input: Feedback information from the user.

[0670] Output: Feedback data stored on the server.

[0671] How it works: The device sends the information entered into the feedback form to the server, which stores it in an analysis database.

[0672] Step 9: Training the generative AI

[0673] The server trains the generative artificial intelligence based on the stored log data and feedback information.

[0674] Input: Conversation logs and feedback data.

[0675] Output: An improved generative artificial intelligence model.

[0676] How it works: The server, through a training system, includes this data in the training set of the generative artificial intelligence, improving the AI's accuracy and quality of responses.

[0677] (Application example 1)

[0678] 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."

[0679] In today's world, there is a demand for improved personal security awareness and for rapid and effective security responses. However, it is difficult for individual users to effectively manage and use complex and diverse security information. In particular, it is often difficult to take appropriate action in emergencies that require a rapid response. For this reason, an interface that allows users to easily obtain security-related support is needed.

[0680] 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.

[0681] In this invention, the server includes means for inputting user information, means for saving user information, means for a user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, means for inputting security questions or problems by voice, means for responding to the user with security-related information, means for providing emergency response instructions, and means for functioning as a security assistant. This allows the user to easily receive security-related assistance and instructions via voice, enabling efficient and effective security responses.

[0682] "Means for inputting user information" refers to devices or software that allow users to input personal data such as their name and voice profile.

[0683] The "means for storing user information" refers to a database or storage device for retaining input user information for a long period of time.

[0684] A "means for a user to initiate a conversation" is an interface such as a button or voice command that a user uses to initiate a conversation with the system.

[0685] "Means for generating a response based on the user's words using generative artificial intelligence" is a system that uses an artificial intelligence algorithm to automatically generate an appropriate response based on the content of the user's statements.

[0686] The "means for converting the generated response into speech in a selected voice" is a speech synthesis engine for converting the response text into a speech file based on a voice profile selected by the user.

[0687] The "means for reproducing audio to the user" is a device for delivering the synthesized audio to the user via a speaker or earphones.

[0688] "Means for storing and analyzing conversation logs" refers to data storage and analysis algorithms for recording the content of conversations between users and the system and subsequently analyzing the data.

[0689] "Means for training generative artificial intelligence" refers to machine learning algorithms and computing resources that improve the performance of generative artificial intelligence based on new data.

[0690] The "means for receiving and analyzing feedback from users" refers to an interface and analysis system for collecting and analyzing feedback data such as satisfaction levels and requests for improvement provided by users.

[0691] "Means for improving generative artificial intelligence based on feedback" refers to a process for using feedback data to improve generative artificial intelligence algorithms and databases.

[0692] The "means for inputting security questions or problems by voice" is a voice recognition device that allows a user to input security questions or problems in a voice format.

[0693] The "means for responding to the user with security-related information" refers to generative artificial intelligence and response generation algorithms that provide appropriate responses to security-related questions or problems entered by the user.

[0694] The "means for providing emergency response instructions" is a system that provides instructions and advice necessary for users to take appropriate action in an emergency.

[0695] "Means to function as a security assistant" refers to services and interfaces that comprehensively provide security-related information and emergency response support.

[0696] MODE FOR CARRYING OUT THE INVENTION

[0697] System Overview

[0698] This invention is a system that inputs and stores user information and uses artificial intelligence to provide the user with virtual security-related conversations. The system plays back responses based on the voice profile selected by the user, enabling personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[0699] User Registration

[0700] Step 1. Enter and save your user information

[0701] A user starts an application and inputs information such as their name and voice type. The terminal receives the input user information and sends it to the server. The server stores the received user information in a database.

[0702] Dialogue generation and voice selection

[0703] Step 2. Initiating a conversation and generating a response

[0704] The user presses the "Start conversation" button in the application to begin the conversation. The device converts the user's voice input into text using a speech recognition engine and sends the text data to the server in real time. The server then passes the received text data and conversation history to a generative AI system, which generates an appropriate response.

[0705] Step 3. Synthesize and play the response

[0706] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The terminal receives the audio file from the server and plays it back to the user.

[0707] Iterative learning and feedback

[0708] Step 4. Saving and analyzing conversation logs

[0709] The server stores the content of each conversation session as a log and periodically analyzes it. The generative AI uses the newly saved log data as training data to improve its performance.

[0710] Step 5. Receive and analyze feedback

[0711] After the conversation ends, the user inputs their satisfaction level and feedback. The device then sends the feedback information to the server. The server analyzes the received feedback information and reflects it in the training data for the generative AI.

[0712] Specific examples of security measures

[0713] Specific examples of security assistant AI

[0714] When a user says, "I think a burglar might have broken in," the server generates a response, "Should I immediately contact the police or call a nearby security guard?" and converts it into the voice selected by the user using a speech synthesis engine. The device then plays back this generated voice and gives instructions to the user.

[0715] Example prompt sentence:

[0716] What is the course of action if a burglar breaks in?

[0717] In this embodiment of the invention, the hardware required is a microphone and speaker, and the software uses the speech_recognition library and the gtts library. The generative AI uses the GPT-2 model and the transformers library. A database and various analysis algorithms are used for data storage and analysis. This allows users to easily receive security-related assistance and instructions via voice.

[0718] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0719] Step 1:

[0720] A user starts an application and inputs user information such as name, voice type, etc. This inputs the user information.

[0721] Step 2:

[0722] The terminal receives the input user information and transmits it to the server, where the input data includes the user name and voice profile.

[0723] Step 3:

[0724] The server stores the received user information in a database, which is then used for conversation generation and voice synthesis.

[0725] Step 4:

[0726] The user presses the start speech button in the application to start the conversation, which starts voice input.

[0727] Step 5:

[0728] The device converts the user's voice input into text using a speech recognition engine (e.g., Google Speech Recognition API) and sends the text data to the server in real time. The voice data is converted into text data and sent to the server.

[0729] Step 6:

[0730] The server passes the received text data to a generative AI (e.g., GPT-2), which generates an appropriate response along with the user's past conversation history. The prompt sentence is used as input data, and the generative AI model generates text data.

[0731] Step 7:

[0732] The server passes the generated response text to a speech synthesis engine (e.g., GTTS), which generates an audio file based on the voice profile selected by the user. The text data and the voice profile are used as input data to generate the audio file.

[0733] Step 8:

[0734] The terminal receives the audio file from the server and plays it for the user, so that the audio file is played and the user can listen to it.

[0735] Step 9:

[0736] The server stores the content of each conversation session as a log and periodically analyzes it. By storing and analyzing the conversation log, the accuracy of the generative AI model is improved.

[0737] Step 10:

[0738] After the conversation ends, the user inputs their satisfaction level and feedback, which generates feedback data.

[0739] Step 11:

[0740] The terminal sends feedback information to the server. The feedback data is sent to the server.

[0741] Step 12:

[0742] The server analyzes the received feedback information and reflects it in the training data of the generative AI. The feedback data is analyzed and used to improve the generative AI model.

[0743] 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.

[0744] System Overview

[0745] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[0746] User Registration

[0747] Step 1. Enter and save your user information

[0748] A user launches the application and enters information such as their name, email address, and preferred voice type.

[0749] The terminal receives the input user information and transmits it to the server.

[0750] The server stores the received user information in a database and completes the initial setup.

[0751] Dialogue generation and voice selection

[0752] Step 2. Initiating a conversation and recognizing emotions

[0753] The user presses the start conversation button in the application and is ready to start the conversation.

[0754] The device passes the user's voice input to a speech recognition engine and converts it into text. For example, a speech such as "I'm tired today" is generated as text data.

[0755] The terminal transmits the converted text data to the server, and also passes the voice data to an emotion engine for emotion analysis.

[0756] Step 3. Emotion-based response generation

[0757] The server inputs the emotional data and text data obtained from the emotion engine into a generative AI system to generate an optimal response, such as, "You seem tired today. Would you like to know how to relax?"

[0758] Step 4. Synthesize and play the response

[0759] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[0760] The terminal receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[0761] Iterative learning and feedback

[0762] Step 5. Saving and analyzing conversation logs

[0763] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[0764] The emotion engine analyzes the user's emotional tendencies from log data and reflects this in training the generative artificial intelligence.

[0765] Step 6. Receive and analyze feedback

[0766] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[0767] The terminal transmits the feedback information to the server.

[0768] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[0769] Specific examples

[0770] Example 1: Stress-relieving conversations

[0771] The user inputs "I'm very tired today."

[0772] Based on the input and emotional data, the server generates a response such as, "You seem tired today. Shall I think of a way to help you relax?" and converts this into the "gentle female voice" selected by the user using a speech synthesis engine.

[0773] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[0774] Example 2: Everyday conversation

[0775] The user types, "What should I have for dinner tonight?"

[0776] Based on the input and emotional data, the server generates the response, "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[0777] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[0778] As described above, this system can recognize the user's emotions and generate responses based on them, providing a more personalized communication experience and effectively reducing stress.

[0779] The processing flow will be explained below.

[0780] Step 1:

[0781] The user launches the application and is presented with a welcome screen where the user enters initial information such as name, email address, and preferred voice type.

[0782] Step 2:

[0783] The device receives the entered user information and transmits the data to the server, including the user's name, email address, and selected voice profile.

[0784] Step 3:

[0785] The server stores the received user information in a database, which serves as the basis for responding to the user's preferences in subsequent conversation sessions.

[0786] Step 4:

[0787] The user clicks the Start Conversation button and is ready to begin the conversation, which causes the system to start a conversation session.

[0788] Step 5:

[0789] The device passes the user's voice input to a voice recognition engine, which converts the voice into text data. For example, the voice saying "I'm tired today" is converted into text "I'm tired today."

[0790] Step 6:

[0791] The device sends the converted text data to the server in real time, and simultaneously sends the voice data to the emotion engine.

[0792] Step 7:

[0793] The server receives the emotion data obtained from the emotion engine and the text data from the voice recognition engine, and inputs them into the generative AI. The emotion data includes information such as "fatigue."

[0794] Step 8:

[0795] Generative AI generates appropriate responses based on emotional and text data, such as "You seem tired today. Take a break and refresh yourself."

[0796] Step 9:

[0797] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. If the user selects a "soft female voice," the response will be spoken in that voice.

[0798] Step 10:

[0799] The terminal plays the audio file received from the server, allowing the user to experience a conversation with a virtual conversation partner.

[0800] Step 11:

[0801] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and emotional tendencies.

[0802] Step 12:

[0803] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized for the user's preferences and emotions.

[0804] Step 13:

[0805] After the conversation, the user inputs their satisfaction level and feedback, such as "I would like more specific advice."

[0806] Step 14:

[0807] The terminal transmits feedback information from the user to the server.

[0808] Step 15:

[0809] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[0810] The above is the specific processing flow of this system, which combines an emotion engine. Through these detailed steps, users can achieve more effective stress relief through personalized dialogue tailored to their emotions.

[0811] Example 2

[0812] 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."

[0813] In today's world, many people are increasingly experiencing stress and loneliness. In these circumstances, there is a growing need for dialogue systems that can understand emotions and respond appropriately based on those emotions. However, conventional dialogue systems lack emotion recognition capabilities, making it difficult to provide the personalized communication users desire. Furthermore, they lack natural-sounding voice synthesis based on voice profile selection and a feedback system for improving the quality of training data. This leads to issues such as reduced user satisfaction and difficulty in improving dialogue quality.

[0814] 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.

[0815] In this invention, the server includes a means for saving input information, a means for generating a response based on the user's words using generative artificial intelligence, and an emotion recognition means for analyzing emotion data and generating a response. This makes it possible to understand emotions based on the information entered by the user and generate an optimal response. Furthermore, a means for voice synthesis of the generated response based on a voice profile selected by the user is provided, making it possible to provide natural and personalized voice responses. Furthermore, by including a means for saving a conversation log as training data for the generative artificial intelligence and a means for receiving and analyzing feedback from the user, the quality of the dialogue system can be continuously improved.

[0816] 1. "Input information" refers to information such as name, email address, and voice profile that a user enters into the system.

[0817] 2. "Storage means" refers to a method or device for storing input information or conversation logs in a database or memory.

[0818] 3. "Means for starting a conversation" refers to the buttons or interface that a user operates to start a conversation.

[0819] 4. "Generative AI" refers to machine learning models and algorithms that generate optimal responses based on user input and emotional data.

[0820] 5. "Voice Profile" means a voice configuration with specific voice characteristics (e.g., pitch, gender, timbre, etc.) selected by a user.

[0821] 6. "Speech synthesis means" means a technology or device for converting text data into speech based on a specified speech profile.

[0822] 7. "Playback means" refers to an output device such as a speaker or earphones that allows the user to hear the generated audio.

[0823] 8. "Conversation log" means a record of the content of a conversation between a user and a system, including emotional data.

[0824] 9. "Emotion recognition means" means technologies or algorithms for analyzing emotions from a user's voice or text data.

[0825] 10. "Feedback" refers to the satisfaction rating and suggestions for improvement provided by the user to the system after the conversation has ended.

[0826] MODE FOR CARRYING OUT THE INVENTION

[0827] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[0828] Hardware and Software Configuration

[0829] Server: The server stores user information, runs generative AI, analyzes emotion data, generates responses, synthesizes responses, stores and analyzes conversation logs, and analyzes feedback. Specifically, it uses a database management system (e.g., MySQL or MongoDB), generative AI (e.g., OpenAI's GPT-4), emotion engine (e.g., IBM Watson Tone Analyzer), and speech synthesis engine (e.g., Amazon Polly).

[0830] Terminal: The terminal inputs user information, receives and converts voice input to text, analyzes emotions in the voice data, plays back replies, and inputs and sends feedback information. Specifically, it uses a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[0831] User: The user enters information or speaks through the application and receives spoken responses provided by the system.

[0832] Specific processing of the system

[0833] 1. Entering and saving user information: The user launches the application and enters information such as their name, email address, preferred voice type, etc. The device sends the entered information to the server, which stores it in a database.

[0834] 2. Starting a conversation and recognizing emotions: The user presses the conversation start button to prepare for the conversation. The device passes the user's voice input to the speech recognition engine, converts it into text, and sends it to the server. At the same time, the voice data is passed to the emotion engine for emotion analysis.

[0835] 3. Emotion-based response generation: The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[0836] For example, a prompt such as "The user says he is tired. Please generate the best response" is sent to the generative artificial intelligence.

[0837] 4. Response voice synthesis and playback: The server passes the generated response text to a voice synthesis engine, which generates an audio file based on the voice profile selected by the user. The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[0838] 5. Saving and analyzing conversation logs, receiving and analyzing feedback: The server saves the content and emotional data of each conversation session as a log and uses it as training data for the generative AI. After the conversation ends, the user enters feedback and sends it to the server via their device. The server analyzes the feedback information and reflects it in the training data for the generative AI to improve the quality of the next conversation.

[0839] Specific examples

[0840] Stress relief conversations:

[0841] The user speaks "I'm very tired today."

[0842] The device converts the speech into text using a voice recognition engine and sends the text data, such as "I'm tired today," to the server.

[0843] The server detects the "tired" emotion using an emotion engine and sends a prompt to the generative AI model.

[0844] The server generates a text response such as "You seem tired today. Want to think of a way to relax?"

[0845] The server uses a voice synthesis engine to convert the voice into a gentle female voice.

[0846] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[0847] Through the above process, the system can recognize the user's emotions and generate optimal responses based on those emotions, thereby providing a personalized communication experience and increasing user satisfaction.

[0848] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0849] Step 1: Enter and save user information

[0850] A user launches the application and enters information such as their name, email address, and preferred voice type.

[0851] Input: User information such as name, email address, and voice profile.

[0852] The terminal receives the entered user information and transmits this information to the server.

[0853] Output: The user information sent to the server.

[0854] The server stores the received user information in a database, typically using a database management system such as MySQL or MongoDB.

[0855] Output: User information stored in the database.

[0856] Step 2: Initiating a conversation and recognizing emotions

[0857] The user presses the start conversation button in the application and is ready to start the conversation.

[0858] The terminal receives the user's voice input and converts it into text using a speech recognition engine.

[0859] Input: User's voice input (e.g. "I'm tired today").

[0860] Output: The audio data that is passed to the speech recognition engine.

[0861] The device sends the generated text data to the server and simultaneously passes the voice data to the emotion engine for emotion analysis. For example, the device generates text data such as "I'm tired today" using the Google Cloud Speech-to-Text API.

[0862] Input: Text data and audio data.

[0863] Output: The audio data passed to the emotion engine.

[0864] The server analyzes emotions from the received voice data using an emotion engine, for example, IBM Watson Tone Analyzer.

[0865] Output: Parsed emotion data.

[0866] Step 3: Emotion-based response generation

[0867] The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[0868] Input: Emotion data and text data.

[0869] Output: The prompt input to the generative AI model (e.g., "The user says they are tired. Please generate the best response.").

[0870] The server receives responses from a generative AI model, such as OpenAI's GPT-4, to generate responses like, "You seem tired today. Want to know how to relax?"

[0871] Output: The generated response text.

[0872] Step 4: Synthesize and play the response

[0873] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[0874] Input: Response text and voice profile.

[0875] Output: The data that is passed to the speech synthesis engine.

[0876] The server sends the generated audio file to the terminal using a speech synthesis engine (e.g., Amazon Polly).

[0877] Output: The generated audio file.

[0878] The terminal receives the generated audio file and plays it for the user.

[0879] Output: The audio that is played to the user.

[0880] Step 5: Save and analyze conversation logs, receive and analyze feedback

[0881] The server stores the content and emotional data of each conversation session as a log, for example, in MongoDB.

[0882] Input: Conversational content and emotion data.

[0883] Output: Saved conversation logs.

[0884] The server uses the stored log data to train the generative artificial intelligence.

[0885] Input: Saved conversation logs.

[0886] Output: Conversation logs used as training data.

[0887] After the conversation, the user enters their satisfaction level and feedback.

[0888] Input: Satisfaction rating and suggestions for improvement.

[0889] The terminal transmits the feedback information input by the user to the server.

[0890] Output: Feedback information sent to the server.

[0891] The server analyzes the received feedback information and reflects it in the training data for the generative artificial intelligence.

[0892] Input: Feedback information.

[0893] Output: An improved generative artificial intelligence model.

[0894] Through these processing steps, the system can recognize the user's emotions and generate optimal responses, providing a personalized communication experience. It can also utilize feedback to improve the system's performance and continuously increase user satisfaction.

[0895] (Application example 2)

[0896] 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."

[0897] Ensuring personal safety and security is becoming increasingly important in modern society. However, traditional security measures are generally implemented by administrators or experts, and lack personalized, interactive security measures that reflect individual needs and feelings. Therefore, a system is needed that allows users to feel reassured in specific situations and receive appropriate advice on implementing specific security measures.

[0898] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0899] In this invention, the server includes means for inputting user information, means for saving the user information, means for the user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, a visual display device worn by the user, a speech recognition device for acquiring the user's speech, means for converting the acquired speech into text data, means for analyzing the text data using an emotion recognition engine, means for generating an optimal response based on the analyzed emotion data, and means for providing audio security information to the user. This allows the user to receive a sense of security and appropriate advice for implementing specific security measures through a dialogue that reflects their individual emotional state and security concerns in real time.

[0900] "User Information" refers to data about a user, such as their name, email address, and preferred voice type.

[0901] "Means for inputting user information" refers to a function that allows a user to input their own information through a smart device.

[0902] "Means for saving user information" refers to a function for storing input user data in a storage device such as a database.

[0903] "Means for a user to start a conversation" refers to an interface that allows a user to start a conversation with the system using voice or gestures.

[0904] "Generative AI" refers to an AI system that generates optimal responses based on user input data.

[0905] "Means for converting the generated response to speech in a selected voice" refers to functionality for converting the generated text response to speech using a voice profile selected by the user.

[0906] "Means for playing audio to the user" refers to a function for playing the generated audio through a speaker or the like so that the user can hear it.

[0907] "Means for saving and analyzing conversation logs" refers to a function for recording the content of conversations with users and analyzing that data.

[0908] "Means for training generative AI" refers to a learning function that uses collected data to improve the performance of generative AI.

[0909] "Means for receiving and analyzing feedback from users" refers to a function for collecting opinions and evaluations provided by users to the system and analyzing them.

[0910] "Means for improving generative AI based on feedback" refers to the function of improving the algorithms and response quality of generative AI based on analyzed feedback data.

[0911] "A visual display device worn by a user" refers to a display device such as smart glasses or a head-mounted display used by a user.

[0912] The "voice recognition device for acquiring the user's voice" refers to a microphone or voice input device for picking up the user's voice.

[0913] The "means for converting acquired voice into text data" refers to a function for converting voice data into text data using a voice conversion engine.

[0914] "Means for analyzing text data based on an emotion recognition engine" refers to a function that evaluates and analyzes the emotional state of converted text data using an emotion analysis engine.

[0915] "Means for generating an optimal response based on analyzed emotion data" refers to a function that generates an optimal response taking into account the results of emotion recognition.

[0916] "Means for providing audio security information to the user" refers to a function for providing the generated response or security information to the user as audio.

[0917] The specific system for realizing this invention is an interactive security assistant system designed to provide users with safety and security. This system uses the following hardware and software to handle everything from inputting user information to generating responses and playing voice.

[0918] Hardware and software used

[0919] 1. Hardware

[0920] Smart glasses: When worn by the user, they function as a display and provide audio output and input.

[0921] Microphone: Built into the smart glasses and used to capture the user's voice.

[0922] Speaker: Built into the smart glasses, used to play generated responses to the user.

[0923] 2. Software

[0924] Speech recognition engine: For example, using the Google Speech-to-Text API, converts the user's voice into text data.

[0925] Emotion recognition engine: For example, use the Azure Cognitive Services Emotion API to analyze emotions based on text data.

[0926] Generative AI: For example, OpenAI GPT-4 is used to generate optimal responses based on user input and sentiment analysis results.

[0927] A speech synthesis engine: For example, using Amazon Polly to convert the generated responses into speech.

[0928] Processing Details

[0929] 1. Enter and save user information

[0930] The user puts on the smart glasses, launches the application, and enters their name, email address, preferred voice type, etc. This information is sent to the server through the smart glasses interface and stored in a database.

[0931] 2. Conversation initiation and emotion recognition

[0932] The user initiates a conversation using voice or touch gestures. The smart glasses' microphone picks up the user's voice and converts it into text using the Google Speech-to-Text API. This text is then sent to a server where an emotion recognition engine (such as the Azure Cognitive Services Emotion API) performs emotion analysis.

[0933] 3. Generating and playing optimal responses

[0934] Based on the emotional data and text data obtained from the emotion recognition engine, a generative artificial intelligence (e.g., OpenAI GPT-4) generates the optimal response. The generated response is passed to a speech synthesis engine (e.g., Amazon Polly), which converts it into an audio file using the voice profile selected by the user. The generated audio is played back from the smartglasses' speakers, allowing the user to experience the dialogue.

[0935] 4. Saving and analyzing conversation logs

[0936] The conversation content and emotional data are stored on the server and used as a dataset to learn the user's tendencies, which improves the quality of the next conversation. Users also provide feedback after the conversation, which the server analyzes and reflects in training the generative AI.

[0937] Specific prompt examples

[0938] For example, if a user says, "I'm at home now, but I'm a little worried about my safety," the system will generate the following optimal response:

[0939] "If a user voice-inputs a security-related question like, 'I'm at home now, but I'm a little worried about my safety,' how would you provide a sense of security? Considering the emotional data, what response should your generative AI model give? Think of five specific example responses."

[0940] In this way, it is possible to provide users with a sense of security and appropriate advice for implementing specific security measures through dialogue that reflects their individual emotional state and security concerns in real time.

[0941] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0942] Step 1:

[0943] The user puts on the smart glasses and launches the application.

[0944] Specific operation: The application is launched in the smart glasses, and first an interface is displayed for the user to enter user information such as name, email address, preferred voice type, etc. The user manually enters this information and presses the send button.

[0945] Input: User information such as name, email address, and preferred voice type.

[0946] Output: The entered user information is sent to the server and stored in the database.

[0947] Step 2:

[0948] The user presses the start conversation button to prepare for the conversation. The microphone in the smart glasses picks up the user's voice and passes it to the speech recognition engine.

[0949] What it does: A user says, "I'm at home now, but I'm a little worried about my safety." The microphone captures this speech, and the Google Speech-to-Text API converts it into text data, which is then sent to the server.

[0950] Input: Speech input from the user.

[0951] Output: Text data is generated by the Google Speech-to-Text API and sent to the server.

[0952] Step 3:

[0953] The server passes the generated text data to an emotion recognition engine to analyze the emotion.

[0954] How it works: The server passes text data to the Azure Cognitive Services Emotion API and requests emotion recognition. The emotion recognition engine analyzes the user's emotions based on the text data and returns the results to the server.

[0955] Input: Text data converted from audio.

[0956] Output: Emotion data as emotion recognition results.

[0957] Step 4:

[0958] The server uses generative artificial intelligence to generate the optimal response based on emotional data and text data.

[0959] How it works: The server instructs OpenAI GPT-4 to generate the optimal response based on emotional and text data. The generative AI analyzes the data and generates an appropriate response text. For example, "Please double-check the area around your house. Are the front door and windows securely closed?"

[0960] Input: Emotion data and text data.

[0961] Output: The generated response text data.

[0962] Step 5:

[0963] The server passes the generated response text to a speech synthesis engine, which generates an audio file using the voice profile selected by the user.

[0964] What happens: The server asks Amazon Polly to convert the response text data into an audio file. Amazon Polly generates the audio file using the voice profile selected by the user and returns the file to the server.

[0965] Input: Generated response text data.

[0966] Output: The generated audio file based on the voice profile selected by the user.

[0967] Step 6:

[0968] The server sends the generated audio file to the smart glasses for playback.

[0969] What happens: The server sends an audio file to the smart glasses, and the smart glasses' speaker plays the audio. The user hears a message saying, "Please double-check your surroundings. Are the front door and windows securely closed?"

[0970] Input: The generated audio file.

[0971] Output: The sound is played and heard by the user.

[0972] Step 7:

[0973] The server stores and analyzes conversation logs, receives user feedback, and uses it to improve the generative AI.

[0974] How it works: The server records the conversation and emotional data and stores it in a database. After the conversation ends, the user is presented with an interface to input their satisfaction and feedback. The input feedback is sent to the server, analyzed, and used as training data for the generative AI.

[0975] Input: Conversation logs, user feedback.

[0976] Output: Improved performance of generative artificial intelligence as a result of analysis.

[0977] 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.

[0978] 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.

[0979] 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.

[0980] [Third embodiment]

[0981] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0982] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0983] 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).

[0984] 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.

[0985] 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.

[0986] 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).

[0987] 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.

[0988] 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.

[0989] 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.

[0990] 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.

[0991] 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.

[0992] 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."

[0993] System Overview

[0994] The present invention is a system that inputs and stores user information and uses artificial intelligence to provide virtual conversations with users. The system plays back responses based on the voice profile selected by the user, realizing personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[0995] User Registration

[0996] Step 1. Enter and save your user information

[0997] The user launches the application and first enters information such as their name, email address, and preferred voice type.

[0998] The terminal receives the input user information and transmits it to the server.

[0999] The server stores the received user information in a database and completes the initial setup.

[1000] Dialogue generation and voice selection

[1001] Step 2. Initiating a conversation and generating a response

[1002] The user presses the start conversation button in the application to begin the conversation.

[1003] The terminal converts the user's voice input into text using a voice recognition engine and transmits the text data to the server in real time.

[1004] The server passes the received text data to a generative artificial intelligence, which generates an appropriate response along with the user's past conversation history.

[1005] Step 3. Synthesize and play the response

[1006] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[1007] The terminal receives the audio file from the server and plays it for the user.

[1008] Iterative learning and feedback

[1009] Step 4. Saving and analyzing conversation logs

[1010] The server stores the contents of each conversation session as a log and analyzes it periodically.

[1011] The generative AI uses the newly saved log data as training data to continue learning about the user's language usage and stress points.

[1012] Step 5. Receive and analyze feedback

[1013] After the conversation, the user enters their satisfaction level and feedback.

[1014] The terminal transmits the feedback information to the server.

[1015] The server analyzes the received feedback information and reflects it in the training data of the generative artificial intelligence, thereby improving the quality of the next conversation.

[1016] Specific examples

[1017] Example 1: Stress-relieving conversations

[1018] The user inputs "I'm very tired today."

[1019] The server responds to the input by generating the message, "You've worked hard today. Is there anything you'd like to say?" and then uses a speech synthesis engine to convert it into the "gentle female voice" selected by the user.

[1020] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[1021] Example 2: Everyday conversation

[1022] The user types, "What should I have for dinner tonight?"

[1023] The server responds to the input by generating "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[1024] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[1025] As described above, this system effectively reduces user stress and provides a highly satisfying communication experience.

[1026] The processing flow will be explained below.

[1027] Step 1:

[1028] The user launches the application and is presented with a welcome screen where they enter initial information such as their name, email address, and preferred voice type.

[1029] Step 2:

[1030] The device receives the entered user information and sends it to the server, including information such as name, email address, and voice type.

[1031] Step 3:

[1032] The server stores the received user information in a database, which is used to tailor responses to the user's preferences in subsequent conversation sessions.

[1033] Step 4:

[1034] The user is ready to start the conversation by pressing the Start Conversation button, which starts the conversation session.

[1035] Step 5:

[1036] The device passes the user's voice input to a speech recognition engine, which converts it into text. For example, a speech phrase like "I'm tired today" is generated as text data.

[1037] Step 6:

[1038] The terminal transmits the converted text data to the server in real time.

[1039] Step 7:

[1040] The server inputs text data into a generative artificial intelligence (GPT model), which generates an appropriate response based on that data, such as, "You seem tired today. Is there anything you'd like to talk about?"

[1041] Step 8:

[1042] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[1043] Step 9:

[1044] The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[1045] Step 10:

[1046] The server stores the content of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[1047] Step 11:

[1048] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized to the user's preferences.

[1049] Step 12:

[1050] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[1051] Step 13:

[1052] The terminal sends feedback information to the server.

[1053] Step 14:

[1054] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[1055] The above is the specific processing flow of this system. Through these detailed steps, users can effectively relieve stress through individually personalized dialogue.

[1056] Example 1

[1057] 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."

[1058] Conventional user information input and dialogue systems lack the ability to accurately convert user voice input into text, generate appropriate responses, and play back the generated responses in a voice that suits the user's preferences. Furthermore, log management and feedback to analyze conversation history and improve the performance of generative AI are not fully utilized. This makes it difficult to improve user satisfaction.

[1059] 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.

[1060] In this invention, the server includes a means for storing user information, a means for generating a response based on the user's words using a generative artificial intelligence, and a means for converting the generated response into speech using a selected voice. This allows the server to convert the user's voice input into text, have the generative artificial intelligence generate an appropriate response based on the text, and then play back the response as speech based on a voice profile selected by the user. Furthermore, by including a means for storing and analyzing a conversation log and a means for improving the generative artificial intelligence based on feedback, the server can reflect the user's past conversation history and feedback, enabling more natural and effective dialogue.

[1061] "Means for inputting user information" refers to an interface that allows a user to input their own information into the system.

[1062] The "means for saving user information" is a function for temporarily or permanently storing input user information in a storage such as a database.

[1063] The "means for a user to start a conversation" refers to an operating means or interface that allows a user to start a conversation with the system.

[1064] "Means for generating responses based on user words using generative artificial intelligence" refers to algorithms or software for generating appropriate responses based on input data obtained from the user.

[1065] The "means for converting the generated response into speech in a selected voice" is a technique for converting the generated textual response into speech data based on a voice profile selected by the user.

[1066] The "means for reproducing audio to the user" refers to a device or function for allowing the user to listen to the generated audio data.

[1067] "Means for saving and analyzing conversation logs" refers to a mechanism for recording conversation history data and analyzing it to improve the system and understand user trends.

[1068] "Means for training a generative artificial intelligence" refers to a learning process that uses stored data and feedback to improve the accuracy and performance of a generative artificial intelligence.

[1069] "Means for receiving and analyzing feedback from users" is a function for receiving and analyzing opinions and satisfaction levels provided by users after using the system.

[1070] "Means for improving generative artificial intelligence based on feedback" refers to means for optimizing the behavior and responses of generative artificial intelligence by utilizing received feedback information.

[1071] The "means for converting speech input to text" is software or hardware for analyzing speech input from a user and converting it into text format data.

[1072] "Means for transmitting converted text data to the generative artificial intelligence" refers to a communication function that passes the text data converted from speech to the generative artificial intelligence to generate a response.

[1073] The "means for passing the generated response text to the speech synthesis engine" is a process for sending the text response generated by the generative artificial intelligence to the speech synthesis engine.

[1074] The "means for transmitting a synthesized voice file from a server to a terminal" is a communication technology for transferring a voice file generated on a server to a terminal and allowing the user to play it.

[1075] The present invention is a system that inputs and stores user information and uses generative artificial intelligence to provide virtual conversations with the user. The system converts the user's voice input into text, and the generative artificial intelligence generates appropriate responses based on that text. The system then converts the generated responses into speech using a voice profile selected by the user and plays them back, enabling personalized communication. The system also stores and analyzes conversation logs, receives and analyzes feedback, and incorporates it into the generative artificial intelligence's training data to continuously improve the system's performance.

[1076] The hardware and software required to implement this system are as follows:

[1077] Device: The smartphone or personal computer used by the user.

[1078] Server: Cloud server (e.g. AWS, Google Cloud).

[1079] Database: A relational database such as MySQL.

[1080] Speech recognition engine: Google Speech-to-Text, etc.

[1081] Generative artificial intelligence: OpenAI GPT-3, etc.

[1082] Speech synthesis engine: Google Text-to-Speech, Amazon Polly, etc.

[1083] Entering and saving user information

[1084] The user starts the application and enters user information, such as name, email address, and preferred voice type, into the registration form and presses the submit button. The device then sends this information to the server, which then stores it in a database.

[1085] Conversation generation and response

[1086] The user presses a button to start a conversation within the application. The device converts the user's voice input into text using a speech recognition engine and sends the generated text data to the server. The server then passes this text data to a generative artificial intelligence (AI) system, which generates an appropriate response along with the user's past conversation history. The generated response text is then passed to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The device then receives this audio file and plays it back to the user.

[1087] Iterative learning and feedback

[1088] The server saves the content of each conversation session as a log and periodically analyzes it. The generative AI then uses the newly saved log data as training data to continue learning about the user's vocabulary and stress points. After the conversation ends, the user enters their satisfaction level and feedback. The device sends this feedback information to the server, which analyzes the received information and reflects it in the generative AI's training data.

[1089] Specific examples

[1090] Example 1: Stress-relieving conversations

[1091] The user speaks, "I'm very tired today."

[1092] The terminal converts this speech into text and sends it to the server.

[1093] The server uses an AI model to generate a response such as, "You've worked hard today. Is there anything you'd like to say?"

[1094] The server passes this text to a speech synthesis engine set to a "gentle female voice" and generates an audio file.

[1095] The terminal plays this sound to the user.

[1096] Example 2: Everyday conversation

[1097] The user speaks, "What should I have for dinner tonight?"

[1098] The terminal converts this speech into text and sends it to the server.

[1099] The server uses an AI model to generate a response like, "How about pasta? It's easy and delicious."

[1100] The server passes this text to a speech synthesis engine set to a "male, low-medium voice" and generates an audio file.

[1101] The terminal plays this sound to the user.

[1102] Prompt Sentence Examples

[1103] Example prompt to generate a response to the input "I'm very tired today":

[1104] User: I'm very tired today.

[1105] AI: You've done well today. Is there anything you'd like to talk about?

[1106] An example prompt to generate a response to the input "What should I have for dinner tonight?" would be:

[1107] User: What should I have for dinner tonight?

[1108] AI: How about pasta? It's easy and delicious.

[1109] As described above, this system converts the user's voice input into text, and then uses generative AI to generate appropriate responses based on that text, and plays back the responses aloud based on the voice profile selected by the user. By managing conversation logs and utilizing feedback, the system's performance can be continuously improved.

[1110] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1111] Step 1: Enter and save user information

[1112] A user launches the application and enters information such as their name, email address, and preferred voice type.

[1113] Input: The type of personal information or voice input the user makes.

[1114] Output: Completed user profile data.

[1115] How it works: The device receives the information the user enters into the input form and sends it as an HTTP POST request to the server, which stores the information in a database and generates profile data.

[1116] Step 2: Start a conversation

[1117] The user presses the "Start Conversation" button in the application to begin the conversation.

[1118] Input: A user action to start a conversation.

[1119] Output: Start flag of conversation session.

[1120] Operation: The terminal detects the user's "start conversation" operation and sends a request to the server to start a conversation session.

[1121] Step 3: Convert voice to text

[1122] The user provides voice input during the conversation.

[1123] Input: Audio data from the user.

[1124] Output: The audio data converted to text.

[1125] How it works: The device captures the user's voice with a microphone and sends it to a speech recognition engine (e.g., Google Speech-to-Text). The speech recognition engine converts the voice data into text, and the device sends the text data to a server.

[1126] Step 4: Generate a response

[1127] The server passes the received text data to a generative artificial intelligence (AI model), which generates an appropriate response along with the user's past conversation history.

[1128] Input: User utterance data in text format.

[1129] Output: The generated response text data.

[1130] How it works: The server inputs text data in the form of a prompt sentence into a generative AI (e.g., OpenAI GPT-3) and obtains the generated response text.

[1131] Step 5: Text-to-speech response

[1132] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[1133] Input: The generated response text and the user's selected voice profile.

[1134] Output: The generated audio file.

[1135] What it does: The server sends the text data to a speech synthesis engine (e.g., Google Text-to-Speech) and generates an audio file using the specified voice profile.

[1136] Step 6: Play the audio file

[1137] The terminal plays the audio file received from the server to the user.

[1138] Input: The audio file sent from the server.

[1139] Output: The audio played to the user.

[1140] How it works: The device downloads the audio file from the server, plays it in an audio player, and allows the user to listen.

[1141] Step 7: Saving and analyzing conversation logs

[1142] The server stores the contents of each conversation session as a log and analyzes it periodically.

[1143] Input: All text data from the conversation session.

[1144] Output: Saved conversation log.

[1145] How it works: The server stores textual conversation data in a "conversation log" table in the database, and then periodically analyzes it using an analysis algorithm.

[1146] Step 8: Receive and analyze feedback

[1147] After the conversation, the user enters their satisfaction level and feedback.

[1148] Input: Feedback information from the user.

[1149] Output: Feedback data stored on the server.

[1150] How it works: The device sends the information entered into the feedback form to the server, which stores it in an analysis database.

[1151] Step 9: Training the generative AI

[1152] The server trains the generative artificial intelligence based on the stored log data and feedback information.

[1153] Input: Conversation logs and feedback data.

[1154] Output: An improved generative artificial intelligence model.

[1155] How it works: The server, through a training system, includes this data in the training set of the generative artificial intelligence, improving the AI's accuracy and quality of responses.

[1156] (Application example 1)

[1157] 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."

[1158] In today's world, there is a demand for improved personal security awareness and for rapid and effective security responses. However, it is difficult for individual users to effectively manage and use complex and diverse security information. In particular, it is often difficult to take appropriate action in emergencies that require a rapid response. For this reason, an interface that allows users to easily obtain security-related support is needed.

[1159] 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.

[1160] In this invention, the server includes means for inputting user information, means for saving user information, means for a user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, means for inputting security questions or problems by voice, means for responding to the user with security-related information, means for providing emergency response instructions, and means for functioning as a security assistant. This allows the user to easily receive security-related assistance and instructions via voice, enabling efficient and effective security responses.

[1161] "Means for inputting user information" refers to devices or software that allow users to input personal data such as their name and voice profile.

[1162] The "means for storing user information" refers to a database or storage device for retaining input user information for a long period of time.

[1163] A "means for a user to initiate a conversation" is an interface such as a button or voice command that a user uses to initiate a conversation with the system.

[1164] "Means for generating a response based on the user's words using generative artificial intelligence" is a system that uses an artificial intelligence algorithm to automatically generate an appropriate response based on the content of the user's statements.

[1165] The "means for converting the generated response into speech in a selected voice" is a speech synthesis engine for converting the response text into a speech file based on a voice profile selected by the user.

[1166] The "means for reproducing audio to the user" is a device for delivering the synthesized audio to the user via a speaker or earphones.

[1167] "Means for storing and analyzing conversation logs" refers to data storage and analysis algorithms for recording the content of conversations between users and the system and subsequently analyzing the data.

[1168] "Means for training generative artificial intelligence" refers to machine learning algorithms and computing resources that improve the performance of generative artificial intelligence based on new data.

[1169] The "means for receiving and analyzing feedback from users" refers to an interface and analysis system for collecting and analyzing feedback data such as satisfaction levels and requests for improvement provided by users.

[1170] "Means for improving generative artificial intelligence based on feedback" refers to a process for using feedback data to improve generative artificial intelligence algorithms and databases.

[1171] The "means for inputting security questions or problems by voice" is a voice recognition device that allows a user to input security questions or problems in a voice format.

[1172] The "means for responding to the user with security-related information" refers to generative artificial intelligence and response generation algorithms that provide appropriate responses to security-related questions or problems entered by the user.

[1173] The "means for providing emergency response instructions" is a system that provides instructions and advice necessary for users to take appropriate action in an emergency.

[1174] "Means to function as a security assistant" refers to services and interfaces that comprehensively provide security-related information and emergency response support.

[1175] MODE FOR CARRYING OUT THE INVENTION

[1176] System Overview

[1177] This invention is a system that inputs and stores user information and uses artificial intelligence to provide the user with virtual security-related conversations. The system plays back responses based on the voice profile selected by the user, enabling personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[1178] User Registration

[1179] Step 1. Enter and save your user information

[1180] A user starts an application and inputs information such as their name and voice type. The terminal receives the input user information and sends it to the server. The server stores the received user information in a database.

[1181] Dialogue generation and voice selection

[1182] Step 2. Initiating a conversation and generating a response

[1183] The user presses the "Start conversation" button in the application to begin the conversation. The device converts the user's voice input into text using a speech recognition engine and sends the text data to the server in real time. The server then passes the received text data and conversation history to a generative AI system, which generates an appropriate response.

[1184] Step 3. Synthesize and play the response

[1185] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The terminal receives the audio file from the server and plays it back to the user.

[1186] Iterative learning and feedback

[1187] Step 4. Saving and analyzing conversation logs

[1188] The server stores the content of each conversation session as a log and periodically analyzes it. The generative AI uses the newly saved log data as training data to improve its performance.

[1189] Step 5. Receive and analyze feedback

[1190] After the conversation ends, the user inputs their satisfaction level and feedback. The device then sends the feedback information to the server. The server analyzes the received feedback information and reflects it in the training data for the generative AI.

[1191] Specific examples of security measures

[1192] Specific examples of security assistant AI

[1193] When a user says, "I think a burglar might have broken in," the server generates a response, "Should I immediately contact the police or call a nearby security guard?" and converts it into the voice selected by the user using a speech synthesis engine. The device then plays back this generated voice and gives instructions to the user.

[1194] Example prompt sentence:

[1195] What is the course of action if a burglar breaks in?

[1196] In this embodiment of the invention, the hardware required is a microphone and speaker, and the software uses the speech_recognition library and the gtts library. The generative AI uses the GPT-2 model and the transformers library. A database and various analysis algorithms are used for data storage and analysis. This allows users to easily receive security-related assistance and instructions via voice.

[1197] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1198] Step 1:

[1199] A user starts an application and inputs user information such as name, voice type, etc. This inputs the user information.

[1200] Step 2:

[1201] The terminal receives the input user information and transmits it to the server, where the input data includes the user name and voice profile.

[1202] Step 3:

[1203] The server stores the received user information in a database, which is then used for conversation generation and voice synthesis.

[1204] Step 4:

[1205] The user presses the start speech button in the application to start the conversation, which starts voice input.

[1206] Step 5:

[1207] The device converts the user's voice input into text using a speech recognition engine (e.g., Google Speech Recognition API) and sends the text data to the server in real time. The voice data is converted into text data and sent to the server.

[1208] Step 6:

[1209] The server passes the received text data to a generative AI (e.g., GPT-2), which generates an appropriate response along with the user's past conversation history. The prompt sentence is used as input data, and the generative AI model generates text data.

[1210] Step 7:

[1211] The server passes the generated response text to a speech synthesis engine (e.g., GTTS), which generates an audio file based on the voice profile selected by the user. The text data and the voice profile are used as input data to generate the audio file.

[1212] Step 8:

[1213] The terminal receives the audio file from the server and plays it for the user, so that the audio file is played and the user can listen to it.

[1214] Step 9:

[1215] The server stores the content of each conversation session as a log and periodically analyzes it. By storing and analyzing the conversation log, the accuracy of the generative AI model is improved.

[1216] Step 10:

[1217] After the conversation ends, the user inputs their satisfaction level and feedback, which generates feedback data.

[1218] Step 11:

[1219] The terminal sends feedback information to the server. The feedback data is sent to the server.

[1220] Step 12:

[1221] The server analyzes the received feedback information and reflects it in the training data of the generative AI. The feedback data is analyzed and used to improve the generative AI model.

[1222] 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.

[1223] System Overview

[1224] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[1225] User Registration

[1226] Step 1. Enter and save your user information

[1227] A user launches the application and enters information such as their name, email address, and preferred voice type.

[1228] The terminal receives the input user information and transmits it to the server.

[1229] The server stores the received user information in a database and completes the initial setup.

[1230] Dialogue generation and voice selection

[1231] Step 2. Initiating a conversation and recognizing emotions

[1232] The user presses the start conversation button in the application and is ready to start the conversation.

[1233] The device passes the user's voice input to a speech recognition engine and converts it into text. For example, a speech such as "I'm tired today" is generated as text data.

[1234] The terminal transmits the converted text data to the server, and also passes the voice data to an emotion engine for emotion analysis.

[1235] Step 3. Emotion-based response generation

[1236] The server inputs the emotional data and text data obtained from the emotion engine into a generative AI system to generate an optimal response, such as, "You seem tired today. Would you like to know how to relax?"

[1237] Step 4. Synthesize and play the response

[1238] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[1239] The terminal receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[1240] Iterative learning and feedback

[1241] Step 5. Saving and analyzing conversation logs

[1242] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[1243] The emotion engine analyzes the user's emotional tendencies from log data and reflects this in training the generative artificial intelligence.

[1244] Step 6. Receive and analyze feedback

[1245] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[1246] The terminal transmits the feedback information to the server.

[1247] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[1248] Specific examples

[1249] Example 1: Stress-relieving conversations

[1250] The user inputs "I'm very tired today."

[1251] Based on the input and emotional data, the server generates a response such as, "You seem tired today. Shall I think of a way to help you relax?" and converts this into the "gentle female voice" selected by the user using a speech synthesis engine.

[1252] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[1253] Example 2: Everyday conversation

[1254] The user types, "What should I have for dinner tonight?"

[1255] Based on the input and emotional data, the server generates the response, "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[1256] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[1257] As described above, this system can recognize the user's emotions and generate responses based on them, providing a more personalized communication experience and effectively reducing stress.

[1258] The processing flow will be explained below.

[1259] Step 1:

[1260] The user launches the application and is presented with a welcome screen where the user enters initial information such as name, email address, and preferred voice type.

[1261] Step 2:

[1262] The device receives the entered user information and transmits the data to the server, including the user's name, email address, and selected voice profile.

[1263] Step 3:

[1264] The server stores the received user information in a database, which serves as the basis for responding to the user's preferences in subsequent conversation sessions.

[1265] Step 4:

[1266] The user clicks the Start Conversation button and is ready to begin the conversation, which causes the system to start a conversation session.

[1267] Step 5:

[1268] The device passes the user's voice input to a voice recognition engine, which converts the voice into text data. For example, the voice saying "I'm tired today" is converted into text "I'm tired today."

[1269] Step 6:

[1270] The device sends the converted text data to the server in real time, and simultaneously sends the voice data to the emotion engine.

[1271] Step 7:

[1272] The server receives the emotion data obtained from the emotion engine and the text data from the voice recognition engine, and inputs them into the generative AI. The emotion data includes information such as "fatigue."

[1273] Step 8:

[1274] Generative AI generates appropriate responses based on emotional and text data, such as "You seem tired today. Take a break and refresh yourself."

[1275] Step 9:

[1276] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. If the user selects a "soft female voice," the response will be spoken in that voice.

[1277] Step 10:

[1278] The terminal plays the audio file received from the server, allowing the user to experience a conversation with a virtual conversation partner.

[1279] Step 11:

[1280] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and emotional tendencies.

[1281] Step 12:

[1282] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized for the user's preferences and emotions.

[1283] Step 13:

[1284] After the conversation, the user inputs their satisfaction level and feedback, such as "I would like more specific advice."

[1285] Step 14:

[1286] The terminal transmits feedback information from the user to the server.

[1287] Step 15:

[1288] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[1289] The above is the specific processing flow of this system, which combines an emotion engine. Through these detailed steps, users can achieve more effective stress relief through personalized dialogue tailored to their emotions.

[1290] Example 2

[1291] 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."

[1292] In today's world, many people are increasingly experiencing stress and loneliness. In these circumstances, there is a growing need for dialogue systems that can understand emotions and respond appropriately based on those emotions. However, conventional dialogue systems lack emotion recognition capabilities, making it difficult to provide the personalized communication users desire. Furthermore, they lack natural-sounding voice synthesis based on voice profile selection and a feedback system for improving the quality of training data. This leads to issues such as reduced user satisfaction and difficulty in improving dialogue quality.

[1293] 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.

[1294] In this invention, the server includes a means for saving input information, a means for generating a response based on the user's words using generative artificial intelligence, and an emotion recognition means for analyzing emotion data and generating a response. This makes it possible to understand emotions based on the information entered by the user and generate an optimal response. Furthermore, a means for voice synthesis of the generated response based on a voice profile selected by the user is provided, making it possible to provide natural and personalized voice responses. Furthermore, by including a means for saving a conversation log as training data for the generative artificial intelligence and a means for receiving and analyzing feedback from the user, the quality of the dialogue system can be continuously improved.

[1295] 1. "Input information" refers to information such as name, email address, and voice profile that a user enters into the system.

[1296] 2. "Storage means" refers to a method or device for storing input information or conversation logs in a database or memory.

[1297] 3. "Means for starting a conversation" refers to the buttons or interface that a user operates to start a conversation.

[1298] 4. "Generative AI" refers to machine learning models and algorithms that generate optimal responses based on user input and emotional data.

[1299] 5. "Voice Profile" means a voice configuration with specific voice characteristics (e.g., pitch, gender, timbre, etc.) selected by a user.

[1300] 6. "Speech synthesis means" means a technology or device for converting text data into speech based on a specified speech profile.

[1301] 7. "Playback means" refers to an output device such as a speaker or earphones that allows the user to hear the generated audio.

[1302] 8. "Conversation log" means a record of the content of a conversation between a user and a system, including emotional data.

[1303] 9. "Emotion recognition means" means technologies or algorithms for analyzing emotions from a user's voice or text data.

[1304] 10. "Feedback" refers to the satisfaction rating and suggestions for improvement provided by the user to the system after the conversation has ended.

[1305] MODE FOR CARRYING OUT THE INVENTION

[1306] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[1307] Hardware and Software Configuration

[1308] Server: The server stores user information, runs generative AI, analyzes emotion data, generates responses, synthesizes responses, stores and analyzes conversation logs, and analyzes feedback. Specifically, it uses a database management system (e.g., MySQL or MongoDB), generative AI (e.g., OpenAI's GPT-4), emotion engine (e.g., IBM Watson Tone Analyzer), and speech synthesis engine (e.g., Amazon Polly).

[1309] Terminal: The terminal inputs user information, receives and converts voice input to text, analyzes emotions in the voice data, plays back replies, and inputs and sends feedback information. Specifically, it uses a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[1310] User: The user enters information or speaks through the application and receives spoken responses provided by the system.

[1311] Specific processing of the system

[1312] 1. Entering and saving user information: The user launches the application and enters information such as their name, email address, preferred voice type, etc. The device sends the entered information to the server, which stores it in a database.

[1313] 2. Starting a conversation and recognizing emotions: The user presses the conversation start button to prepare for the conversation. The device passes the user's voice input to the speech recognition engine, converts it into text, and sends it to the server. At the same time, the voice data is passed to the emotion engine for emotion analysis.

[1314] 3. Emotion-based response generation: The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[1315] For example, a prompt such as "The user says he is tired. Please generate the best response" is sent to the generative artificial intelligence.

[1316] 4. Response voice synthesis and playback: The server passes the generated response text to a voice synthesis engine, which generates an audio file based on the voice profile selected by the user. The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[1317] 5. Saving and analyzing conversation logs, receiving and analyzing feedback: The server saves the content and emotional data of each conversation session as a log and uses it as training data for the generative AI. After the conversation ends, the user enters feedback and sends it to the server via their device. The server analyzes the feedback information and reflects it in the training data for the generative AI to improve the quality of the next conversation.

[1318] Specific examples

[1319] Stress relief conversations:

[1320] The user speaks "I'm very tired today."

[1321] The device converts the speech into text using a voice recognition engine and sends the text data, such as "I'm tired today," to the server.

[1322] The server detects the "tired" emotion using an emotion engine and sends a prompt to the generative AI model.

[1323] The server generates a text response such as "You seem tired today. Want to think of a way to relax?"

[1324] The server uses a voice synthesis engine to convert the voice into a gentle female voice.

[1325] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[1326] Through the above process, the system can recognize the user's emotions and generate optimal responses based on those emotions, thereby providing a personalized communication experience and increasing user satisfaction.

[1327] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1328] Step 1: Enter and save user information

[1329] A user launches the application and enters information such as their name, email address, and preferred voice type.

[1330] Input: User information such as name, email address, and voice profile.

[1331] The terminal receives the entered user information and transmits this information to the server.

[1332] Output: The user information sent to the server.

[1333] The server stores the received user information in a database, typically using a database management system such as MySQL or MongoDB.

[1334] Output: User information stored in the database.

[1335] Step 2: Initiating a conversation and recognizing emotions

[1336] The user presses the start conversation button in the application and is ready to start the conversation.

[1337] The terminal receives the user's voice input and converts it into text using a speech recognition engine.

[1338] Input: User's voice input (e.g. "I'm tired today").

[1339] Output: The audio data that is passed to the speech recognition engine.

[1340] The device sends the generated text data to the server and simultaneously passes the voice data to the emotion engine for emotion analysis. For example, the device generates text data such as "I'm tired today" using the Google Cloud Speech-to-Text API.

[1341] Input: Text data and audio data.

[1342] Output: The audio data passed to the emotion engine.

[1343] The server analyzes emotions from the received voice data using an emotion engine, for example, IBM Watson Tone Analyzer.

[1344] Output: Parsed emotion data.

[1345] Step 3: Emotion-based response generation

[1346] The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[1347] Input: Emotion data and text data.

[1348] Output: The prompt input to the generative AI model (e.g., "The user says they are tired. Please generate the best response.").

[1349] The server receives responses from a generative AI model, such as OpenAI's GPT-4, to generate responses like, "You seem tired today. Want to know how to relax?"

[1350] Output: The generated response text.

[1351] Step 4: Synthesize and play the response

[1352] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[1353] Input: Response text and voice profile.

[1354] Output: The data that is passed to the speech synthesis engine.

[1355] The server sends the generated audio file to the terminal using a speech synthesis engine (e.g., Amazon Polly).

[1356] Output: The generated audio file.

[1357] The terminal receives the generated audio file and plays it for the user.

[1358] Output: The audio that is played to the user.

[1359] Step 5: Save and analyze conversation logs, receive and analyze feedback

[1360] The server stores the content and emotional data of each conversation session as a log, for example, in MongoDB.

[1361] Input: Conversational content and emotion data.

[1362] Output: Saved conversation logs.

[1363] The server uses the stored log data to train the generative artificial intelligence.

[1364] Input: Saved conversation logs.

[1365] Output: Conversation logs used as training data.

[1366] After the conversation, the user enters their satisfaction level and feedback.

[1367] Input: Satisfaction rating and suggestions for improvement.

[1368] The terminal transmits the feedback information input by the user to the server.

[1369] Output: Feedback information sent to the server.

[1370] The server analyzes the received feedback information and reflects it in the training data for the generative artificial intelligence.

[1371] Input: Feedback information.

[1372] Output: An improved generative artificial intelligence model.

[1373] Through these processing steps, the system can recognize the user's emotions and generate optimal responses, providing a personalized communication experience. It can also utilize feedback to improve the system's performance and continuously increase user satisfaction.

[1374] (Application example 2)

[1375] 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."

[1376] Ensuring personal safety and security is becoming increasingly important in modern society. However, traditional security measures are generally implemented by administrators or experts, and lack personalized, interactive security measures that reflect individual needs and feelings. Therefore, a system is needed that allows users to feel reassured in specific situations and receive appropriate advice on implementing specific security measures.

[1377] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1378] In this invention, the server includes means for inputting user information, means for saving the user information, means for the user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, a visual display device worn by the user, a speech recognition device for acquiring the user's speech, means for converting the acquired speech into text data, means for analyzing the text data using an emotion recognition engine, means for generating an optimal response based on the analyzed emotion data, and means for providing audio security information to the user. This allows the user to receive a sense of security and appropriate advice for implementing specific security measures through a dialogue that reflects their individual emotional state and security concerns in real time.

[1379] "User Information" refers to data about a user, such as their name, email address, and preferred voice type.

[1380] "Means for inputting user information" refers to a function that allows a user to input their own information through a smart device.

[1381] "Means for saving user information" refers to a function for storing input user data in a storage device such as a database.

[1382] "Means for a user to start a conversation" refers to an interface that allows a user to start a conversation with the system using voice or gestures.

[1383] "Generative AI" refers to an AI system that generates optimal responses based on user input data.

[1384] "Means for converting the generated response to speech in a selected voice" refers to functionality for converting the generated text response to speech using a voice profile selected by the user.

[1385] "Means for playing audio to the user" refers to a function for playing the generated audio through a speaker or the like so that the user can hear it.

[1386] "Means for saving and analyzing conversation logs" refers to a function for recording the content of conversations with users and analyzing that data.

[1387] "Means for training generative AI" refers to a learning function that uses collected data to improve the performance of generative AI.

[1388] "Means for receiving and analyzing feedback from users" refers to a function for collecting opinions and evaluations provided by users to the system and analyzing them.

[1389] "Means for improving generative AI based on feedback" refers to the function of improving the algorithms and response quality of generative AI based on analyzed feedback data.

[1390] "A visual display device worn by a user" refers to a display device such as smart glasses or a head-mounted display used by a user.

[1391] The "voice recognition device for acquiring the user's voice" refers to a microphone or voice input device for picking up the user's voice.

[1392] The "means for converting acquired voice into text data" refers to a function for converting voice data into text data using a voice conversion engine.

[1393] "Means for analyzing text data based on an emotion recognition engine" refers to a function that evaluates and analyzes the emotional state of converted text data using an emotion analysis engine.

[1394] "Means for generating an optimal response based on analyzed emotion data" refers to a function that generates an optimal response taking into account the results of emotion recognition.

[1395] "Means for providing audio security information to the user" refers to a function for providing the generated response or security information to the user as audio.

[1396] The specific system for realizing this invention is an interactive security assistant system designed to provide users with safety and security. This system uses the following hardware and software to handle everything from inputting user information to generating responses and playing voice.

[1397] Hardware and software used

[1398] 1. Hardware

[1399] Smart glasses: When worn by the user, they function as a display and provide audio output and input.

[1400] Microphone: Built into the smart glasses and used to capture the user's voice.

[1401] Speaker: Built into the smart glasses, used to play generated responses to the user.

[1402] 2. Software

[1403] Speech recognition engine: For example, using the Google Speech-to-Text API, converts the user's voice into text data.

[1404] Emotion recognition engine: For example, use the Azure Cognitive Services Emotion API to analyze emotions based on text data.

[1405] Generative AI: For example, OpenAI GPT-4 is used to generate optimal responses based on user input and sentiment analysis results.

[1406] A speech synthesis engine: For example, using Amazon Polly to convert the generated responses into speech.

[1407] Processing Details

[1408] 1. Enter and save user information

[1409] The user puts on the smart glasses, launches the application, and enters their name, email address, preferred voice type, etc. This information is sent to the server through the smart glasses interface and stored in a database.

[1410] 2. Conversation initiation and emotion recognition

[1411] The user initiates a conversation using voice or touch gestures. The smart glasses' microphone picks up the user's voice and converts it into text using the Google Speech-to-Text API. This text is then sent to a server where an emotion recognition engine (such as the Azure Cognitive Services Emotion API) performs emotion analysis.

[1412] 3. Generating and playing optimal responses

[1413] Based on the emotional data and text data obtained from the emotion recognition engine, a generative artificial intelligence (e.g., OpenAI GPT-4) generates the optimal response. The generated response is passed to a speech synthesis engine (e.g., Amazon Polly), which converts it into an audio file using the voice profile selected by the user. The generated audio is played back from the smartglasses' speakers, allowing the user to experience the dialogue.

[1414] 4. Saving and analyzing conversation logs

[1415] The conversation content and emotional data are stored on the server and used as a dataset to learn the user's tendencies, which improves the quality of the next conversation. Users also provide feedback after the conversation, which the server analyzes and reflects in training the generative AI.

[1416] Specific prompt examples

[1417] For example, if a user says, "I'm at home now, but I'm a little worried about my safety," the system will generate the following optimal response:

[1418] "If a user voice-inputs a security-related question like, 'I'm at home now, but I'm a little worried about my safety,' how would you provide a sense of security? Considering the emotional data, what response should your generative AI model give? Think of five specific example responses."

[1419] In this way, it is possible to provide users with a sense of security and appropriate advice for implementing specific security measures through dialogue that reflects their individual emotional state and security concerns in real time.

[1420] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1421] Step 1:

[1422] The user puts on the smart glasses and launches the application.

[1423] Specific operation: The application is launched in the smart glasses, and first an interface is displayed for the user to enter user information such as name, email address, preferred voice type, etc. The user manually enters this information and presses the send button.

[1424] Input: User information such as name, email address, and preferred voice type.

[1425] Output: The entered user information is sent to the server and stored in the database.

[1426] Step 2:

[1427] The user presses the start conversation button to prepare for the conversation. The microphone in the smart glasses picks up the user's voice and passes it to the speech recognition engine.

[1428] What it does: A user says, "I'm at home now, but I'm a little worried about my safety." The microphone captures this speech, and the Google Speech-to-Text API converts it into text data, which is then sent to the server.

[1429] Input: Speech input from the user.

[1430] Output: Text data is generated by the Google Speech-to-Text API and sent to the server.

[1431] Step 3:

[1432] The server passes the generated text data to an emotion recognition engine to analyze the emotion.

[1433] How it works: The server passes text data to the Azure Cognitive Services Emotion API and requests emotion recognition. The emotion recognition engine analyzes the user's emotions based on the text data and returns the results to the server.

[1434] Input: Text data converted from audio.

[1435] Output: Emotion data as emotion recognition results.

[1436] Step 4:

[1437] The server uses generative artificial intelligence to generate the optimal response based on emotional data and text data.

[1438] How it works: The server instructs OpenAI GPT-4 to generate the optimal response based on emotional and text data. The generative AI analyzes the data and generates an appropriate response text. For example, "Please double-check the area around your house. Are the front door and windows securely closed?"

[1439] Input: Emotion data and text data.

[1440] Output: The generated response text data.

[1441] Step 5:

[1442] The server passes the generated response text to a speech synthesis engine, which generates an audio file using the voice profile selected by the user.

[1443] What happens: The server asks Amazon Polly to convert the response text data into an audio file. Amazon Polly generates the audio file using the voice profile selected by the user and returns the file to the server.

[1444] Input: Generated response text data.

[1445] Output: The generated audio file based on the voice profile selected by the user.

[1446] Step 6:

[1447] The server sends the generated audio file to the smart glasses for playback.

[1448] What happens: The server sends an audio file to the smart glasses, and the smart glasses' speaker plays the audio. The user hears a message saying, "Please double-check your surroundings. Are the front door and windows securely closed?"

[1449] Input: The generated audio file.

[1450] Output: The sound is played and heard by the user.

[1451] Step 7:

[1452] The server stores and analyzes conversation logs, receives user feedback, and uses it to improve the generative AI.

[1453] How it works: The server records the conversation and emotional data and stores it in a database. After the conversation ends, the user is presented with an interface to input their satisfaction and feedback. The input feedback is sent to the server, analyzed, and used as training data for the generative AI.

[1454] Input: Conversation logs, user feedback.

[1455] Output: Improved performance of generative artificial intelligence as a result of analysis.

[1456] 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.

[1457] 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.

[1458] 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.

[1459] [Fourth embodiment]

[1460] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1461] 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.

[1462] 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).

[1463] 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.

[1464] 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.

[1465] 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).

[1466] 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.

[1467] 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.

[1468] 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.

[1469] 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.

[1470] 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.

[1471] 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.

[1472] 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."

[1473] System Overview

[1474] The present invention is a system that inputs and stores user information and uses artificial intelligence to provide virtual conversations with users. The system plays back responses based on the voice profile selected by the user, realizing personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[1475] User Registration

[1476] Step 1. Enter and save your user information

[1477] The user launches the application and first enters information such as their name, email address, and preferred voice type.

[1478] The terminal receives the input user information and transmits it to the server.

[1479] The server stores the received user information in a database and completes the initial setup.

[1480] Dialogue generation and voice selection

[1481] Step 2. Initiating a conversation and generating a response

[1482] The user presses the start conversation button in the application to begin the conversation.

[1483] The terminal converts the user's voice input into text using a voice recognition engine and transmits the text data to the server in real time.

[1484] The server passes the received text data to a generative artificial intelligence, which generates an appropriate response along with the user's past conversation history.

[1485] Step 3. Synthesize and play the response

[1486] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[1487] The terminal receives the audio file from the server and plays it for the user.

[1488] Iterative learning and feedback

[1489] Step 4. Saving and analyzing conversation logs

[1490] The server stores the contents of each conversation session as a log and analyzes it periodically.

[1491] The generative AI uses the newly saved log data as training data to continue learning about the user's language usage and stress points.

[1492] Step 5. Receive and analyze feedback

[1493] After the conversation, the user enters their satisfaction level and feedback.

[1494] The terminal transmits the feedback information to the server.

[1495] The server analyzes the received feedback information and reflects it in the training data of the generative artificial intelligence, thereby improving the quality of the next conversation.

[1496] Specific examples

[1497] Example 1: Stress-relieving conversations

[1498] The user inputs "I'm very tired today."

[1499] The server responds to the input by generating the message, "You've worked hard today. Is there anything you'd like to say?" and then uses a speech synthesis engine to convert it into the "gentle female voice" selected by the user.

[1500] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[1501] Example 2: Everyday conversation

[1502] The user types, "What should I have for dinner tonight?"

[1503] The server responds to the input by generating "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[1504] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[1505] As described above, this system effectively reduces user stress and provides a highly satisfying communication experience.

[1506] The processing flow will be explained below.

[1507] Step 1:

[1508] The user launches the application and is presented with a welcome screen where they enter initial information such as their name, email address, and preferred voice type.

[1509] Step 2:

[1510] The device receives the entered user information and sends it to the server, including information such as name, email address, and voice type.

[1511] Step 3:

[1512] The server stores the received user information in a database, which is used to tailor responses to the user's preferences in subsequent conversation sessions.

[1513] Step 4:

[1514] The user is ready to start the conversation by pressing the Start Conversation button, which starts the conversation session.

[1515] Step 5:

[1516] The device passes the user's voice input to a speech recognition engine, which converts it into text. For example, a speech phrase like "I'm tired today" is generated as text data.

[1517] Step 6:

[1518] The terminal transmits the converted text data to the server in real time.

[1519] Step 7:

[1520] The server inputs text data into a generative artificial intelligence (GPT model), which generates an appropriate response based on that data, such as, "You seem tired today. Is there anything you'd like to talk about?"

[1521] Step 8:

[1522] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[1523] Step 9:

[1524] The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[1525] Step 10:

[1526] The server stores the content of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[1527] Step 11:

[1528] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized to the user's preferences.

[1529] Step 12:

[1530] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[1531] Step 13:

[1532] The terminal sends feedback information to the server.

[1533] Step 14:

[1534] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[1535] The above is the specific processing flow of this system. Through these detailed steps, users can effectively relieve stress through individually personalized dialogue.

[1536] Example 1

[1537] 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."

[1538] Conventional user information input and dialogue systems lack the ability to accurately convert user voice input into text, generate appropriate responses, and play back the generated responses in a voice that suits the user's preferences. Furthermore, log management and feedback to analyze conversation history and improve the performance of generative AI are not fully utilized. This makes it difficult to improve user satisfaction.

[1539] 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.

[1540] In this invention, the server includes a means for storing user information, a means for generating a response based on the user's words using a generative artificial intelligence, and a means for converting the generated response into speech using a selected voice. This allows the server to convert the user's voice input into text, have the generative artificial intelligence generate an appropriate response based on the text, and then play back the response as speech based on a voice profile selected by the user. Furthermore, by including a means for storing and analyzing a conversation log and a means for improving the generative artificial intelligence based on feedback, the server can reflect the user's past conversation history and feedback, enabling more natural and effective dialogue.

[1541] "Means for inputting user information" refers to an interface that allows a user to input their own information into the system.

[1542] The "means for saving user information" is a function for temporarily or permanently storing input user information in a storage such as a database.

[1543] The "means for a user to start a conversation" refers to an operating means or interface that allows a user to start a conversation with the system.

[1544] "Means for generating responses based on user words using generative artificial intelligence" refers to algorithms or software for generating appropriate responses based on input data obtained from the user.

[1545] The "means for converting the generated response into speech in a selected voice" is a technique for converting the generated textual response into speech data based on a voice profile selected by the user.

[1546] The "means for reproducing audio to the user" refers to a device or function for allowing the user to listen to the generated audio data.

[1547] "Means for saving and analyzing conversation logs" refers to a mechanism for recording conversation history data and analyzing it to improve the system and understand user trends.

[1548] "Means for training a generative artificial intelligence" refers to a learning process that uses stored data and feedback to improve the accuracy and performance of a generative artificial intelligence.

[1549] "Means for receiving and analyzing feedback from users" is a function for receiving and analyzing opinions and satisfaction levels provided by users after using the system.

[1550] "Means for improving generative artificial intelligence based on feedback" refers to means for optimizing the behavior and responses of generative artificial intelligence by utilizing received feedback information.

[1551] The "means for converting speech input to text" is software or hardware for analyzing speech input from a user and converting it into text format data.

[1552] "Means for transmitting converted text data to the generative artificial intelligence" refers to a communication function that passes the text data converted from speech to the generative artificial intelligence to generate a response.

[1553] The "means for passing the generated response text to the speech synthesis engine" is a process for sending the text response generated by the generative artificial intelligence to the speech synthesis engine.

[1554] The "means for transmitting a synthesized voice file from a server to a terminal" is a communication technology for transferring a voice file generated on a server to a terminal and allowing the user to play it.

[1555] The present invention is a system that inputs and stores user information and uses generative artificial intelligence to provide virtual conversations with the user. The system converts the user's voice input into text, and the generative artificial intelligence generates appropriate responses based on that text. The system then converts the generated responses into speech using a voice profile selected by the user and plays them back, enabling personalized communication. The system also stores and analyzes conversation logs, receives and analyzes feedback, and incorporates it into the generative artificial intelligence's training data to continuously improve the system's performance.

[1556] The hardware and software required to implement this system are as follows:

[1557] Device: The smartphone or personal computer used by the user.

[1558] Server: Cloud server (e.g. AWS, Google Cloud).

[1559] Database: A relational database such as MySQL.

[1560] Speech recognition engine: Google Speech-to-Text, etc.

[1561] Generative artificial intelligence: OpenAI GPT-3, etc.

[1562] Speech synthesis engine: Google Text-to-Speech, Amazon Polly, etc.

[1563] Entering and saving user information

[1564] The user starts the application and enters user information, such as name, email address, and preferred voice type, into the registration form and presses the submit button. The device then sends this information to the server, which then stores it in a database.

[1565] Conversation generation and response

[1566] The user presses a button to start a conversation within the application. The device converts the user's voice input into text using a speech recognition engine and sends the generated text data to the server. The server then passes this text data to a generative artificial intelligence (AI) system, which generates an appropriate response along with the user's past conversation history. The generated response text is then passed to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The device then receives this audio file and plays it back to the user.

[1567] Iterative learning and feedback

[1568] The server saves the content of each conversation session as a log and periodically analyzes it. The generative AI then uses the newly saved log data as training data to continue learning about the user's vocabulary and stress points. After the conversation ends, the user enters their satisfaction level and feedback. The device sends this feedback information to the server, which analyzes the received information and reflects it in the generative AI's training data.

[1569] Specific examples

[1570] Example 1: Stress-relieving conversations

[1571] The user speaks, "I'm very tired today."

[1572] The terminal converts this speech into text and sends it to the server.

[1573] The server uses an AI model to generate a response such as, "You've worked hard today. Is there anything you'd like to say?"

[1574] The server passes this text to a speech synthesis engine set to a "gentle female voice" and generates an audio file.

[1575] The terminal plays this sound to the user.

[1576] Example 2: Everyday conversation

[1577] The user speaks, "What should I have for dinner tonight?"

[1578] The terminal converts this speech into text and sends it to the server.

[1579] The server uses an AI model to generate a response like, "How about pasta? It's easy and delicious."

[1580] The server passes this text to a speech synthesis engine set to a "male, low-medium voice" and generates an audio file.

[1581] The terminal plays this sound to the user.

[1582] Prompt Sentence Examples

[1583] Example prompt to generate a response to the input "I'm very tired today":

[1584] User: I'm very tired today.

[1585] AI: You've done well today. Is there anything you'd like to talk about?

[1586] An example prompt to generate a response to the input "What should I have for dinner tonight?" would be:

[1587] User: What should I have for dinner tonight?

[1588] AI: How about pasta? It's easy and delicious.

[1589] As described above, this system converts the user's voice input into text, and then uses generative AI to generate appropriate responses based on that text, and plays back the responses aloud based on the voice profile selected by the user. By managing conversation logs and utilizing feedback, the system's performance can be continuously improved.

[1590] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1591] Step 1: Enter and save user information

[1592] A user launches the application and enters information such as their name, email address, and preferred voice type.

[1593] Input: The type of personal information or voice input the user makes.

[1594] Output: Completed user profile data.

[1595] How it works: The device receives the information the user enters into the input form and sends it as an HTTP POST request to the server, which stores the information in a database and generates profile data.

[1596] Step 2: Start a conversation

[1597] The user presses the "Start Conversation" button in the application to begin the conversation.

[1598] Input: A user action to start a conversation.

[1599] Output: Start flag of conversation session.

[1600] Operation: The terminal detects the user's "start conversation" operation and sends a request to the server to start a conversation session.

[1601] Step 3: Convert voice to text

[1602] The user provides voice input during the conversation.

[1603] Input: Audio data from the user.

[1604] Output: The audio data converted to text.

[1605] How it works: The device captures the user's voice with a microphone and sends it to a speech recognition engine (e.g., Google Speech-to-Text). The speech recognition engine converts the voice data into text, and the device sends the text data to a server.

[1606] Step 4: Generate a response

[1607] The server passes the received text data to a generative artificial intelligence (AI model), which generates an appropriate response along with the user's past conversation history.

[1608] Input: User utterance data in text format.

[1609] Output: The generated response text data.

[1610] How it works: The server inputs text data in the form of a prompt sentence into a generative AI (e.g., OpenAI GPT-3) and obtains the generated response text.

[1611] Step 5: Text-to-speech response

[1612] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[1613] Input: The generated response text and the user's selected voice profile.

[1614] Output: The generated audio file.

[1615] What it does: The server sends the text data to a speech synthesis engine (e.g., Google Text-to-Speech) and generates an audio file using the specified voice profile.

[1616] Step 6: Play the audio file

[1617] The terminal plays the audio file received from the server to the user.

[1618] Input: The audio file sent from the server.

[1619] Output: The audio played to the user.

[1620] How it works: The device downloads the audio file from the server, plays it in an audio player, and allows the user to listen.

[1621] Step 7: Saving and analyzing conversation logs

[1622] The server stores the contents of each conversation session as a log and analyzes it periodically.

[1623] Input: All text data from the conversation session.

[1624] Output: Saved conversation log.

[1625] How it works: The server stores textual conversation data in a "conversation log" table in the database, and then periodically analyzes it using an analysis algorithm.

[1626] Step 8: Receive and analyze feedback

[1627] After the conversation, the user enters their satisfaction level and feedback.

[1628] Input: Feedback information from the user.

[1629] Output: Feedback data stored on the server.

[1630] How it works: The device sends the information entered into the feedback form to the server, which stores it in an analysis database.

[1631] Step 9: Training the generative AI

[1632] The server trains the generative artificial intelligence based on the stored log data and feedback information.

[1633] Input: Conversation logs and feedback data.

[1634] Output: An improved generative artificial intelligence model.

[1635] How it works: The server, through a training system, includes this data in the training set of the generative artificial intelligence, improving the AI's accuracy and quality of responses.

[1636] (Application example 1)

[1637] 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."

[1638] In today's world, there is a demand for improved personal security awareness and for rapid and effective security responses. However, it is difficult for individual users to effectively manage and use complex and diverse security information. In particular, it is often difficult to take appropriate action in emergencies that require a rapid response. For this reason, an interface that allows users to easily obtain security-related support is needed.

[1639] 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.

[1640] In this invention, the server includes means for inputting user information, means for saving user information, means for a user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, means for inputting security questions or problems by voice, means for responding to the user with security-related information, means for providing emergency response instructions, and means for functioning as a security assistant. This allows the user to easily receive security-related assistance and instructions via voice, enabling efficient and effective security responses.

[1641] "Means for inputting user information" refers to devices or software that allow users to input personal data such as their name and voice profile.

[1642] The "means for storing user information" refers to a database or storage device for retaining input user information for a long period of time.

[1643] A "means for a user to initiate a conversation" is an interface such as a button or voice command that a user uses to initiate a conversation with the system.

[1644] "Means for generating a response based on the user's words using generative artificial intelligence" is a system that uses an artificial intelligence algorithm to automatically generate an appropriate response based on the content of the user's statements.

[1645] The "means for converting the generated response into speech in a selected voice" is a speech synthesis engine for converting the response text into a speech file based on a voice profile selected by the user.

[1646] The "means for reproducing audio to the user" is a device for delivering the synthesized audio to the user via a speaker or earphones.

[1647] "Means for storing and analyzing conversation logs" refers to data storage and analysis algorithms for recording the content of conversations between users and the system and subsequently analyzing the data.

[1648] "Means for training generative artificial intelligence" refers to machine learning algorithms and computing resources that improve the performance of generative artificial intelligence based on new data.

[1649] The "means for receiving and analyzing feedback from users" refers to an interface and analysis system for collecting and analyzing feedback data such as satisfaction levels and requests for improvement provided by users.

[1650] "Means for improving generative artificial intelligence based on feedback" refers to a process for using feedback data to improve generative artificial intelligence algorithms and databases.

[1651] The "means for inputting security questions or problems by voice" is a voice recognition device that allows a user to input security questions or problems in a voice format.

[1652] The "means for responding to the user with security-related information" refers to generative artificial intelligence and response generation algorithms that provide appropriate responses to security-related questions or problems entered by the user.

[1653] The "means for providing emergency response instructions" is a system that provides instructions and advice necessary for users to take appropriate action in an emergency.

[1654] "Means to function as a security assistant" refers to services and interfaces that comprehensively provide security-related information and emergency response support.

[1655] MODE FOR CARRYING OUT THE INVENTION

[1656] System Overview

[1657] This invention is a system that inputs and stores user information and uses artificial intelligence to provide the user with virtual security-related conversations. The system plays back responses based on the voice profile selected by the user, enabling personalized communication. The system also has the ability to analyze conversation logs and continuously improve the performance of the artificial intelligence.

[1658] User Registration

[1659] Step 1. Enter and save your user information

[1660] A user starts an application and inputs information such as their name and voice type. The terminal receives the input user information and sends it to the server. The server stores the received user information in a database.

[1661] Dialogue generation and voice selection

[1662] Step 2. Initiating a conversation and generating a response

[1663] The user presses the "Start conversation" button in the application to begin the conversation. The device converts the user's voice input into text using a speech recognition engine and sends the text data to the server in real time. The server then passes the received text data and conversation history to a generative AI system, which generates an appropriate response.

[1664] Step 3. Synthesize and play the response

[1665] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. The terminal receives the audio file from the server and plays it back to the user.

[1666] Iterative learning and feedback

[1667] Step 4. Saving and analyzing conversation logs

[1668] The server stores the content of each conversation session as a log and periodically analyzes it. The generative AI uses the newly saved log data as training data to improve its performance.

[1669] Step 5. Receive and analyze feedback

[1670] After the conversation ends, the user inputs their satisfaction level and feedback. The device then sends the feedback information to the server. The server analyzes the received feedback information and reflects it in the training data for the generative AI.

[1671] Specific examples of security measures

[1672] Specific examples of security assistant AI

[1673] When a user says, "I think a burglar might have broken in," the server generates a response, "Should I immediately contact the police or call a nearby security guard?" and converts it into the voice selected by the user using a speech synthesis engine. The device then plays back this generated voice and gives instructions to the user.

[1674] Example prompt sentence:

[1675] What is the course of action if a burglar breaks in?

[1676] In this embodiment of the invention, the hardware required is a microphone and speaker, and the software uses the speech_recognition library and the gtts library. The generative AI uses the GPT-2 model and the transformers library. A database and various analysis algorithms are used for data storage and analysis. This allows users to easily receive security-related assistance and instructions via voice.

[1677] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1678] Step 1:

[1679] A user starts an application and inputs user information such as name, voice type, etc. This inputs the user information.

[1680] Step 2:

[1681] The terminal receives the input user information and transmits it to the server, where the input data includes the user name and voice profile.

[1682] Step 3:

[1683] The server stores the received user information in a database, which is then used for conversation generation and voice synthesis.

[1684] Step 4:

[1685] The user presses the start speech button in the application to start the conversation, which starts voice input.

[1686] Step 5:

[1687] The device converts the user's voice input into text using a speech recognition engine (e.g., Google Speech Recognition API) and sends the text data to the server in real time. The voice data is converted into text data and sent to the server.

[1688] Step 6:

[1689] The server passes the received text data to a generative AI (e.g., GPT-2), which generates an appropriate response along with the user's past conversation history. The prompt sentence is used as input data, and the generative AI model generates text data.

[1690] Step 7:

[1691] The server passes the generated response text to a speech synthesis engine (e.g., GTTS), which generates an audio file based on the voice profile selected by the user. The text data and the voice profile are used as input data to generate the audio file.

[1692] Step 8:

[1693] The terminal receives the audio file from the server and plays it for the user, so that the audio file is played and the user can listen to it.

[1694] Step 9:

[1695] The server stores the content of each conversation session as a log and periodically analyzes it. By storing and analyzing the conversation log, the accuracy of the generative AI model is improved.

[1696] Step 10:

[1697] After the conversation ends, the user inputs their satisfaction level and feedback, which generates feedback data.

[1698] Step 11:

[1699] The terminal sends feedback information to the server. The feedback data is sent to the server.

[1700] Step 12:

[1701] The server analyzes the received feedback information and reflects it in the training data of the generative AI. The feedback data is analyzed and used to improve the generative AI model.

[1702] 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.

[1703] System Overview

[1704] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[1705] User Registration

[1706] Step 1. Enter and save your user information

[1707] A user launches the application and enters information such as their name, email address, and preferred voice type.

[1708] The terminal receives the input user information and transmits it to the server.

[1709] The server stores the received user information in a database and completes the initial setup.

[1710] Dialogue generation and voice selection

[1711] Step 2. Initiating a conversation and recognizing emotions

[1712] The user presses the start conversation button in the application and is ready to start the conversation.

[1713] The device passes the user's voice input to a speech recognition engine and converts it into text. For example, a speech such as "I'm tired today" is generated as text data.

[1714] The terminal transmits the converted text data to the server, and also passes the voice data to an emotion engine for emotion analysis.

[1715] Step 3. Emotion-based response generation

[1716] The server inputs the emotional data and text data obtained from the emotion engine into a generative AI system to generate an optimal response, such as, "You seem tired today. Would you like to know how to relax?"

[1717] Step 4. Synthesize and play the response

[1718] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user, for example, a "gentle female voice."

[1719] The terminal receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[1720] Iterative learning and feedback

[1721] Step 5. Saving and analyzing conversation logs

[1722] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and stress points.

[1723] The emotion engine analyzes the user's emotional tendencies from log data and reflects this in training the generative artificial intelligence.

[1724] Step 6. Receive and analyze feedback

[1725] After the conversation, the user can input their satisfaction and feedback. For example, they can rate the content of the conversation on a 5-point scale and input opinions such as "I would like more specific advice."

[1726] The terminal transmits the feedback information to the server.

[1727] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[1728] Specific examples

[1729] Example 1: Stress-relieving conversations

[1730] The user inputs "I'm very tired today."

[1731] Based on the input and emotional data, the server generates a response such as, "You seem tired today. Shall I think of a way to help you relax?" and converts this into the "gentle female voice" selected by the user using a speech synthesis engine.

[1732] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[1733] Example 2: Everyday conversation

[1734] The user types, "What should I have for dinner tonight?"

[1735] Based on the input and emotional data, the server generates the response, "How about pasta? It's easy and delicious," and converts it into the "male mid-low voice" selected by the user using a speech synthesis engine.

[1736] The terminal plays back the generated voice, and the user enjoys a relaxed conversation like a conversation with a friend.

[1737] As described above, this system can recognize the user's emotions and generate responses based on them, providing a more personalized communication experience and effectively reducing stress.

[1738] The processing flow will be explained below.

[1739] Step 1:

[1740] The user launches the application and is presented with a welcome screen where the user enters initial information such as name, email address, and preferred voice type.

[1741] Step 2:

[1742] The device receives the entered user information and transmits the data to the server, including the user's name, email address, and selected voice profile.

[1743] Step 3:

[1744] The server stores the received user information in a database, which serves as the basis for responding to the user's preferences in subsequent conversation sessions.

[1745] Step 4:

[1746] The user clicks the Start Conversation button and is ready to begin the conversation, which causes the system to start a conversation session.

[1747] Step 5:

[1748] The device passes the user's voice input to a voice recognition engine, which converts the voice into text data. For example, the voice saying "I'm tired today" is converted into text "I'm tired today."

[1749] Step 6:

[1750] The device sends the converted text data to the server in real time, and simultaneously sends the voice data to the emotion engine.

[1751] Step 7:

[1752] The server receives the emotion data obtained from the emotion engine and the text data from the voice recognition engine, and inputs them into the generative AI. The emotion data includes information such as "fatigue."

[1753] Step 8:

[1754] Generative AI generates appropriate responses based on emotional and text data, such as "You seem tired today. Take a break and refresh yourself."

[1755] Step 9:

[1756] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user. If the user selects a "soft female voice," the response will be spoken in that voice.

[1757] Step 10:

[1758] The terminal plays the audio file received from the server, allowing the user to experience a conversation with a virtual conversation partner.

[1759] Step 11:

[1760] The server stores the content and emotional data of each conversation session as a log, which is used as a dataset to learn the user's vocabulary and emotional tendencies.

[1761] Step 12:

[1762] The server periodically analyzes the log data and feeds it back to the generative AI as new training data, improving the AI's ability to generate responses optimized for the user's preferences and emotions.

[1763] Step 13:

[1764] After the conversation, the user inputs their satisfaction level and feedback, such as "I would like more specific advice."

[1765] Step 14:

[1766] The terminal transmits feedback information from the user to the server.

[1767] Step 15:

[1768] The server analyzes the received feedback information and reflects it in the training data of the generative AI, improving the quality of the next conversation and providing the user with a more satisfying communication experience.

[1769] The above is the specific processing flow of this system, which combines an emotion engine. Through these detailed steps, users can achieve more effective stress relief through personalized dialogue tailored to their emotions.

[1770] Example 2

[1771] 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."

[1772] In today's world, many people are increasingly experiencing stress and loneliness. In these circumstances, there is a growing need for dialogue systems that can understand emotions and respond appropriately based on those emotions. However, conventional dialogue systems lack emotion recognition capabilities, making it difficult to provide the personalized communication users desire. Furthermore, they lack natural-sounding voice synthesis based on voice profile selection and a feedback system for improving the quality of training data. This leads to issues such as reduced user satisfaction and difficulty in improving dialogue quality.

[1773] 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.

[1774] In this invention, the server includes a means for saving input information, a means for generating a response based on the user's words using generative artificial intelligence, and an emotion recognition means for analyzing emotion data and generating a response. This makes it possible to understand emotions based on the information entered by the user and generate an optimal response. Furthermore, a means for voice synthesis of the generated response based on a voice profile selected by the user is provided, making it possible to provide natural and personalized voice responses. Furthermore, by including a means for saving a conversation log as training data for the generative artificial intelligence and a means for receiving and analyzing feedback from the user, the quality of the dialogue system can be continuously improved.

[1775] 1. "Input information" refers to information such as name, email address, and voice profile that a user enters into the system.

[1776] 2. "Storage means" refers to a method or device for storing input information or conversation logs in a database or memory.

[1777] 3. "Means for starting a conversation" refers to the buttons or interface that a user operates to start a conversation.

[1778] 4. "Generative AI" refers to machine learning models and algorithms that generate optimal responses based on user input and emotional data.

[1779] 5. "Voice Profile" means a voice configuration with specific voice characteristics (e.g., pitch, gender, timbre, etc.) selected by a user.

[1780] 6. "Speech synthesis means" means a technology or device for converting text data into speech based on a specified speech profile.

[1781] 7. "Playback means" refers to an output device such as a speaker or earphones that allows the user to hear the generated audio.

[1782] 8. "Conversation log" means a record of the content of a conversation between a user and a system, including emotional data.

[1783] 9. "Emotion recognition means" means technologies or algorithms for analyzing emotions from a user's voice or text data.

[1784] 10. "Feedback" refers to the satisfaction rating and suggestions for improvement provided by the user to the system after the conversation has ended.

[1785] MODE FOR CARRYING OUT THE INVENTION

[1786] This invention is a system that inputs and saves user information and combines a generative artificial intelligence (AI) and an emotion engine to provide virtual conversations with the user. This system plays back responses aloud based on a voice profile selected by the user, realizing personalized communication. It also has the ability to analyze conversation logs and improve the performance of the generative AI. Furthermore, the emotion engine can recognize the user's emotions and generate optimal responses based on those emotions.

[1787] Hardware and Software Configuration

[1788] Server: The server stores user information, runs generative AI, analyzes emotion data, generates responses, synthesizes responses, stores and analyzes conversation logs, and analyzes feedback. Specifically, it uses a database management system (e.g., MySQL or MongoDB), generative AI (e.g., OpenAI's GPT-4), emotion engine (e.g., IBM Watson Tone Analyzer), and speech synthesis engine (e.g., Amazon Polly).

[1789] Terminal: The terminal inputs user information, receives and converts voice input to text, analyzes emotions in the voice data, plays back replies, and inputs and sends feedback information. Specifically, it uses a voice recognition engine (e.g., Google Cloud Speech-to-Text API).

[1790] User: The user enters information or speaks through the application and receives spoken responses provided by the system.

[1791] Specific processing of the system

[1792] 1. Entering and saving user information: The user launches the application and enters information such as their name, email address, preferred voice type, etc. The device sends the entered information to the server, which stores it in a database.

[1793] 2. Starting a conversation and recognizing emotions: The user presses the conversation start button to prepare for the conversation. The device passes the user's voice input to the speech recognition engine, converts it into text, and sends it to the server. At the same time, the voice data is passed to the emotion engine for emotion analysis.

[1794] 3. Emotion-based response generation: The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[1795] For example, a prompt such as "The user says he is tired. Please generate the best response" is sent to the generative artificial intelligence.

[1796] 4. Response voice synthesis and playback: The server passes the generated response text to a voice synthesis engine, which generates an audio file based on the voice profile selected by the user. The device receives the generated audio file from the server and plays it back to the user, allowing the user to experience a conversation with a virtual person.

[1797] 5. Saving and analyzing conversation logs, receiving and analyzing feedback: The server saves the content and emotional data of each conversation session as a log and uses it as training data for the generative AI. After the conversation ends, the user enters feedback and sends it to the server via their device. The server analyzes the feedback information and reflects it in the training data for the generative AI to improve the quality of the next conversation.

[1798] Specific examples

[1799] Stress relief conversations:

[1800] The user speaks "I'm very tired today."

[1801] The device converts the speech into text using a voice recognition engine and sends the text data, such as "I'm tired today," to the server.

[1802] The server detects the "tired" emotion using an emotion engine and sends a prompt to the generative AI model.

[1803] The server generates a text response such as "You seem tired today. Want to think of a way to relax?"

[1804] The server uses a voice synthesis engine to convert the voice into a gentle female voice.

[1805] The terminal plays back the generated voice, and the user enjoys the dialogue to reduce stress.

[1806] Through the above process, the system can recognize the user's emotions and generate optimal responses based on those emotions, thereby providing a personalized communication experience and increasing user satisfaction.

[1807] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1808] Step 1: Enter and save user information

[1809] A user launches the application and enters information such as their name, email address, and preferred voice type.

[1810] Input: User information such as name, email address, and voice profile.

[1811] The terminal receives the entered user information and transmits this information to the server.

[1812] Output: The user information sent to the server.

[1813] The server stores the received user information in a database, typically using a database management system such as MySQL or MongoDB.

[1814] Output: User information stored in the database.

[1815] Step 2: Initiating a conversation and recognizing emotions

[1816] The user presses the start conversation button in the application and is ready to start the conversation.

[1817] The terminal receives the user's voice input and converts it into text using a speech recognition engine.

[1818] Input: User's voice input (e.g. "I'm tired today").

[1819] Output: The audio data that is passed to the speech recognition engine.

[1820] The device sends the generated text data to the server and simultaneously passes the voice data to the emotion engine for emotion analysis. For example, the device generates text data such as "I'm tired today" using the Google Cloud Speech-to-Text API.

[1821] Input: Text data and audio data.

[1822] Output: The audio data passed to the emotion engine.

[1823] The server analyzes emotions from the received voice data using an emotion engine, for example, IBM Watson Tone Analyzer.

[1824] Output: Parsed emotion data.

[1825] Step 3: Emotion-based response generation

[1826] The server inputs the emotional data and text data obtained from the emotion engine into a generative artificial intelligence system to generate the optimal response.

[1827] Input: Emotion data and text data.

[1828] Output: The prompt input to the generative AI model (e.g., "The user says they are tired. Please generate the best response.").

[1829] The server receives responses from a generative AI model, such as OpenAI's GPT-4, to generate responses like, "You seem tired today. Want to know how to relax?"

[1830] Output: The generated response text.

[1831] Step 4: Synthesize and play the response

[1832] The server passes the generated response text to a speech synthesis engine, which generates an audio file based on the voice profile selected by the user.

[1833] Input: Response text and voice profile.

[1834] Output: The data that is passed to the speech synthesis engine.

[1835] The server sends the generated audio file to the terminal using a speech synthesis engine (e.g., Amazon Polly).

[1836] Output: The generated audio file.

[1837] The terminal receives the generated audio file and plays it for the user.

[1838] Output: The audio that is played to the user.

[1839] Step 5: Save and analyze conversation logs, receive and analyze feedback

[1840] The server stores the content and emotional data of each conversation session as a log, for example, in MongoDB.

[1841] Input: Conversational content and emotion data.

[1842] Output: Saved conversation logs.

[1843] The server uses the stored log data to train the generative artificial intelligence.

[1844] Input: Saved conversation logs.

[1845] Output: Conversation logs used as training data.

[1846] After the conversation, the user enters their satisfaction level and feedback.

[1847] Input: Satisfaction rating and suggestions for improvement.

[1848] The terminal transmits the feedback information input by the user to the server.

[1849] Output: Feedback information sent to the server.

[1850] The server analyzes the received feedback information and reflects it in the training data for the generative artificial intelligence.

[1851] Input: Feedback information.

[1852] Output: An improved generative artificial intelligence model.

[1853] Through these processing steps, the system can recognize the user's emotions and generate optimal responses, providing a personalized communication experience. It can also utilize feedback to improve the system's performance and continuously increase user satisfaction.

[1854] (Application example 2)

[1855] 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."

[1856] Ensuring personal safety and security is becoming increasingly important in modern society. However, traditional security measures are generally implemented by administrators or experts, and lack personalized, interactive security measures that reflect individual needs and feelings. Therefore, a system is needed that allows users to feel reassured in specific situations and receive appropriate advice on implementing specific security measures.

[1857] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1858] In this invention, the server includes means for inputting user information, means for saving the user information, means for the user to start a conversation, means for generating a response based on the user's words using a generative artificial intelligence, means for converting the generated response into speech in a selected voice, means for playing the speech to the user, means for saving and analyzing a conversation log, means for training the generative artificial intelligence, means for receiving and analyzing feedback from the user, means for improving the generative artificial intelligence based on the feedback, a visual display device worn by the user, a speech recognition device for acquiring the user's speech, means for converting the acquired speech into text data, means for analyzing the text data using an emotion recognition engine, means for generating an optimal response based on the analyzed emotion data, and means for providing audio security information to the user. This allows the user to receive a sense of security and appropriate advice for implementing specific security measures through a dialogue that reflects their individual emotional state and security concerns in real time.

[1859] "User Information" refers to data about a user, such as their name, email address, and preferred voice type.

[1860] "Means for inputting user information" refers to a function that allows a user to input their own information through a smart device.

[1861] "Means for saving user information" refers to a function for storing input user data in a storage device such as a database.

[1862] "Means for a user to start a conversation" refers to an interface that allows a user to start a conversation with the system using voice or gestures.

[1863] "Generative AI" refers to an AI system that generates optimal responses based on user input data.

[1864] "Means for converting the generated response to speech in a selected voice" refers to functionality for converting the generated text response to speech using a voice profile selected by the user.

[1865] "Means for playing audio to the user" refers to a function for playing the generated audio through a speaker or the like so that the user can hear it.

[1866] "Means for saving and analyzing conversation logs" refers to a function for recording the content of conversations with users and analyzing that data.

[1867] "Means for training generative AI" refers to a learning function that uses collected data to improve the performance of generative AI.

[1868] "Means for receiving and analyzing feedback from users" refers to a function for collecting opinions and evaluations provided by users to the system and analyzing them.

[1869] "Means for improving generative AI based on feedback" refers to the function of improving the algorithms and response quality of generative AI based on analyzed feedback data.

[1870] "A visual display device worn by a user" refers to a display device such as smart glasses or a head-mounted display used by a user.

[1871] The "voice recognition device for acquiring the user's voice" refers to a microphone or voice input device for picking up the user's voice.

[1872] The "means for converting acquired voice into text data" refers to a function for converting voice data into text data using a voice conversion engine.

[1873] "Means for analyzing text data based on an emotion recognition engine" refers to a function that evaluates and analyzes the emotional state of converted text data using an emotion analysis engine.

[1874] "Means for generating an optimal response based on analyzed emotion data" refers to a function that generates an optimal response taking into account the results of emotion recognition.

[1875] "Means for providing audio security information to the user" refers to a function for providing the generated response or security information to the user as audio.

[1876] The specific system for realizing this invention is an interactive security assistant system designed to provide users with safety and security. This system uses the following hardware and software to handle everything from inputting user information to generating responses and playing voice.

[1877] Hardware and software used

[1878] 1. Hardware

[1879] Smart glasses: When worn by the user, they function as a display and provide audio output and input.

[1880] Microphone: Built into the smart glasses and used to capture the user's voice.

[1881] Speaker: Built into the smart glasses, used to play generated responses to the user.

[1882] 2. Software

[1883] Speech recognition engine: For example, using the Google Speech-to-Text API, converts the user's voice into text data.

[1884] Emotion recognition engine: For example, use the Azure Cognitive Services Emotion API to analyze emotions based on text data.

[1885] Generative AI: For example, OpenAI GPT-4 is used to generate optimal responses based on user input and sentiment analysis results.

[1886] A speech synthesis engine: For example, using Amazon Polly to convert the generated responses into speech.

[1887] Processing Details

[1888] 1. Enter and save user information

[1889] The user puts on the smart glasses, launches the application, and enters their name, email address, preferred voice type, etc. This information is sent to the server through the smart glasses interface and stored in a database.

[1890] 2. Conversation initiation and emotion recognition

[1891] The user initiates a conversation using voice or touch gestures. The smart glasses' microphone picks up the user's voice and converts it into text using the Google Speech-to-Text API. This text is then sent to a server where an emotion recognition engine (such as the Azure Cognitive Services Emotion API) performs emotion analysis.

[1892] 3. Generating and playing optimal responses

[1893] Based on the emotional data and text data obtained from the emotion recognition engine, a generative artificial intelligence (e.g., OpenAI GPT-4) generates the optimal response. The generated response is passed to a speech synthesis engine (e.g., Amazon Polly), which converts it into an audio file using the voice profile selected by the user. The generated audio is played back from the smartglasses' speakers, allowing the user to experience the dialogue.

[1894] 4. Saving and analyzing conversation logs

[1895] The conversation content and emotional data are stored on the server and used as a dataset to learn the user's tendencies, which improves the quality of the next conversation. Users also provide feedback after the conversation, which the server analyzes and reflects in training the generative AI.

[1896] Specific prompt examples

[1897] For example, if a user says, "I'm at home now, but I'm a little worried about my safety," the system will generate the following optimal response:

[1898] "If a user voice-inputs a security-related question like, 'I'm at home now, but I'm a little worried about my safety,' how would you provide a sense of security? Considering the emotional data, what response should your generative AI model give? Think of five specific example responses."

[1899] In this way, it is possible to provide users with a sense of security and appropriate advice for implementing specific security measures through dialogue that reflects their individual emotional state and security concerns in real time.

[1900] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1901] Step 1:

[1902] The user puts on the smart glasses and launches the application.

[1903] Specific operation: The application is launched in the smart glasses, and first an interface is displayed for the user to enter user information such as name, email address, preferred voice type, etc. The user manually enters this information and presses the send button.

[1904] Input: User information such as name, email address, and preferred voice type.

[1905] Output: The entered user information is sent to the server and stored in the database.

[1906] Step 2:

[1907] The user presses the start conversation button to prepare for the conversation. The microphone in the smart glasses picks up the user's voice and passes it to the speech recognition engine.

[1908] What it does: A user says, "I'm at home now, but I'm a little worried about my safety." The microphone captures this speech, and the Google Speech-to-Text API converts it into text data, which is then sent to the server.

[1909] Input: Speech input from the user.

[1910] Output: Text data is generated by the Google Speech-to-Text API and sent to the server.

[1911] Step 3:

[1912] The server passes the generated text data to an emotion recognition engine to analyze the emotion.

[1913] How it works: The server passes text data to the Azure Cognitive Services Emotion API and requests emotion recognition. The emotion recognition engine analyzes the user's emotions based on the text data and returns the results to the server.

[1914] Input: Text data converted from audio.

[1915] Output: Emotion data as emotion recognition results.

[1916] Step 4:

[1917] The server uses generative artificial intelligence to generate the optimal response based on emotional data and text data.

[1918] How it works: The server instructs OpenAI GPT-4 to generate the optimal response based on emotional and text data. The generative AI analyzes the data and generates an appropriate response text. For example, "Please double-check the area around your house. Are the front door and windows securely closed?"

[1919] Input: Emotion data and text data.

[1920] Output: The generated response text data.

[1921] Step 5:

[1922] The server passes the generated response text to a speech synthesis engine, which generates an audio file using the voice profile selected by the user.

[1923] What happens: The server asks Amazon Polly to convert the response text data into an audio file. Amazon Polly generates the audio file using the voice profile selected by the user and returns the file to the server.

[1924] Input: Generated response text data.

[1925] Output: The generated audio file based on the voice profile selected by the user.

[1926] Step 6:

[1927] The server sends the generated audio file to the smart glasses for playback.

[1928] What happens: The server sends an audio file to the smart glasses, and the smart glasses' speaker plays the audio. The user hears a message saying, "Please double-check your surroundings. Are the front door and windows securely closed?"

[1929] Input: The generated audio file.

[1930] Output: The sound is played and heard by the user.

[1931] Step 7:

[1932] The server stores and analyzes conversation logs, receives user feedback, and uses it to improve the generative AI.

[1933] How it works: The server records the conversation and emotional data and stores it in a database. After the conversation ends, the user is presented with an interface to input their satisfaction and feedback. The input feedback is sent to the server, analyzed, and used as training data for the generative AI.

[1934] Input: Conversation logs, user feedback.

[1935] Output: Improved performance of generative artificial intelligence as a result of analysis.

[1936] 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.

[1937] 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.

[1938] 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 robot 414.

[1939] 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.

[1940] 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.

[1941] 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.

[1942] 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).

[1943] 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.

[1944] 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."

[1945] 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.

[1946] 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).

[1947] 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.

[1948] 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.

[1949] 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.

[1950] 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.

[1951] 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.

[1952] 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.

[1953] 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.

[1954] 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.

[1955] 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.

[1956] 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.

[1957] The following is further disclosed regarding the above embodiment.

[1958] (Claim 1)

[1959] a means for inputting user information;

[1960] a means for storing user information;

[1961] a means for a user to initiate a conversation;

[1962] A means for generating a response in response to a user's words using generative artificial intelligence;

[1963] means for converting the generated respons...

Claims

1. a means for inputting user information; a means for storing user information; a means for a user to initiate a conversation; A means for generating a response in response to a user's words using generative artificial intelligence; means for converting the generated response into speech in a selected voice; means for playing audio to a user; A means of storing and analyzing conversation logs; A means for training the generative artificial intelligence; and means for receiving and analyzing feedback from users; A way to improve generative artificial intelligence using feedback A system including:

2. 10. The system of claim 1, wherein the generated response is voice synthesized based on a user-selected voice profile.

3. 10. The system of claim 1, further comprising a generative artificial intelligence that generates optimal responses based on the user's past conversation history and current language.

Images