system

Abstract

Provide a system. A device for receiving voice input; means for converting voice input into text data; A means for including a generative AI for analyzing text data in a server; A means for receiving the analysis results from the server; A means for converting the analysis results into audio data; means for providing audio data to a user; A system including:

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] Conventional interactive robots often have difficulty in communicating with users, particularly due to low speech recognition accuracy and inappropriate responses. Furthermore, they face challenges in efficiently analyzing user-entered information and generating appropriate responses. Therefore, there is a need for systems that can respond quickly and accurately to user requests. [Means for solving the problem]

[0005] The present invention provides a system including a means for receiving voice input, a means for converting the voice input into text data, a means for including a generative AI in a server for analyzing the text data, a means for receiving analysis results from the server, a means for converting the analysis results into voice data, and a means for providing the voice data to a user. This allows the user's voice input to be efficiently converted into text data, which is then analyzed by the generative AI on the server to generate an appropriate response. The generated response is provided to the user as voice data, enabling smooth dialogue and guidance.

[0006] A "means for receiving voice input" is a device or method for capturing voice information from a user and incorporating it into the system.

[0007] "Means for converting voice input into text data" refers to a device or method for analyzing captured voice information and converting it into text format data.

[0008] "Means for including generative AI in a server" refers to a device or method for incorporating artificial intelligence technology into a server to analyze received text data and generate an appropriate response.

[0009] The "means for receiving the analysis result from the server" refers to a device or method for transferring the text response generated by the server to the terminal and receiving it.

[0010] The "means for converting the analysis result into voice data" refers to a device or method for converting the text response received from the server into voice data.

[0011] The "means for providing audio data to the user" refers to a device or method for allowing the user to listen to the generated audio data.

[0012] A "voice recognition engine" is a software or hardware technology that analyzes voice input and converts it into text data.

[0013] A "speech synthesis engine" is a software or hardware technology that converts text data into speech and provides it to the user.

[0014] A "terminal" is a device for interacting with a user, and is equipped with a microphone, a speaker, a voice recognition engine, and a voice synthesis engine. [Brief explanation of the drawings]

[0015] [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

[0016] 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.

[0017] First, the terms used in the following description will be explained.

[0018] 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).

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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."

[0036] This invention relates to a robot equipped with generative AI that provides information through dialogue with a user. This system receives voice input from a user, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, and ultimately provides a voice response to the user.

[0037] Specifically, it is implemented in the following manner.

[0038] composition

[0039] 1. Terminal

[0040] Microphone: Captures audio input from the user.

[0041] Speech recognition engine: Converts captured voice input into text data.

[0042] Communication module: Transmits the converted text data to the server.

[0043] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[0044] Speaker: Provides audio data to the user.

[0045] 2. Server

[0046] Analysis module: Analyzes the received text data.

[0047] Generative AI: Generates appropriate responses based on analysis results.

[0048] Program processing and explanation

[0049] Receiving voice input

[0050] Terminal

[0051] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[0052] Converting audio data to text

[0053] Terminal

[0054] The captured voice data is sent to a voice recognition engine, which converts the voice into text data such as "Please tell me about nearby restaurants."

[0055] Sending text data

[0056] Communication Module

[0057] The converted text data is sent to the server through the communication module.

[0058] Analysis of text data and response generation using generative AI

[0059] server

[0060] The server uses an analysis module to analyze the received text data and determine that the user is inquiring about nearby restaurants. The generative AI then generates a response based on the analysis results, such as "I will list restaurants within 1 km of here."

[0061] Sending response data

[0062] server

[0063] The generated response is sent from the server to the terminal.

[0064] Converting response data to audio

[0065] Terminal

[0066] The received text response is sent to a speech synthesis engine and converted into voice data.

[0067] Voice response to the user

[0068] Terminal

[0069] Finally, the speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[0070] Specific examples

[0071] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[0072] 1. Voice Input

[0073] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[0074] 2. Speech-to-text

[0075] The device converts the speech to text, generating the text "Can you tell me where to find a restaurant nearby?"

[0076] 3. Sending text data

[0077] The converted text data is sent to the server.

[0078] 4. Parsing and Response Generation

[0079] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[0080] 5. Transcribing and delivering responses

[0081] The server sends a response to the device, which converts it into voice and provides it to the user. The robot then provides voice guidance, saying, "I'll list restaurants within 1 km of here."

[0082] Real-world usage scenarios

[0083] This system is expected to be used in shopping malls, airports, hotels, etc. For example, if it is installed as a guide robot in a shopping mall, when a visitor asks, "Where is the restroom?", the system will provide quick and accurate guidance.

[0084] The present invention allows users to quickly obtain accurate information through natural voice dialogue, contributing to improved user satisfaction.

[0085] The processing flow will be explained below.

[0086] Step 1:

[0087] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[0088] Step 2:

[0089] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[0090] Step 3:

[0091] The device sends the converted text data to the server via the communication module, and the text data is sent as an HTTP request to the server's API endpoint.

[0092] Step 4:

[0093] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[0094] Step 5:

[0095] The server issues instructions to the generative AI based on the analysis results, and the generative AI uses the analysis results to generate a response text such as "I will list restaurants within 1 km of here."

[0096] Step 6:

[0097] The server sends the generated response text to the terminal as an HTTP response, and the generated text data is returned to the terminal via the communication module.

[0098] Step 7:

[0099] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech generated is, "I'll list restaurants within 1 km of here."

[0100] Step 8:

[0101] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here."

[0102] Example 1

[0103] 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."

[0104] Conventional voice dialogue systems have insufficient speech recognition accuracy and response generation quality, making it difficult for users to have smooth and natural dialogue. In addition, the overall system processing speed is slow, making it difficult to provide instant responses. This has led to problems that degrade the user experience.

[0105] 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.

[0106] In this invention, the server includes means for analyzing text data using generative artificial intelligence to generate a response, means for transmitting the generated response data to the terminal, and means for the terminal to convert the response into voice data and provide the user with the response. This makes it possible to streamline the process from voice input to response generation and provide the user with a quick and accurate response.

[0107] The "means for receiving voice input" is a means for capturing voice from a user using a device such as a microphone.

[0108] The "means for converting speech input into text data" refers to a means including a speech recognition engine that analyzes received speech and converts it into a corresponding text format.

[0109] The "means for transmitting text data to a server" refers to means including a communication module for transmitting the converted text data to a server via a network.

[0110] "Means for analyzing text data and generating responses using generative artificial intelligence" refers to means for analyzing text data on a server using natural language processing technology and generating appropriate responses using generative artificial intelligence.

[0111] The "means for receiving response data from the server" refers to means including a communication module for receiving the generated response data from the server to the terminal.

[0112] The "means for converting received response data into voice data" refers to a means for converting received response data in text format into voice data using a voice synthesis engine.

[0113] The "means for providing audio data to the user" refers to a means for reproducing the converted audio data to the user through a speaker.

[0114] The present invention relates to a system equipped with generative artificial intelligence (AI) that provides information through dialogue with a user. Specifically, the system receives voice input, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, ultimately providing a voice response to the user.

[0115] System configuration

[0116] Terminal

[0117] Microphone: Responsible for capturing audio input from the user.

[0118] Speech recognition engine: Converts the captured voice into text data. Here, a common cloud service can be used as the speech recognition engine. Examples include Google® Cloud Speech-to-Text and Microsoft® Azure® Speech Services.

[0119] Communication module: Transmits the converted text data to the server using an internet connection, including a Wi-Fi module and a 4G / 5G module.

[0120] Speech synthesis engine: Converts the response received from the server into voice data. For voice synthesis, Amazon Polly or Google Cloud Text-to-Speech are used.

[0121] Speaker: Used to provide audio data to the user.

[0122] server

[0123] Analysis module: Used to analyze text data sent from the device, using NLP libraries (e.g., SpaCy or NLTK) to understand the user's intent.

[0124] Generative AI: Generates appropriate responses based on the analysis results. For this purpose, OpenAI's GPT-3 (registered trademark) and BERT are used as generative AI.

[0125] Specific operation example

[0126] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the system operates as follows:

[0127] 1. Receiving voice input

[0128] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by a microphone.

[0129] 2. Converting voice data to text

[0130] The captured voice data is sent to a voice recognition engine (for example, Google Cloud Speech-to-Text) and converted into text data such as "Please tell me where to find a nearby restaurant."

[0131] 3. Sending text data

[0132] The converted text data is sent to the server via a communication module, which uses an internet connection.

[0133] 4. Text Data Analysis and Response Generation

[0134] The server analyzes the received text data using an analysis module (e.g., SpaCy) to determine that the user is inquiring about nearby restaurants. Next, a generative AI (e.g., GPT-3) generates a response based on the analysis results, such as "I'll list restaurants within 1 km of here."

[0135] 5. Sending response data

[0136] The generated response is sent from the server to the terminal, typically using HTTPS as the communication protocol.

[0137] 6. Speech conversion of response data

[0138] The device sends the received text response to a speech synthesis engine (e.g., Amazon Polly), which converts it into voice data.

[0139] 7. Audio response to the user

[0140] The speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[0141] Prompt Sentence Examples

[0142] "User said: 'What restaurants are nearby?'"

[0143] In this way, the system enables users to quickly and accurately obtain information through natural voice interaction.

[0144] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0145] Step 1:

[0146] Receiving voice input

[0147] Subject: Terminal

[0148] What it does: The device's microphone captures audio input from the user.

[0149] Input: User speaks to the robot, "What restaurants are nearby?"

[0150] Output: The captured audio data.

[0151] What it does: The microphone converts your voice into an electronic signal and stores that data in temporary memory.

[0152] Step 2:

[0153] Converting audio data to text

[0154] Subject: Terminal

[0155] How it works: Captured voice data is sent to a speech recognition engine, which analyzes the voice data and converts it into corresponding text.

[0156] Input: The captured audio data.

[0157] Output: Text data: "Please tell me where to find a restaurant nearby."

[0158] How it works: The speech recognition engine converts the audio signal into the time-frequency domain and parses it into text. Google Cloud Speech-to-Text can be used as the speech recognition engine.

[0159] Step 3:

[0160] Sending text data

[0161] Subject: Terminal

[0162] Operation: The converted text data is sent to the server through the communication module.

[0163] Input: The converted text data.

[0164] Output: The text data sent to the server.

[0165] Specific operation: Uses Wi-Fi module and 4G / 5G module to send data to the server via HTTPS protocol.

[0166] Step 4:

[0167] Analysis of text data and response generation using generative AI

[0168] Subject: Server

[0169] How it works: The server analyzes the received text data using the analysis module. Based on the analysis results, the generative AI generates an appropriate response.

[0170] Input: Text data sent to the server.

[0171] Output: The response "List of restaurants within 1km of here."

[0172] How it works: The server uses SpaCy, an NLP library, to parse the text data, then uses a generative AI model such as OpenAI's GPT-3 to generate a response.

[0173] Step 5:

[0174] Sending response data

[0175] Subject: Server

[0176] Operation: The generated response data is sent from the server to the device.

[0177] Input: The text data generated as a response.

[0178] Output: The response data sent to the device.

[0179] Specific operation: The server uses the HTTPS protocol to send the generated response data to the terminal.

[0180] Step 6:

[0181] Converting response data to audio

[0182] Subject: Terminal

[0183] How it works: The received text response is sent to a speech synthesis engine and converted into audio data.

[0184] Input: The text format of the response data received from the server.

[0185] Output: Audio data.

[0186] Specific operation: Uses Amazon Polly as a speech synthesis engine to convert text to speech.

[0187] Step 7:

[0188] Voice response to the user

[0189] Subject: Terminal

[0190] Action: A voice response is played through the speaker and provided to the user.

[0191] Input: Speech data generated by the speech synthesis engine.

[0192] Output: The audio response provided to the user.

[0193] Specific operation: The speaker plays the audio data and provides it to the user as sound.

[0194] (Application example 1)

[0195] 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."

[0196] Conventional information guidance systems in brick-and-mortar stores have the problem of being unable to provide immediate and accurate responses to specific product or location information that users are looking for. Particularly in large stores with many products, users often experience inconvenience due to being unable to quickly obtain information. To solve this problem, an efficient system that provides real-time information based on the user's voice input is needed.

[0197] 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.

[0198] In this invention, the server includes means for receiving voice input, means for converting the voice input into text data, means for including in the server a generative AI for analyzing the text data, means for receiving analysis results from the server, means for converting the analysis results into voice data, means for providing the voice data to the user, and means for providing voice guidance about store information based on user questions in the physical store. This enables users to obtain information they desire in the physical store, such as product locations and sale information, in real time, improving the user experience.

[0199] A "means for receiving voice input" is a device or sensor for detecting a user's voice, including a microphone and related hardware.

[0200] "Means for converting voice input into text data" refers to a voice recognition engine that converts voice into text, thereby converting voice data into character data.

[0201] "Means including a server with a generative AI for analyzing text data" refers to a server equipped with artificial intelligence that receives text data, analyzes its contents, and generates an appropriate response.

[0202] "Means for receiving analysis results from the server" refers to a communication module or related software for receiving the analysis results generated by the server.

[0203] "Means for converting the analysis results into voice data" refers to a voice synthesis engine that converts the text data of the analysis results into voice data.

[0204] The "means for providing audio data to the user" refers to an output device such as a speaker for reproducing the converted audio data and providing it to the user.

[0205] "Means for providing voice guidance on in-store information based on questions posed by users in a physical store" refers to a system that includes generative AI and communication means for providing voice guidance on in-store information, such as product locations and sale information, in response to voice questions posed by users in a physical store.

[0206] The present invention relates to a system that provides information desired by a user in a physical store through voice dialogue. This system performs voice input, voice recognition, text analysis, response generation using generative AI, and voice synthesis to provide a voice response. An embodiment of the present invention will be described in detail below.

[0207] Basic configuration

[0208] 1. Voice Input

[0209] The system uses a microphone to receive the user's voice. When the user asks the robot, "Where is the toilet paper counter?", the voice is captured by the microphone.

[0210] 2. Voice Recognition

[0211] The voice data captured by the microphone is sent to the device's speech recognition engine, which analyzes the voice data and converts it into text data. Specifically, the speech recognition engine uses existing technologies such as the Google Speech Recognition API.

[0212] 3. Sending text data

[0213] The converted text data is sent to the server via the communication module, using an HTTP POST request.

[0214] 4. Text Analysis and Response Generation

[0215] The server receives the text data and analyzes it using an analysis module. Based on the analysis results, a generative AI generates an optimal response. The generative AI can use an advanced natural language processing model such as GPT-3.

[0216] 5. Sending response data

[0217] The generated response text is sent to the terminal again through the communication module.

[0218] 6. Speech Synthesis

[0219] The response text sent to the device is converted into voice data by a speech synthesis engine, using technologies such as Google Text-to-Speech (gTTS).

[0220] 7. Audio Provision

[0221] Finally, the voice data is sent to the user through a speaker, and the robot announces, "Toilet paper is sold at Aisle 5."

[0222] Equipment and software used

[0223] Hardware: microphone, speaker, communication module

[0224] software:

[0225] SpeechRecognition (Python library)

[0226] gTTS (Google Text-to-Speech)

[0227] requests (a Python library for sending HTTP requests)

[0228] Advanced natural language processing models (generative AI, e.g. GPT-3)

[0229] Specific examples

[0230] For example, if a user says "Tell me what products are on sale," the following happens:

[0231] 1. The user asks the robot verbally, "Please tell me what products are on sale."

[0232] 2. The device converts the speech to text, generating the text "Tell me what items are on sale."

[0233] 3. This text data is sent to the server.

[0234] 4. The server analyzes the text data, and the generative AI generates a response such as, "Here are the items currently on sale:..."

[0235] 5. The server sends a response to the device, which converts it into audio and provides it to the user.

[0236] Example prompt sentence:

[0237] "Where is the toilet paper section?"

[0238] "Please tell me what products are on sale."

[0239] In this way, the present invention can efficiently and quickly provide information that users desire in a physical store, thereby improving the user experience.

[0240] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0241] Step 1:

[0242] The user provides voice input. The user asks a question to the robot in the physical store, for example, "Where is the toilet paper section?" This voice input is captured by the microphone.

[0243] Input: User's voice

[0244] Output: Audio data captured by the microphone

[0245] Step 2:

[0246] The device sends the captured voice data to a speech recognition engine, which converts the voice data into text using a service such as the Google Speech Recognition API.

[0247] Input: Captured audio data

[0248] Output: Transformed text data (e.g., "Where is the toilet paper counter?")

[0249] Step 3:

[0250] The terminal sends the converted text data to the server via the communication module, using a standard HTTP POST request.

[0251] Input: Converted text data

[0252] Output: Text data sent to the server

[0253] Step 4:

[0254] The server receives the text data and analyzes it using an analysis module. A generative AI, such as GPT-3, generates an appropriate response based on the analysis results, such as "The toilet paper section is on Aisle 5."

[0255] Input: Text data received by the server

[0256] Data processing / data calculation: Text analysis and response generation using generative AI

[0257] Output: The generated response text (e.g., "Toilet paper is available on Aisle 5")

[0258] Step 5:

[0259] The server then sends the generated response text to the terminal via the communication module, again using an HTTP POST request.

[0260] Input: Generated response text

[0261] Output: Response text sent to the terminal

[0262] Step 6:

[0263] The device sends the received response text to a speech synthesis engine, which converts the text data into voice data using a service such as Google Text-to-Speech (gTTS).

[0264] Input: Received response text

[0265] Output: Speech data (e.g. "Toilet paper is on Aisle 5")

[0266] Step 7:

[0267] The terminal provides the user with voice data through a speaker, and ultimately the user can hear voice guidance from the robot.

[0268] Input: Audio data

[0269] Output: A spoken response played through the speaker (e.g., "Toilet paper is on Aisle 5")

[0270] In this way, the data input and output are clearly defined at each step, making it a system that allows users to quickly obtain useful information.

[0271] 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.

[0272] This invention combines a robot equipped with generative AI that provides information through dialogue with the user with an emotion engine that recognizes the user's emotions. This system performs voice input, voice recognition, text analysis, response generation using generative AI, emotion recognition, and voice synthesis, and ultimately provides a voice response to the user. The specific form is shown below.

[0273] composition

[0274] 1. Terminal

[0275] Microphone: Captures audio input from the user.

[0276] Speech recognition engine: Converts captured voice input into text data.

[0277] Emotion engine: Analyzes user emotions from voice input.

[0278] Communication module: Sends text data and emotion information to the server.

[0279] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[0280] Speaker: Provides audio data to the user.

[0281] 2. Server

[0282] Analysis module: Analyzes the received text data.

[0283] Generative AI: Generates appropriate responses based on analysis results.

[0284] Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state.

[0285] Program processing and explanation

[0286] Receiving voice input

[0287] Terminal

[0288] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[0289] Speech data text conversion and sentiment analysis

[0290] Terminal

[0291] The captured voice data is sent to a speech recognition engine, which converts the voice into text data such as "Please tell me where to find a nearby restaurant." At the same time, an emotion engine analyzes the voice data and identifies the user's emotional state (e.g., happy, worried, etc.).

[0292] Sending text data and emotional information

[0293] Communication Module

[0294] The converted text data and emotion information are sent to the server via the communication module. The text data and emotion information are sent as an HTTP request to the server's API endpoint.

[0295] Analysis of text data and response generation using generative AI

[0296] server

[0297] The server uses an analysis module to analyze the received text data and determine the user's intent (whether they want to know about nearby restaurants). Next, the generative AI generates a response text based on the analysis results, such as "I will list restaurants within 1 km of here."

[0298] Response adjustment based on emotional information

[0299] server

[0300] The generated response text is passed to an emotion adjustment module, which adjusts it based on the user's emotional state. For example, if the user is recognized as distressed, the response might be adjusted to "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0301] Sending adjusted response data

[0302] server

[0303] The adjusted response text is sent to the terminal as an HTTP response.

[0304] Converting response data to audio

[0305] Terminal

[0306] The received text response is sent to a speech synthesis engine, which converts it into speech, generating a voice that says, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0307] Voice response to the user

[0308] Terminal

[0309] Finally, the speaker provides the user with a voice response saying, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[0310] Specific examples

[0311] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[0312] 1. Voice Input

[0313] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[0314] 2. Speech-to-text and emotion recognition

[0315] The device converts the speech into text and simultaneously recognizes the user's emotions. The text "Please tell me about nearby restaurants" and the emotional information "I'm in trouble" are generated.

[0316] 3. Data transmission

[0317] The converted text data and emotion information are sent to the server.

[0318] 4. Parsing and Response Generation

[0319] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[0320] 5. Response Adjustment

[0321] The emotion regulation module adjusts the response to, "I'll list restaurants within 1km of here. Let me know if I can help you."

[0322] 6. Transcribing and providing responses

[0323] The adjusted response is sent to the device, where a speech synthesis engine converts it into audio and provides it to the user. The robot then provides a voice guide, saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[0324] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to their emotions.

[0325] The processing flow will be explained below.

[0326] Step 1:

[0327] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[0328] Step 2:

[0329] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[0330] Step 3:

[0331] The device sends the text data to the emotion engine, which analyzes the voice data and identifies the user's emotional state as "troubled."

[0332] Step 4:

[0333] The device sends the converted text data and emotion information to the server via the communication module. The text data and emotion information are sent to the server's API endpoint as an HTTP request.

[0334] Step 5:

[0335] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[0336] Step 6:

[0337] The server issues instructions to the generative AI based on the analysis results, and the generative AI generates a response text such as "I will list restaurants within 1 km of here."

[0338] Step 7:

[0339] The server passes the generated response text to the emotion adjustment module, which adjusts the response based on the user's emotional state ("I'm in trouble"). For example, it generates an adjusted response text such as "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0340] Step 8:

[0341] The server sends the adjusted response text to the terminal as an HTTP response. The adjusted text data is returned to the terminal.

[0342] Step 9:

[0343] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech synthesis engine generates a voice saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[0344] Step 10:

[0345] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0346] Example 2

[0347] 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."

[0348] Conventional dialogue systems have difficulty providing responses that take into account not only the user's intentions but also their emotional state. As a result, users are often dissatisfied with the responses from the system, and there is a demand for more natural dialogue and accurate information provision.

[0349] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for analyzing text data using generative artificial intelligence and generating an appropriate response, a means for adjusting the generated response based on the user's emotions, and a means for receiving the adjusted analysis result. This makes it possible to provide natural and accurate information according to the user's emotional state.

[0350] "Voice input" is the process of capturing a user's spoken words or instructions with a device such as a microphone and recognizing them as digital signals.

[0351] "Means for converting into text data" refers to a technique or device that converts captured voice input into character string data.

[0352] "Means for recognizing emotions" refers to technologies or devices that analyze and determine a user's emotional state from voice or other input data.

[0353] "Transmitting means" refers to the communications technology or equipment that sends the converted and analyzed data from one point to another.

[0354] "Generative AI" is an AI technology that generates appropriate responses based on input data.

[0355] The "means for adjusting the response" refers to a technique or device that modifies the generated response to an appropriate form depending on the emotional state of the user.

[0356] The "means for receiving the analysis results" refers to the technology or device that inputs the analyzed data and generated responses sent from the server into the terminal.

[0357] "Means for converting into audio data" refers to technology or devices that convert character data into audio signals that can be heard by the user.

[0358] The "means for providing" refers to a technique or device for transmitting the generated audio data to the user through an output device such as a speaker.

[0359] This invention provides a system that enables natural dialogue between a user and a robot. The system's main components include voice input, voice recognition, emotion recognition, generative artificial intelligence (AI), emotion regulation, voice synthesis, and a communication module for processing and transmitting the respective data.

[0360] Hardware and software used

[0361] Terminal

[0362] Microphone: Captures audio input from the user.

[0363] Speech recognition engine: Converts captured voice input into text data, for example, using the Google Cloud Speech-to-Text API.

[0364] Emotion engine: Analyzes user emotions from voice input, for example using IBM Watson® Tone Analyzer.

[0365] Communication module: Sends text data and emotion information to the server.

[0366] Speech synthesis engine: Converts the analysis results received from the server into voice data. For example, use Amazon Polly.

[0367] Speaker: Provides audio data to the user.

[0368] server

[0369] Analysis module: Analyzes the received text data.

[0370] Generative artificial intelligence (AI): Generates appropriate responses based on analysis results, for example, using OpenAI GPT-3.

[0371] Emotion adjustment module: adjusts the generated response based on the user's emotional state.

[0372] Specific examples of processing

[0373] The system operates as follows.

[0374] 1. The user speaks to the robot: "Tell me about a nearby restaurant." The device's microphone captures this voice.

[0375] 2. Text conversion and emotion recognition: The captured voice data is sent to a speech recognition engine and converted into text data: "Please tell me where to find a nearby restaurant." At the same time, the emotion engine analyzes the voice data and identifies the emotion information, "I'm in trouble."

[0376] 3. Sending data to the server: The converted text data and emotion information are sent to the server through the communication module.

[0377] 4. Analysis of text data and generation of a response using generative AI: The analysis module on the server analyzes the received text data and determines the user's intent (want to know about nearby restaurants). Based on the analysis results, the generative AI generates a response text such as "I will list restaurants within 1 km of here." The following prompt sentence is used:

[0378] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[0379] 5. Response adjustment based on emotion information: The generated response text is passed to the emotion adjustment module, which adjusts the response based on the emotion tag “in need” to “I’ll list restaurants within 1km of here. Let me know if I can help you.”

[0380] 6. Sending adjusted response data: The adjusted response text is sent to the terminal again through the communication module.

[0381] 7. Speech synthesis and provision: The received response text data is sent to a speech synthesis engine, which converts it into voice data saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you." Finally, the voice is played over the speaker and provided to the user.

[0382] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to emotions.

[0383] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0384] Program processing flow

[0385] Step 1: Capturing Audio Input

[0386] Terminal

[0387] Input: The user says, "What restaurants are nearby?"

[0388] Processing: A microphone captures this audio and stores it as digital data.

[0389] Output: The user's voice is stored as digital audio data.

[0390] Specific actions

[0391] When a user speaks to the robot, their voice is converted into data by a microphone, and the data is temporarily stored in a buffer on the device.

[0392] Step 2: Converting speech data to text and analyzing sentiment

[0393] Terminal

[0394] Input: Captured digital audio data.

[0395] Processing: The speech recognition engine converts the voice data into text data, and the emotion engine analyzes the voice data to recognize the emotional state.

[0396] Output: Text data "Please tell me a nearby restaurant" and emotion information "I'm in trouble."

[0397] Specific actions

[0398] A speech recognition engine (e.g., Google Cloud Speech-to-Text API) converts the speech into text, generating the string "Please tell me where to find a nearby restaurant." At the same time, an emotion engine (e.g., IBM Watson Tone Analyzer) analyzes the speech data and generates an emotion tag such as "I'm in trouble."

[0399] Step 3: Sending text data and emotion information to the server

[0400] Communication Module

[0401] Input: Text data and emotion information.

[0402] Processing: Sends an HTTP POST request to the server.

[0403] Output: Text data and emotion information sent to the server.

[0404] Specific actions

[0405] The communication module generates an HTTP POST request containing text data and emotion information in the payload and sends it to the server's API endpoint.

[0406] POST / api / v1 / parse

[0407] Host: example.com

[0408] Content-Type: application / json

[0409] {

[0410] "text": "Can you tell me nearby restaurants?",

[0411] "emotion": "troubled"

[0412] }

[0413] Step 4: Analyzing text data and generating responses using AI

[0414] server

[0415] Input: Text data and emotion information.

[0416] Processing: The analytics module analyzes the text data to determine the user's intent, and the generative AI model generates an appropriate response.

[0417] Output: Response text "Listing restaurants within 1km of here."

[0418] Specific actions

[0419] The analysis module analyzes the text data and extracts the user's intent, which is "I want to know about nearby restaurants." The following prompt sentence is used to generate the response:

[0420] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[0421] The AI ​​model (e.g., OpenAI GPT-3) generates a response text such as "I will list restaurants within 1 km of here."

[0422] Step 5: Adjusting responses based on emotional information

[0423] server

[0424] Input: Response text and sentiment information.

[0425] Processing: The emotion regulation module adjusts the response based on the user's emotions.

[0426] Output: Tailored response text: "I'll list restaurants within 1km of here. Let me know if I can help you."

[0427] Specific actions

[0428] The emotion adjustment module adjusts the response based on the emotion tag "distressed," resulting in the final response: "I'll list restaurants within 1km of here. Let me know if I can help you."

[0429] Step 6: Sending adjusted response data to the device

[0430] server

[0431] Input: The adjusted response text.

[0432] Processing: Send to the terminal as an HTTP response.

[0433] Output: The adjusted response text sent to the terminal.

[0434] Specific actions

[0435] The server generates an HTTP response and sends it back to the device, including the adjusted response text.

[0436] HTTP / 1.1 200 OK

[0437] Content-Type: application / json

[0438] {

[0439] "response": "I'll list restaurants within 1km of here. Let me know if I can help you."

[0440] }

[0441] Step 7: Transcribing response data

[0442] Terminal

[0443] Input: The adjusted response text.

[0444] Processing: The speech synthesis engine converts the text data into speech data.

[0445] Output: Speech data saying "I'll list restaurants within 1km of here. Let me know if I can help you."

[0446] Specific actions

[0447] The adjusted response text is sent to a speech synthesis engine (e.g., Amazon Polly) to obtain the corresponding voice data, which is then saved as an audio file for playback.

[0448] Step 8: Audible response to the user

[0449] Terminal

[0450] Input: Audio data.

[0451] Processing: Outputs sound through the speaker.

[0452] Output: An audio response that the user can hear.

[0453] Specific actions

[0454] A voice comes through the speaker and tells the user, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[0455] (Application example 2)

[0456] 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."

[0457] Current voice input systems are unable to respond taking into account the user's emotional state, and therefore often provide inappropriate information to users. Furthermore, guidance services in brick-and-mortar stores are also required to understand what the user is looking for and provide appropriate information, but no system exists that can achieve this. This poses the challenge of making it difficult to improve the user experience and provide efficient guidance.

[0458] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means including a generative AI for analyzing text data, means for adjusting a response generated based on the user's emotional information, and means for receiving the analysis results from the server. This makes it possible to provide appropriate information that takes the user's emotions into consideration.

[0459] A "means for receiving voice input" is a device or function that captures voice from a user and converts it into input data.

[0460] The "means for converting speech input into text data" refers to a technology or device that converts received speech input into text data using speech recognition technology.

[0461] "Means for analyzing user emotions" refers to a technology or device that analyzes and identifies the user's emotional state from voice data or text data.

[0462] The "means for transmitting text data and emotion information to a server" is a communication device or function for transmitting the analyzed text data and emotion information to a server.

[0463] "Means including a generative AI on a server for analyzing text data" refers to a device or system that includes a generative AI function on a server for analyzing received text data and generating an appropriate response.

[0464] The "means for adjusting the response generated based on the user's emotional information" refers to a function or device that adjusts the response based on the analysis results in accordance with the user's emotional state.

[0465] The "means for receiving the analysis results from the server" is a communication device or function for receiving the analysis results provided by the server.

[0466] "Means for converting analysis results into audio data" refers to a technology or device that converts text-based responses into audio data.

[0467] The "means for providing audio data to the user" refers to a playback device or function such as a speaker or headphone that allows the user to listen to the converted audio data.

[0468] To implement this invention, a system is required that executes each step of voice input, voice recognition, emotion analysis, response generation using generative AI, response adjustment, and voice synthesis. The system consists of a terminal that interacts with the user and a server that analyzes data and generates responses.

[0469] Hardware and Software Configuration

[0470] Device:

[0471] 1. Microphone: Captures audio input from the user.

[0472] 2. Speech recognition engine: Converts captured voice into text data. This function uses the Google Speech-to-Text API.

[0473] 3. Emotion engine: Analyzes the user's emotions from voice input. This analysis uses the Microsoft Azure Emotion API.

[0474] 4. Communication module: Sends text data and emotion information to the server using HTTP and REST API as the communication protocol.

[0475] 5. Speech synthesis engine: Converts the analysis results received from the server into voice data. This function uses the Google Text-to-Speech API.

[0476] 6. Speaker: Provides the converted audio data to the user.

[0477] server:

[0478] 1. Analysis module: Analyzes the received text data using natural language processing technology.

[0479] 2. Generative AI: Generates appropriate responses based on the analysis results. The OpenAI GPT-3 model is used for generative AI.

[0480] 3. Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state. For example, if the user is in a distressed emotional state, the response will be changed to a more considerate one.

[0481] Data processing and calculation

[0482] The device receives voice data and converts it into text data using a voice recognition engine. At the same time, the emotion engine analyzes the emotional information. This data is then sent to the server via a communication module.

[0483] On the server, the analysis module analyzes the text data, and the generative AI generates a response. The generated response is adjusted by the emotion adjustment module based on the user's emotional state. The final response is sent back to the device.

[0484] The response text data returned to the terminal is converted into voice data by a voice synthesis engine and provided to the user through a speaker.

[0485] Specific examples

[0486] The user asks aloud, "Where is the recommended sweets corner?" This voice is captured by a microphone and converted into text data by a speech recognition engine. At the same time, an emotion engine analyzes the user's emotional state (distress). The text data and emotion information are sent to the server.

[0487] The server's analysis module analyzes the text data, and the generative AI generates an appropriate response. This response is then adjusted by the emotion adjustment module to, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions." This adjusted response is sent to the device, where it is converted into voice data by the voice synthesis engine. It is then provided to the user via the speaker.

[0488] Prompt Sentence Examples

[0489] "A customer is asking, 'Where's the recommended snack section?' Distressed emotion detected. Suggest an appropriate response."

[0490] The above is an embodiment of the present invention.

[0491] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0492] Step 1:

[0493] Receiving voice input

[0494] The user speaks into the microphone on their smartphone or smart glasses, asking, "Where is the recommended sweets corner?"

[0495] Input: User's voice data.

[0496] What it does: The microphone captures sound and stores it as audio data.

[0497] Output: The captured audio data.

[0498] Step 2:

[0499] Speech data text conversion and sentiment analysis

[0500] The device sends the captured voice data to a speech recognition engine to convert it into text data, while an emotion engine analyzes the voice data to identify the user's emotional state.

[0501] Input: Captured audio data.

[0502] How it works: The speech recognition engine converts the speech into text data such as "Where is the recommended snack corner?" The emotion engine analyzes the speech data and generates emotion information such as "I'm in trouble."

[0503] Output: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[0504] Step 3:

[0505] Sending text data and emotional information

[0506] The terminal transmits the converted text data and emotion information to the server via the communication module.

[0507] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[0508] Operation: The communication module sends text data and emotion information as an HTTP request to the server's API endpoint.

[0509] Output: Text data and emotion information sent to the server.

[0510] Step 4:

[0511] Analysis of text data and response generation using generative AI

[0512] The server uses an analysis module to analyze the received text data and determine the user's intent. Next, a generative AI generates an appropriate response based on the analysis results.

[0513] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[0514] How it works: The analysis module analyzes the text data and determines the user's intent (the user wants to know where the candy corner is). The generative AI generates a response text such as "The candy corner is in the center of the third floor."

[0515] Output: Response text "The candy corner is in the center of the third floor."

[0516] Step 5:

[0517] Response adjustment based on emotional information

[0518] The server passes the generated response text to an emotion adjustment module, which adjusts the response based on the user's emotional state.

[0519] Input: Emotion information "I'm in trouble" and response text "The candy corner is in the center of the third floor."

[0520] How it works: The emotion adjustment module adjusts the response text depending on the emotion, resulting in a final response of "The candy corner is in the center of the third floor. Let me know if you need any further directions."

[0521] Output: Adjusted response text: "The candy section is in the center of the third floor. Let me know if you need any further directions."

[0522] Step 6:

[0523] Sending adjusted response data

[0524] The server sends the adjusted response text to the terminal as an HTTP response.

[0525] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[0526] Operation: The server sends the response text to the terminal as an HTTP response.

[0527] Output: The adjusted response text sent to the terminal.

[0528] Step 7:

[0529] Converting response data to audio

[0530] The terminal sends the received response text to a speech synthesis engine and converts it into voice data.

[0531] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[0532] How it works: A speech synthesis engine converts text data into speech.

[0533] Output: Speech data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[0534] Step 8:

[0535] Voice response to the user

[0536] The terminal ultimately provides a voice response to the user through a speaker.

[0537] Input: Voice data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[0538] Action: The speaker plays the audio data and provides a response to the user.

[0539] Output: A voice description that says, "The sweets corner is in the center of the third floor. Let me know if you need any further directions."

[0540] 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.

[0541] 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.

[0542] 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.

[0543] [Second embodiment]

[0544] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0545] 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.

[0546] 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).

[0547] 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.

[0548] 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.

[0549] 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).

[0550] 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.

[0551] 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.

[0552] 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.

[0553] 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.

[0554] 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.

[0555] 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."

[0556] This invention relates to a robot equipped with generative AI that provides information through dialogue with a user. This system receives voice input from a user, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, and ultimately provides a voice response to the user.

[0557] Specifically, it is implemented in the following manner.

[0558] composition

[0559] 1. Terminal

[0560] Microphone: Captures audio input from the user.

[0561] Speech recognition engine: Converts captured voice input into text data.

[0562] Communication module: Transmits the converted text data to the server.

[0563] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[0564] Speaker: Provides audio data to the user.

[0565] 2. Server

[0566] Analysis module: Analyzes the received text data.

[0567] Generative AI: Generates appropriate responses based on analysis results.

[0568] Program processing and explanation

[0569] Receiving voice input

[0570] Terminal

[0571] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[0572] Converting audio data to text

[0573] Terminal

[0574] The captured voice data is sent to a voice recognition engine, which converts the voice into text data such as "Please tell me about nearby restaurants."

[0575] Sending text data

[0576] Communication Module

[0577] The converted text data is sent to the server through the communication module.

[0578] Analysis of text data and response generation using generative AI

[0579] server

[0580] The server uses an analysis module to analyze the received text data and determine that the user is inquiring about nearby restaurants. The generative AI then generates a response based on the analysis results, such as "I will list restaurants within 1 km of here."

[0581] Sending response data

[0582] server

[0583] The generated response is sent from the server to the terminal.

[0584] Converting response data to audio

[0585] Terminal

[0586] The received text response is sent to a speech synthesis engine and converted into voice data.

[0587] Voice response to the user

[0588] Terminal

[0589] Finally, the speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[0590] Specific examples

[0591] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[0592] 1. Voice Input

[0593] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[0594] 2. Speech-to-text

[0595] The device converts the speech to text, generating the text "Can you tell me where to find a restaurant nearby?"

[0596] 3. Sending text data

[0597] The converted text data is sent to the server.

[0598] 4. Parsing and Response Generation

[0599] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[0600] 5. Transcribing and delivering responses

[0601] The server sends a response to the device, which converts it into voice and provides it to the user. The robot then provides voice guidance, saying, "I'll list restaurants within 1 km of here."

[0602] Real-world usage scenarios

[0603] This system is expected to be used in shopping malls, airports, hotels, etc. For example, if it is installed as a guide robot in a shopping mall, when a visitor asks, "Where is the restroom?", the system will provide quick and accurate guidance.

[0604] The present invention allows users to quickly obtain accurate information through natural voice dialogue, contributing to improved user satisfaction.

[0605] The processing flow will be explained below.

[0606] Step 1:

[0607] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[0608] Step 2:

[0609] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[0610] Step 3:

[0611] The device sends the converted text data to the server via the communication module, and the text data is sent as an HTTP request to the server's API endpoint.

[0612] Step 4:

[0613] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[0614] Step 5:

[0615] The server issues instructions to the generative AI based on the analysis results, and the generative AI uses the analysis results to generate a response text such as "I will list restaurants within 1 km of here."

[0616] Step 6:

[0617] The server sends the generated response text to the terminal as an HTTP response, and the generated text data is returned to the terminal via the communication module.

[0618] Step 7:

[0619] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech generated is, "I'll list restaurants within 1 km of here."

[0620] Step 8:

[0621] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here."

[0622] Example 1

[0623] 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."

[0624] Conventional voice dialogue systems have insufficient speech recognition accuracy and response generation quality, making it difficult for users to have smooth and natural dialogue. In addition, the overall system processing speed is slow, making it difficult to provide instant responses. This has led to problems that degrade the user experience.

[0625] 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.

[0626] In this invention, the server includes means for analyzing text data using generative artificial intelligence to generate a response, means for transmitting the generated response data to the terminal, and means for the terminal to convert the response into voice data and provide the user with the response. This makes it possible to streamline the process from voice input to response generation and provide the user with a quick and accurate response.

[0627] The "means for receiving voice input" is a means for capturing voice from a user using a device such as a microphone.

[0628] The "means for converting speech input into text data" refers to a means including a speech recognition engine that analyzes received speech and converts it into a corresponding text format.

[0629] The "means for transmitting text data to a server" refers to means including a communication module for transmitting the converted text data to a server via a network.

[0630] "Means for analyzing text data and generating responses using generative artificial intelligence" refers to means for analyzing text data on a server using natural language processing technology and generating appropriate responses using generative artificial intelligence.

[0631] The "means for receiving response data from the server" refers to means including a communication module for receiving the generated response data from the server to the terminal.

[0632] The "means for converting received response data into voice data" refers to a means for converting received response data in text format into voice data using a voice synthesis engine.

[0633] The "means for providing audio data to the user" refers to a means for reproducing the converted audio data to the user through a speaker.

[0634] The present invention relates to a system equipped with generative artificial intelligence (AI) that provides information through dialogue with a user. Specifically, the system receives voice input, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, ultimately providing a voice response to the user.

[0635] System configuration

[0636] Terminal

[0637] Microphone: Responsible for capturing audio input from the user.

[0638] Speech recognition engine: Converts the captured voice into text data. Common cloud services can be used as speech recognition engines, such as Google Cloud Speech-to-Text and Microsoft Azure Speech Services.

[0639] Communication module: Transmits the converted text data to the server using an internet connection, including a Wi-Fi module and a 4G / 5G module.

[0640] Speech synthesis engine: Converts the response received from the server into voice data. For voice synthesis, Amazon Polly or Google Cloud Text-to-Speech are used.

[0641] Speaker: Used to provide audio data to the user.

[0642] server

[0643] Analysis module: Used to analyze text data sent from the device, using NLP libraries (e.g., SpaCy or NLTK) to understand the user's intent.

[0644] Generative AI: Generate appropriate responses based on the analysis results. For this purpose, OpenAI's GPT-3 and BERT are used as generative AI.

[0645] Specific operation example

[0646] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the system operates as follows:

[0647] 1. Receiving voice input

[0648] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by a microphone.

[0649] 2. Converting voice data to text

[0650] The captured voice data is sent to a voice recognition engine (for example, Google Cloud Speech-to-Text) and converted into text data such as "Please tell me where to find a nearby restaurant."

[0651] 3. Sending text data

[0652] The converted text data is sent to the server via a communication module, which uses an internet connection.

[0653] 4. Text Data Analysis and Response Generation

[0654] The server analyzes the received text data using an analysis module (e.g., SpaCy) to determine that the user is inquiring about nearby restaurants. Next, a generative AI (e.g., GPT-3) generates a response based on the analysis results, such as "I'll list restaurants within 1 km of here."

[0655] 5. Sending response data

[0656] The generated response is sent from the server to the terminal, typically using HTTPS as the communication protocol.

[0657] 6. Speech conversion of response data

[0658] The device sends the received text response to a speech synthesis engine (e.g., Amazon Polly), which converts it into voice data.

[0659] 7. Audio response to the user

[0660] The speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[0661] Prompt Sentence Examples

[0662] "User said: 'What restaurants are nearby?'"

[0663] In this way, the system enables users to quickly and accurately obtain information through natural voice interaction.

[0664] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0665] Step 1:

[0666] Receiving voice input

[0667] Subject: Terminal

[0668] What it does: The device's microphone captures audio input from the user.

[0669] Input: User speaks to the robot, "What restaurants are nearby?"

[0670] Output: The captured audio data.

[0671] What it does: The microphone converts your voice into an electronic signal and stores that data in temporary memory.

[0672] Step 2:

[0673] Converting audio data to text

[0674] Subject: Terminal

[0675] How it works: Captured voice data is sent to a speech recognition engine, which analyzes the voice data and converts it into corresponding text.

[0676] Input: The captured audio data.

[0677] Output: Text data: "Please tell me where to find a restaurant nearby."

[0678] How it works: The speech recognition engine converts the audio signal into the time-frequency domain and parses it into text. Google Cloud Speech-to-Text can be used as the speech recognition engine.

[0679] Step 3:

[0680] Sending text data

[0681] Subject: Terminal

[0682] Operation: The converted text data is sent to the server through the communication module.

[0683] Input: The converted text data.

[0684] Output: The text data sent to the server.

[0685] Specific operation: Uses Wi-Fi module and 4G / 5G module to send data to the server via HTTPS protocol.

[0686] Step 4:

[0687] Analysis of text data and response generation using generative AI

[0688] Subject: Server

[0689] How it works: The server analyzes the received text data using the analysis module. Based on the analysis results, the generative AI generates an appropriate response.

[0690] Input: Text data sent to the server.

[0691] Output: The response "List of restaurants within 1km of here."

[0692] How it works: The server uses SpaCy, an NLP library, to parse the text data, then uses a generative AI model such as OpenAI's GPT-3 to generate a response.

[0693] Step 5:

[0694] Sending response data

[0695] Subject: Server

[0696] Operation: The generated response data is sent from the server to the device.

[0697] Input: The text data generated as a response.

[0698] Output: The response data sent to the device.

[0699] Specific operation: The server uses the HTTPS protocol to send the generated response data to the terminal.

[0700] Step 6:

[0701] Converting response data to audio

[0702] Subject: Terminal

[0703] How it works: The received text response is sent to a speech synthesis engine and converted into audio data.

[0704] Input: The text format of the response data received from the server.

[0705] Output: Audio data.

[0706] Specific operation: Uses Amazon Polly as a speech synthesis engine to convert text to speech.

[0707] Step 7:

[0708] Voice response to the user

[0709] Subject: Terminal

[0710] Action: A voice response is played through the speaker and provided to the user.

[0711] Input: Speech data generated by the speech synthesis engine.

[0712] Output: The audio response provided to the user.

[0713] Specific operation: The speaker plays the audio data and provides it to the user as sound.

[0714] (Application example 1)

[0715] 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."

[0716] Conventional information guidance systems in brick-and-mortar stores have the problem of being unable to provide immediate and accurate responses to specific product or location information that users are looking for. Particularly in large stores with many products, users often experience inconvenience due to being unable to quickly obtain information. To solve this problem, an efficient system that provides real-time information based on the user's voice input is needed.

[0717] 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.

[0718] In this invention, the server includes means for receiving voice input, means for converting the voice input into text data, means for including in the server a generative AI for analyzing the text data, means for receiving analysis results from the server, means for converting the analysis results into voice data, means for providing the voice data to the user, and means for providing voice guidance about store information based on user questions in the physical store. This enables users to obtain information they desire in the physical store, such as product locations and sale information, in real time, improving the user experience.

[0719] A "means for receiving voice input" is a device or sensor for detecting a user's voice, including a microphone and related hardware.

[0720] "Means for converting voice input into text data" refers to a voice recognition engine that converts voice into text, thereby converting voice data into character data.

[0721] "Means including a server with a generative AI for analyzing text data" refers to a server equipped with artificial intelligence that receives text data, analyzes its contents, and generates an appropriate response.

[0722] "Means for receiving analysis results from the server" refers to a communication module or related software for receiving the analysis results generated by the server.

[0723] "Means for converting the analysis results into voice data" refers to a voice synthesis engine that converts the text data of the analysis results into voice data.

[0724] The "means for providing audio data to the user" refers to an output device such as a speaker for reproducing the converted audio data and providing it to the user.

[0725] "Means for providing voice guidance on in-store information based on questions posed by users in a physical store" refers to a system that includes generative AI and communication means for providing voice guidance on in-store information, such as product locations and sale information, in response to voice questions posed by users in a physical store.

[0726] The present invention relates to a system that provides information desired by a user in a physical store through voice dialogue. This system performs voice input, voice recognition, text analysis, response generation using generative AI, and voice synthesis to provide a voice response. An embodiment of the present invention will be described in detail below.

[0727] Basic configuration

[0728] 1. Voice Input

[0729] The system uses a microphone to receive the user's voice. When the user asks the robot, "Where is the toilet paper counter?", the voice is captured by the microphone.

[0730] 2. Voice Recognition

[0731] The voice data captured by the microphone is sent to the device's speech recognition engine, which analyzes the voice data and converts it into text data. Specifically, the speech recognition engine uses existing technologies such as the Google Speech Recognition API.

[0732] 3. Sending text data

[0733] The converted text data is sent to the server via the communication module, using an HTTP POST request.

[0734] 4. Text Analysis and Response Generation

[0735] The server receives the text data and analyzes it using an analysis module. Based on the analysis results, a generative AI generates an optimal response. The generative AI can use an advanced natural language processing model such as GPT-3.

[0736] 5. Sending response data

[0737] The generated response text is sent to the terminal again through the communication module.

[0738] 6. Speech Synthesis

[0739] The response text sent to the device is converted into voice data by a speech synthesis engine, using technologies such as Google Text-to-Speech (gTTS).

[0740] 7. Audio Provision

[0741] Finally, the voice data is sent to the user through a speaker, and the robot announces, "Toilet paper is sold at Aisle 5."

[0742] Equipment and software used

[0743] Hardware: microphone, speaker, communication module

[0744] software:

[0745] SpeechRecognition (Python library)

[0746] gTTS (Google Text-to-Speech)

[0747] requests (a Python library for sending HTTP requests)

[0748] Advanced natural language processing models (generative AI, e.g. GPT-3)

[0749] Specific examples

[0750] For example, if a user says "Tell me what products are on sale," the following happens:

[0751] 1. The user asks the robot verbally, "Please tell me what products are on sale."

[0752] 2. The device converts the speech to text, generating the text "Tell me what items are on sale."

[0753] 3. This text data is sent to the server.

[0754] 4. The server analyzes the text data, and the generative AI generates a response such as, "Here are the items currently on sale:..."

[0755] 5. The server sends a response to the device, which converts it into audio and provides it to the user.

[0756] Example prompt sentence:

[0757] "Where is the toilet paper section?"

[0758] "Please tell me what products are on sale."

[0759] In this way, the present invention can efficiently and quickly provide information that users desire in a physical store, thereby improving the user experience.

[0760] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0761] Step 1:

[0762] The user provides voice input. The user asks a question to the robot in the physical store, for example, "Where is the toilet paper section?" This voice input is captured by the microphone.

[0763] Input: User's voice

[0764] Output: Audio data captured by the microphone

[0765] Step 2:

[0766] The device sends the captured voice data to a speech recognition engine, which converts the voice data into text using a service such as the Google Speech Recognition API.

[0767] Input: Captured audio data

[0768] Output: Transformed text data (e.g., "Where is the toilet paper counter?")

[0769] Step 3:

[0770] The terminal sends the converted text data to the server via the communication module, using a standard HTTP POST request.

[0771] Input: Converted text data

[0772] Output: Text data sent to the server

[0773] Step 4:

[0774] The server receives the text data and analyzes it using an analysis module. A generative AI, such as GPT-3, generates an appropriate response based on the analysis results, such as "The toilet paper section is on Aisle 5."

[0775] Input: Text data received by the server

[0776] Data processing / data calculation: Text analysis and response generation using generative AI

[0777] Output: The generated response text (e.g., "Toilet paper is available on Aisle 5")

[0778] Step 5:

[0779] The server then sends the generated response text to the terminal via the communication module, again using an HTTP POST request.

[0780] Input: Generated response text

[0781] Output: Response text sent to the terminal

[0782] Step 6:

[0783] The device sends the received response text to a speech synthesis engine, which converts the text data into voice data using a service such as Google Text-to-Speech (gTTS).

[0784] Input: Received response text

[0785] Output: Speech data (e.g. "Toilet paper is on Aisle 5")

[0786] Step 7:

[0787] The terminal provides the user with voice data through a speaker, and ultimately the user can hear voice guidance from the robot.

[0788] Input: Audio data

[0789] Output: A spoken response played through the speaker (e.g., "Toilet paper is on Aisle 5")

[0790] In this way, the data input and output are clearly defined at each step, making it a system that allows users to quickly obtain useful information.

[0791] 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.

[0792] This invention combines a robot equipped with generative AI that provides information through dialogue with the user with an emotion engine that recognizes the user's emotions. This system performs voice input, voice recognition, text analysis, response generation using generative AI, emotion recognition, and voice synthesis, and ultimately provides a voice response to the user. The specific form is shown below.

[0793] composition

[0794] 1. Terminal

[0795] Microphone: Captures audio input from the user.

[0796] Speech recognition engine: Converts captured voice input into text data.

[0797] Emotion engine: Analyzes user emotions from voice input.

[0798] Communication module: Sends text data and emotion information to the server.

[0799] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[0800] Speaker: Provides audio data to the user.

[0801] 2. Server

[0802] Analysis module: Analyzes the received text data.

[0803] Generative AI: Generates appropriate responses based on analysis results.

[0804] Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state.

[0805] Program processing and explanation

[0806] Receiving voice input

[0807] Terminal

[0808] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[0809] Speech data text conversion and sentiment analysis

[0810] Terminal

[0811] The captured voice data is sent to a speech recognition engine, which converts the voice into text data such as "Please tell me where to find a nearby restaurant." At the same time, an emotion engine analyzes the voice data and identifies the user's emotional state (e.g., happy, worried, etc.).

[0812] Sending text data and emotional information

[0813] Communication Module

[0814] The converted text data and emotion information are sent to the server via the communication module. The text data and emotion information are sent as an HTTP request to the server's API endpoint.

[0815] Analysis of text data and response generation using generative AI

[0816] server

[0817] The server uses an analysis module to analyze the received text data and determine the user's intent (whether they want to know about nearby restaurants). Next, the generative AI generates a response text based on the analysis results, such as "I will list restaurants within 1 km of here."

[0818] Response adjustment based on emotional information

[0819] server

[0820] The generated response text is passed to an emotion adjustment module, which adjusts it based on the user's emotional state. For example, if the user is recognized as distressed, the response might be adjusted to "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0821] Sending adjusted response data

[0822] server

[0823] The adjusted response text is sent to the terminal as an HTTP response.

[0824] Converting response data to audio

[0825] Terminal

[0826] The received text response is sent to a speech synthesis engine, which converts it into speech, generating a voice that says, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0827] Voice response to the user

[0828] Terminal

[0829] Finally, the speaker provides the user with a voice response saying, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[0830] Specific examples

[0831] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[0832] 1. Voice Input

[0833] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[0834] 2. Speech-to-text and emotion recognition

[0835] The device converts the speech into text and simultaneously recognizes the user's emotions. The text "Please tell me about nearby restaurants" and the emotional information "I'm in trouble" are generated.

[0836] 3. Data transmission

[0837] The converted text data and emotion information are sent to the server.

[0838] 4. Parsing and Response Generation

[0839] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[0840] 5. Response Adjustment

[0841] The emotion regulation module adjusts the response to, "I'll list restaurants within 1km of here. Let me know if I can help you."

[0842] 6. Transcribing and providing responses

[0843] The adjusted response is sent to the device, where a speech synthesis engine converts it into audio and provides it to the user. The robot then provides a voice guide, saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[0844] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to their emotions.

[0845] The processing flow will be explained below.

[0846] Step 1:

[0847] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[0848] Step 2:

[0849] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[0850] Step 3:

[0851] The device sends the text data to the emotion engine, which analyzes the voice data and identifies the user's emotional state as "troubled."

[0852] Step 4:

[0853] The device sends the converted text data and emotion information to the server via the communication module. The text data and emotion information are sent to the server's API endpoint as an HTTP request.

[0854] Step 5:

[0855] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[0856] Step 6:

[0857] The server issues instructions to the generative AI based on the analysis results, and the generative AI generates a response text such as "I will list restaurants within 1 km of here."

[0858] Step 7:

[0859] The server passes the generated response text to the emotion adjustment module, which adjusts the response based on the user's emotional state ("I'm in trouble"). For example, it generates an adjusted response text such as "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0860] Step 8:

[0861] The server sends the adjusted response text to the terminal as an HTTP response. The adjusted text data is returned to the terminal.

[0862] Step 9:

[0863] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech synthesis engine generates a voice saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[0864] Step 10:

[0865] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[0866] Example 2

[0867] 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."

[0868] Conventional dialogue systems have difficulty providing responses that take into account not only the user's intentions but also their emotional state. As a result, users are often dissatisfied with the responses from the system, and there is a demand for more natural dialogue and accurate information provision.

[0869] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for analyzing text data using generative artificial intelligence and generating an appropriate response, a means for adjusting the generated response based on the user's emotions, and a means for receiving the adjusted analysis result. This makes it possible to provide natural and accurate information according to the user's emotional state.

[0870] "Voice input" is the process of capturing a user's spoken words or instructions with a device such as a microphone and recognizing them as digital signals.

[0871] "Means for converting into text data" refers to a technique or device that converts captured voice input into character string data.

[0872] "Means for recognizing emotions" refers to technologies or devices that analyze and determine a user's emotional state from voice or other input data.

[0873] "Transmitting means" refers to the communications technology or equipment that sends the converted and analyzed data from one point to another.

[0874] "Generative AI" is an AI technology that generates appropriate responses based on input data.

[0875] The "means for adjusting the response" refers to a technique or device that modifies the generated response to an appropriate form depending on the emotional state of the user.

[0876] The "means for receiving the analysis results" refers to the technology or device that inputs the analyzed data and generated responses sent from the server into the terminal.

[0877] "Means for converting into audio data" refers to technology or devices that convert character data into audio signals that can be heard by the user.

[0878] The "means for providing" refers to a technique or device for transmitting the generated audio data to the user through an output device such as a speaker.

[0879] This invention provides a system that enables natural dialogue between a user and a robot. The main components of the system include voice input, voice recognition, emotion recognition, generative artificial intelligence (AI), emotion regulation, voice synthesis, and a communication module for processing and transmitting the respective data.

[0880] Hardware and software used

[0881] Terminal

[0882] Microphone: Captures audio input from the user.

[0883] Speech recognition engine: Converts captured voice input into text data, for example, using the Google Cloud Speech-to-Text API.

[0884] Emotion engine: Analyzes user emotions from voice input, for example using IBM Watson Tone Analyzer.

[0885] Communication module: Sends text data and emotion information to the server.

[0886] Speech synthesis engine: Converts the analysis results received from the server into voice data. For example, use Amazon Polly.

[0887] Speaker: Provides audio data to the user.

[0888] server

[0889] Analysis module: Analyzes the received text data.

[0890] Generative artificial intelligence (AI): Generates appropriate responses based on analysis results, for example, using OpenAI GPT-3.

[0891] Emotion adjustment module: adjusts the generated response based on the user's emotional state.

[0892] Specific examples of processing

[0893] The system operates as follows.

[0894] 1. The user speaks to the robot: "Tell me about a nearby restaurant." The device's microphone captures this voice.

[0895] 2. Text conversion and emotion recognition: The captured voice data is sent to a speech recognition engine and converted into text data: "Please tell me where to find a nearby restaurant." At the same time, the emotion engine analyzes the voice data and identifies the emotion information, "I'm in trouble."

[0896] 3. Sending data to the server: The converted text data and emotion information are sent to the server through the communication module.

[0897] 4. Analysis of text data and generation of a response using generative AI: The analysis module on the server analyzes the received text data and determines the user's intent (want to know about nearby restaurants). Based on the analysis results, the generative AI generates a response text such as "I will list restaurants within 1 km of here." The following prompt sentence is used:

[0898] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[0899] 5. Response adjustment based on emotion information: The generated response text is passed to the emotion adjustment module, which adjusts the response based on the emotion tag “in need” to “I’ll list restaurants within 1km of here. Let me know if I can help you.”

[0900] 6. Sending adjusted response data: The adjusted response text is sent to the terminal again through the communication module.

[0901] 7. Speech synthesis and provision: The received response text data is sent to a speech synthesis engine, which converts it into voice data saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you." Finally, the voice is played over the speaker and provided to the user.

[0902] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to emotions.

[0903] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0904] Program processing flow

[0905] Step 1: Capturing Audio Input

[0906] Terminal

[0907] Input: The user says, "What restaurants are nearby?"

[0908] Processing: A microphone captures this audio and stores it as digital data.

[0909] Output: The user's voice is stored as digital audio data.

[0910] Specific actions

[0911] When a user speaks to the robot, their voice is converted into data by a microphone, and the data is temporarily stored in a buffer on the device.

[0912] Step 2: Converting speech data to text and analyzing sentiment

[0913] Terminal

[0914] Input: Captured digital audio data.

[0915] Processing: The speech recognition engine converts the voice data into text data, and the emotion engine analyzes the voice data to recognize the emotional state.

[0916] Output: Text data "Please tell me a nearby restaurant" and emotion information "I'm in trouble."

[0917] Specific actions

[0918] A speech recognition engine (e.g., Google Cloud Speech-to-Text API) converts the speech into text, generating the string "Please tell me where to find a nearby restaurant." At the same time, an emotion engine (e.g., IBM Watson Tone Analyzer) analyzes the speech data and generates an emotion tag such as "I'm in trouble."

[0919] Step 3: Sending text data and emotion information to the server

[0920] Communication Module

[0921] Input: Text data and emotion information.

[0922] Processing: Sends an HTTP POST request to the server.

[0923] Output: Text data and emotion information sent to the server.

[0924] Specific actions

[0925] The communication module generates an HTTP POST request containing text data and emotion information in the payload and sends it to the server's API endpoint.

[0926] POST / api / v1 / parse

[0927] Host: example.com

[0928] Content-Type: application / json

[0929] {

[0930] "text": "Can you tell me nearby restaurants?",

[0931] "emotion": "troubled"

[0932] }

[0933] Step 4: Analyzing text data and generating responses using AI

[0934] server

[0935] Input: Text data and emotion information.

[0936] Processing: The analytics module analyzes the text data to determine the user's intent, and the generative AI model generates an appropriate response.

[0937] Output: Response text "Listing restaurants within 1km of here."

[0938] Specific actions

[0939] The analysis module analyzes the text data and extracts the user's intent, which is "I want to know about nearby restaurants." The following prompt sentence is used to generate the response:

[0940] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[0941] The AI ​​model (e.g., OpenAI GPT-3) generates a response text such as "I will list restaurants within 1 km of here."

[0942] Step 5: Adjusting responses based on emotional information

[0943] server

[0944] Input: Response text and sentiment information.

[0945] Processing: The emotion regulation module adjusts the response based on the user's emotions.

[0946] Output: Tailored response text: "I'll list restaurants within 1km of here. Let me know if I can help you."

[0947] Specific actions

[0948] The emotion adjustment module adjusts the response based on the emotion tag "distressed," resulting in the final response: "I'll list restaurants within 1km of here. Let me know if I can help you."

[0949] Step 6: Sending adjusted response data to the device

[0950] server

[0951] Input: The adjusted response text.

[0952] Processing: Send to the terminal as an HTTP response.

[0953] Output: The adjusted response text sent to the terminal.

[0954] Specific actions

[0955] The server generates an HTTP response and sends it back to the device, including the adjusted response text.

[0956] HTTP / 1.1 200 OK

[0957] Content-Type: application / json

[0958] {

[0959] "response": "I'll list restaurants within 1km of here. Let me know if I can help you."

[0960] }

[0961] Step 7: Transcribing response data

[0962] Terminal

[0963] Input: The adjusted response text.

[0964] Processing: The speech synthesis engine converts the text data into speech data.

[0965] Output: Speech data saying "I'll list restaurants within 1km of here. Let me know if I can help you."

[0966] Specific actions

[0967] The adjusted response text is sent to a speech synthesis engine (e.g., Amazon Polly) to obtain the corresponding voice data, which is then saved as an audio file for playback.

[0968] Step 8: Audible response to the user

[0969] Terminal

[0970] Input: Audio data.

[0971] Processing: Outputs sound through the speaker.

[0972] Output: An audio response that the user can hear.

[0973] Specific actions

[0974] A voice comes through the speaker and tells the user, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[0975] (Application example 2)

[0976] 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."

[0977] Current voice input systems are unable to respond taking into account the user's emotional state, and therefore often provide inappropriate information to users. Furthermore, guidance services in brick-and-mortar stores are also required to understand what the user is looking for and provide appropriate information, but no system exists that can achieve this. This poses the challenge of making it difficult to improve the user experience and provide efficient guidance.

[0978] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means including a generative AI for analyzing text data, means for adjusting a response generated based on the user's emotional information, and means for receiving the analysis results from the server. This makes it possible to provide appropriate information that takes the user's emotions into consideration.

[0979] A "means for receiving voice input" is a device or function that captures voice from a user and converts it into input data.

[0980] The "means for converting speech input into text data" refers to a technology or device that converts received speech input into text data using speech recognition technology.

[0981] "Means for analyzing user emotions" refers to a technology or device that analyzes and identifies the user's emotional state from voice data or text data.

[0982] The "means for transmitting text data and emotion information to a server" is a communication device or function for transmitting the analyzed text data and emotion information to a server.

[0983] "Means including a generative AI on a server for analyzing text data" refers to a device or system that includes a generative AI function on a server for analyzing received text data and generating an appropriate response.

[0984] The "means for adjusting the response generated based on the user's emotional information" refers to a function or device that adjusts the response based on the analysis results in accordance with the user's emotional state.

[0985] The "means for receiving the analysis results from the server" is a communication device or function for receiving the analysis results provided by the server.

[0986] "Means for converting analysis results into audio data" refers to a technology or device that converts text-based responses into audio data.

[0987] The "means for providing audio data to the user" refers to a playback device or function such as a speaker or headphone that allows the user to listen to the converted audio data.

[0988] To implement this invention, a system is required that executes each step of voice input, voice recognition, emotion analysis, response generation using generative AI, response adjustment, and voice synthesis. The system consists of a terminal that interacts with the user and a server that analyzes data and generates responses.

[0989] Hardware and Software Configuration

[0990] Device:

[0991] 1. Microphone: Captures audio input from the user.

[0992] 2. Speech recognition engine: Converts captured voice into text data. This function uses the Google Speech-to-Text API.

[0993] 3. Emotion engine: Analyzes the user's emotions from voice input. This analysis uses the Microsoft Azure Emotion API.

[0994] 4. Communication module: Sends text data and emotion information to the server using HTTP and REST API as the communication protocol.

[0995] 5. Speech synthesis engine: Converts the analysis results received from the server into voice data. This function uses the Google Text-to-Speech API.

[0996] 6. Speaker: Provides the converted audio data to the user.

[0997] server:

[0998] 1. Analysis module: Analyzes the received text data using natural language processing technology.

[0999] 2. Generative AI: Generates appropriate responses based on the analysis results. The OpenAI GPT-3 model is used for generative AI.

[1000] 3. Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state. For example, if the user is in a distressed emotional state, the response will be changed to a more considerate one.

[1001] Data processing and calculation

[1002] The device receives voice data and converts it into text data using a voice recognition engine. At the same time, the emotion engine analyzes the emotional information. This data is then sent to the server via a communication module.

[1003] On the server, the analysis module analyzes the text data, and the generative AI generates a response. The generated response is adjusted by the emotion adjustment module based on the user's emotional state. The final response is sent back to the device.

[1004] The response text data returned to the terminal is converted into voice data by a voice synthesis engine and provided to the user through a speaker.

[1005] Specific examples

[1006] The user asks aloud, "Where is the recommended sweets corner?" This voice is captured by a microphone and converted into text data by a speech recognition engine. At the same time, an emotion engine analyzes the user's emotional state (distress). The text data and emotion information are sent to the server.

[1007] The server's analysis module analyzes the text data, and the generative AI generates an appropriate response. This response is then adjusted by the emotion adjustment module to, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions." This adjusted response is sent to the device, where it is converted into voice data by the voice synthesis engine. It is then provided to the user via the speaker.

[1008] Prompt Sentence Examples

[1009] "A customer is asking, 'Where's the recommended snack section?' Distressed emotion detected. Suggest an appropriate response."

[1010] The above is an embodiment of the present invention.

[1011] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1012] Step 1:

[1013] Receiving voice input

[1014] The user speaks into the microphone on their smartphone or smart glasses, asking, "Where is the recommended sweets corner?"

[1015] Input: User's voice data.

[1016] What it does: The microphone captures sound and stores it as audio data.

[1017] Output: The captured audio data.

[1018] Step 2:

[1019] Speech data text conversion and sentiment analysis

[1020] The device sends the captured voice data to a speech recognition engine to convert it into text data, while an emotion engine analyzes the voice data to identify the user's emotional state.

[1021] Input: Captured audio data.

[1022] How it works: The speech recognition engine converts the speech into text data such as "Where is the recommended snack corner?" The emotion engine analyzes the speech data and generates emotion information such as "I'm in trouble."

[1023] Output: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[1024] Step 3:

[1025] Sending text data and emotional information

[1026] The terminal transmits the converted text data and emotion information to the server via the communication module.

[1027] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[1028] Operation: The communication module sends text data and emotion information as an HTTP request to the server's API endpoint.

[1029] Output: Text data and emotion information sent to the server.

[1030] Step 4:

[1031] Analysis of text data and response generation using generative AI

[1032] The server uses an analysis module to analyze the received text data and determine the user's intent. Next, a generative AI generates an appropriate response based on the analysis results.

[1033] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[1034] How it works: The analysis module analyzes the text data and determines the user's intent (the user wants to know where the candy corner is). The generative AI generates a response text such as "The candy corner is in the center of the third floor."

[1035] Output: Response text "The candy corner is in the center of the third floor."

[1036] Step 5:

[1037] Response adjustment based on emotional information

[1038] The server passes the generated response text to an emotion adjustment module, which adjusts the response based on the user's emotional state.

[1039] Input: Emotion information "I'm in trouble" and response text "The candy corner is in the center of the third floor."

[1040] How it works: The emotion adjustment module adjusts the response text depending on the emotion, resulting in a final response of "The candy corner is in the center of the third floor. Let me know if you need any further directions."

[1041] Output: Adjusted response text: "The candy section is in the center of the third floor. Let me know if you need any further directions."

[1042] Step 6:

[1043] Sending adjusted response data

[1044] The server sends the adjusted response text to the terminal as an HTTP response.

[1045] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[1046] Operation: The server sends the response text to the terminal as an HTTP response.

[1047] Output: The adjusted response text sent to the terminal.

[1048] Step 7:

[1049] Converting response data to audio

[1050] The terminal sends the received response text to a speech synthesis engine and converts it into voice data.

[1051] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[1052] How it works: A speech synthesis engine converts text data into speech.

[1053] Output: Speech data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[1054] Step 8:

[1055] Voice response to the user

[1056] The terminal ultimately provides a voice response to the user through a speaker.

[1057] Input: Voice data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[1058] Action: The speaker plays the audio data and provides a response to the user.

[1059] Output: A voice description that says, "The sweets corner is in the center of the third floor. Let me know if you need any further directions."

[1060] 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.

[1061] 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.

[1062] 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.

[1063] [Third embodiment]

[1064] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1065] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1066] 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).

[1067] 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.

[1068] 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.

[1069] 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).

[1070] 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.

[1071] 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.

[1072] 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.

[1073] 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.

[1074] 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.

[1075] 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."

[1076] This invention relates to a robot equipped with generative AI that provides information through dialogue with a user. This system receives voice input from a user, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, and ultimately provides a voice response to the user.

[1077] Specifically, it is implemented in the following manner.

[1078] composition

[1079] 1. Terminal

[1080] Microphone: Captures audio input from the user.

[1081] Speech recognition engine: Converts captured voice input into text data.

[1082] Communication module: Transmits the converted text data to the server.

[1083] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[1084] Speaker: Provides audio data to the user.

[1085] 2. Server

[1086] Analysis module: Analyzes the received text data.

[1087] Generative AI: Generates appropriate responses based on analysis results.

[1088] Program processing and explanation

[1089] Receiving voice input

[1090] Terminal

[1091] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[1092] Converting audio data to text

[1093] Terminal

[1094] The captured voice data is sent to a voice recognition engine, which converts the voice into text data such as "Please tell me about nearby restaurants."

[1095] Sending text data

[1096] Communication Module

[1097] The converted text data is sent to the server through the communication module.

[1098] Analysis of text data and response generation using generative AI

[1099] server

[1100] The server uses an analysis module to analyze the received text data and determine that the user is inquiring about nearby restaurants. The generative AI then generates a response based on the analysis results, such as "I will list restaurants within 1 km of here."

[1101] Sending response data

[1102] server

[1103] The generated response is sent from the server to the terminal.

[1104] Converting response data to audio

[1105] Terminal

[1106] The received text response is sent to a speech synthesis engine and converted into voice data.

[1107] Voice response to the user

[1108] Terminal

[1109] Finally, the speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[1110] Specific examples

[1111] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[1112] 1. Voice Input

[1113] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[1114] 2. Speech-to-text

[1115] The device converts the speech to text, generating the text "Can you tell me where to find a restaurant nearby?"

[1116] 3. Sending text data

[1117] The converted text data is sent to the server.

[1118] 4. Parsing and Response Generation

[1119] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[1120] 5. Transcribing and delivering responses

[1121] The server sends a response to the device, which converts it into voice and provides it to the user. The robot then provides voice guidance, saying, "I'll list restaurants within 1 km of here."

[1122] Real-world usage scenarios

[1123] This system is expected to be used in shopping malls, airports, hotels, etc. For example, if it is installed as a guide robot in a shopping mall, when a visitor asks, "Where is the restroom?", the system will provide quick and accurate guidance.

[1124] The present invention allows users to quickly obtain accurate information through natural voice dialogue, contributing to improved user satisfaction.

[1125] The processing flow will be explained below.

[1126] Step 1:

[1127] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[1128] Step 2:

[1129] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[1130] Step 3:

[1131] The device sends the converted text data to the server via the communication module, and the text data is sent as an HTTP request to the server's API endpoint.

[1132] Step 4:

[1133] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[1134] Step 5:

[1135] The server issues instructions to the generative AI based on the analysis results, and the generative AI uses the analysis results to generate a response text such as "I will list restaurants within 1 km of here."

[1136] Step 6:

[1137] The server sends the generated response text to the terminal as an HTTP response, and the generated text data is returned to the terminal via the communication module.

[1138] Step 7:

[1139] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech generated is, "I'll list restaurants within 1 km of here."

[1140] Step 8:

[1141] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here."

[1142] Example 1

[1143] 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."

[1144] Conventional voice dialogue systems have insufficient speech recognition accuracy and response generation quality, making it difficult for users to have smooth and natural dialogue. In addition, the overall system processing speed is slow, making it difficult to provide instant responses. This has led to problems that degrade the user experience.

[1145] 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.

[1146] In this invention, the server includes means for analyzing text data using generative artificial intelligence to generate a response, means for transmitting the generated response data to the terminal, and means for the terminal to convert the response into voice data and provide the user with the response. This makes it possible to streamline the process from voice input to response generation and provide the user with a quick and accurate response.

[1147] The "means for receiving voice input" is a means for capturing voice from a user using a device such as a microphone.

[1148] The "means for converting speech input into text data" refers to a means including a speech recognition engine that analyzes received speech and converts it into a corresponding text format.

[1149] The "means for transmitting text data to a server" refers to means including a communication module for transmitting the converted text data to a server via a network.

[1150] "Means for analyzing text data and generating responses using generative artificial intelligence" refers to means for analyzing text data on a server using natural language processing technology and generating appropriate responses using generative artificial intelligence.

[1151] The "means for receiving response data from the server" refers to means including a communication module for receiving the generated response data from the server to the terminal.

[1152] The "means for converting received response data into voice data" refers to a means for converting received response data in text format into voice data using a voice synthesis engine.

[1153] The "means for providing audio data to the user" refers to a means for reproducing the converted audio data to the user through a speaker.

[1154] The present invention relates to a system equipped with generative artificial intelligence (AI) that provides information through dialogue with a user. Specifically, the system receives voice input, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, ultimately providing a voice response to the user.

[1155] System configuration

[1156] Terminal

[1157] Microphone: Responsible for capturing audio input from the user.

[1158] Speech recognition engine: Converts the captured voice into text data. Common cloud services can be used as speech recognition engines, such as Google Cloud Speech-to-Text and Microsoft Azure Speech Services.

[1159] Communication module: Transmits the converted text data to the server using an internet connection, including a Wi-Fi module and a 4G / 5G module.

[1160] Speech synthesis engine: Converts the response received from the server into voice data. For voice synthesis, Amazon Polly or Google Cloud Text-to-Speech are used.

[1161] Speaker: Used to provide audio data to the user.

[1162] server

[1163] Analysis module: Used to analyze text data sent from the device, using NLP libraries (e.g., SpaCy or NLTK) to understand the user's intent.

[1164] Generative AI: Generate appropriate responses based on the analysis results. For this purpose, OpenAI's GPT-3 and BERT are used as generative AI.

[1165] Specific operation example

[1166] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the system operates as follows:

[1167] 1. Receiving voice input

[1168] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by a microphone.

[1169] 2. Converting voice data to text

[1170] The captured voice data is sent to a voice recognition engine (for example, Google Cloud Speech-to-Text) and converted into text data such as "Please tell me where to find a nearby restaurant."

[1171] 3. Sending text data

[1172] The converted text data is sent to the server via a communication module, which uses an internet connection.

[1173] 4. Text Data Analysis and Response Generation

[1174] The server analyzes the received text data using an analysis module (e.g., SpaCy) to determine that the user is inquiring about nearby restaurants. Next, a generative AI (e.g., GPT-3) generates a response based on the analysis results, such as "I'll list restaurants within 1 km of here."

[1175] 5. Sending response data

[1176] The generated response is sent from the server to the terminal, typically using HTTPS as the communication protocol.

[1177] 6. Speech conversion of response data

[1178] The device sends the received text response to a speech synthesis engine (e.g., Amazon Polly), which converts it into voice data.

[1179] 7. Audio response to the user

[1180] The speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[1181] Prompt Sentence Examples

[1182] "User said: 'What restaurants are nearby?'"

[1183] In this way, the system enables users to quickly and accurately obtain information through natural voice interaction.

[1184] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1185] Step 1:

[1186] Receiving voice input

[1187] Subject: Terminal

[1188] What it does: The device's microphone captures audio input from the user.

[1189] Input: User speaks to the robot, "What restaurants are nearby?"

[1190] Output: The captured audio data.

[1191] What it does: The microphone converts your voice into an electronic signal and stores that data in temporary memory.

[1192] Step 2:

[1193] Converting audio data to text

[1194] Subject: Terminal

[1195] How it works: Captured voice data is sent to a speech recognition engine, which analyzes the voice data and converts it into corresponding text.

[1196] Input: The captured audio data.

[1197] Output: Text data: "Please tell me where to find a restaurant nearby."

[1198] How it works: The speech recognition engine converts the audio signal into the time-frequency domain and parses it into text. Google Cloud Speech-to-Text can be used as the speech recognition engine.

[1199] Step 3:

[1200] Sending text data

[1201] Subject: Terminal

[1202] Operation: The converted text data is sent to the server through the communication module.

[1203] Input: The converted text data.

[1204] Output: The text data sent to the server.

[1205] Specific operation: Uses Wi-Fi module and 4G / 5G module to send data to the server via HTTPS protocol.

[1206] Step 4:

[1207] Analysis of text data and response generation using generative AI

[1208] Subject: Server

[1209] How it works: The server analyzes the received text data using the analysis module. Based on the analysis results, the generative AI generates an appropriate response.

[1210] Input: Text data sent to the server.

[1211] Output: The response "List of restaurants within 1km of here."

[1212] How it works: The server uses SpaCy, an NLP library, to parse the text data, then uses a generative AI model such as OpenAI's GPT-3 to generate a response.

[1213] Step 5:

[1214] Sending response data

[1215] Subject: Server

[1216] Operation: The generated response data is sent from the server to the device.

[1217] Input: The text data generated as a response.

[1218] Output: The response data sent to the device.

[1219] Specific operation: The server uses the HTTPS protocol to send the generated response data to the terminal.

[1220] Step 6:

[1221] Converting response data to audio

[1222] Subject: Terminal

[1223] How it works: The received text response is sent to a speech synthesis engine and converted into audio data.

[1224] Input: The text format of the response data received from the server.

[1225] Output: Audio data.

[1226] Specific operation: Uses Amazon Polly as a speech synthesis engine to convert text to speech.

[1227] Step 7:

[1228] Voice response to the user

[1229] Subject: Terminal

[1230] Action: A voice response is played through the speaker and provided to the user.

[1231] Input: Speech data generated by the speech synthesis engine.

[1232] Output: The audio response provided to the user.

[1233] Specific operation: The speaker plays the audio data and provides it to the user as sound.

[1234] (Application example 1)

[1235] 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."

[1236] Conventional information guidance systems in brick-and-mortar stores have the problem of being unable to provide immediate and accurate responses to specific product or location information that users are looking for. Particularly in large stores with many products, users often experience inconvenience due to being unable to quickly obtain information. To solve this problem, an efficient system that provides real-time information based on the user's voice input is needed.

[1237] 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.

[1238] In this invention, the server includes means for receiving voice input, means for converting the voice input into text data, means for including in the server a generative AI for analyzing the text data, means for receiving analysis results from the server, means for converting the analysis results into voice data, means for providing the voice data to the user, and means for providing voice guidance about store information based on user questions in the physical store. This enables users to obtain information they desire in the physical store, such as product locations and sale information, in real time, improving the user experience.

[1239] A "means for receiving voice input" is a device or sensor for detecting a user's voice, including a microphone and related hardware.

[1240] "Means for converting voice input into text data" refers to a voice recognition engine that converts voice into text, thereby converting voice data into character data.

[1241] "Means including a server with a generative AI for analyzing text data" refers to a server equipped with artificial intelligence that receives text data, analyzes its contents, and generates an appropriate response.

[1242] "Means for receiving analysis results from the server" refers to a communication module or related software for receiving the analysis results generated by the server.

[1243] "Means for converting the analysis results into voice data" refers to a voice synthesis engine that converts the text data of the analysis results into voice data.

[1244] The "means for providing audio data to the user" refers to an output device such as a speaker for reproducing the converted audio data and providing it to the user.

[1245] "Means for providing voice guidance on in-store information based on questions posed by users in a physical store" refers to a system that includes generative AI and communication means for providing voice guidance on in-store information, such as product locations and sale information, in response to voice questions posed by users in a physical store.

[1246] The present invention relates to a system that provides information desired by a user in a physical store through voice dialogue. This system performs voice input, voice recognition, text analysis, response generation using generative AI, and voice synthesis to provide a voice response. An embodiment of the present invention will be described in detail below.

[1247] Basic configuration

[1248] 1. Voice Input

[1249] The system uses a microphone to receive the user's voice. When the user asks the robot, "Where is the toilet paper counter?", the voice is captured by the microphone.

[1250] 2. Voice Recognition

[1251] The voice data captured by the microphone is sent to the device's speech recognition engine, which analyzes the voice data and converts it into text data. Specifically, the speech recognition engine uses existing technologies such as the Google Speech Recognition API.

[1252] 3. Sending text data

[1253] The converted text data is sent to the server via the communication module, using an HTTP POST request.

[1254] 4. Text Analysis and Response Generation

[1255] The server receives the text data and analyzes it using an analysis module. Based on the analysis results, a generative AI generates an optimal response. The generative AI can use an advanced natural language processing model such as GPT-3.

[1256] 5. Sending response data

[1257] The generated response text is sent to the terminal again through the communication module.

[1258] 6. Speech Synthesis

[1259] The response text sent to the device is converted into voice data by a speech synthesis engine, using technologies such as Google Text-to-Speech (gTTS).

[1260] 7. Audio Provision

[1261] Finally, the voice data is sent to the user through a speaker, and the robot announces, "Toilet paper is sold at Aisle 5."

[1262] Equipment and software used

[1263] Hardware: microphone, speaker, communication module

[1264] software:

[1265] SpeechRecognition (Python library)

[1266] gTTS (Google Text-to-Speech)

[1267] requests (a Python library for sending HTTP requests)

[1268] Advanced natural language processing models (generative AI, e.g. GPT-3)

[1269] Specific examples

[1270] For example, if a user says "Tell me what products are on sale," the following happens:

[1271] 1. The user asks the robot verbally, "Please tell me what products are on sale."

[1272] 2. The device converts the speech to text, generating the text "Tell me what items are on sale."

[1273] 3. This text data is sent to the server.

[1274] 4. The server analyzes the text data, and the generative AI generates a response such as, "Here are the items currently on sale:..."

[1275] 5. The server sends a response to the device, which converts it into audio and provides it to the user.

[1276] Example prompt sentence:

[1277] "Where is the toilet paper section?"

[1278] "Please tell me what products are on sale."

[1279] In this way, the present invention can efficiently and quickly provide information that users desire in a physical store, thereby improving the user experience.

[1280] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1281] Step 1:

[1282] The user provides voice input. The user asks a question to the robot in the physical store, for example, "Where is the toilet paper section?" This voice input is captured by the microphone.

[1283] Input: User's voice

[1284] Output: Audio data captured by the microphone

[1285] Step 2:

[1286] The device sends the captured voice data to a speech recognition engine, which converts the voice data into text using a service such as the Google Speech Recognition API.

[1287] Input: Captured audio data

[1288] Output: Transformed text data (e.g., "Where is the toilet paper counter?")

[1289] Step 3:

[1290] The terminal sends the converted text data to the server via the communication module, using a standard HTTP POST request.

[1291] Input: Converted text data

[1292] Output: Text data sent to the server

[1293] Step 4:

[1294] The server receives the text data and analyzes it using an analysis module. A generative AI, such as GPT-3, generates an appropriate response based on the analysis results, such as "The toilet paper section is on Aisle 5."

[1295] Input: Text data received by the server

[1296] Data processing / data calculation: Text analysis and response generation using generative AI

[1297] Output: The generated response text (e.g., "Toilet paper is available on Aisle 5")

[1298] Step 5:

[1299] The server then sends the generated response text to the terminal via the communication module, again using an HTTP POST request.

[1300] Input: Generated response text

[1301] Output: Response text sent to the terminal

[1302] Step 6:

[1303] The device sends the received response text to a speech synthesis engine, which converts the text data into voice data using a service such as Google Text-to-Speech (gTTS).

[1304] Input: Received response text

[1305] Output: Speech data (e.g. "Toilet paper is on Aisle 5")

[1306] Step 7:

[1307] The terminal provides the user with voice data through a speaker, and ultimately the user can hear voice guidance from the robot.

[1308] Input: Audio data

[1309] Output: A spoken response played through the speaker (e.g., "Toilet paper is on Aisle 5")

[1310] In this way, the data input and output are clearly defined at each step, making it a system that allows users to quickly obtain useful information.

[1311] 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.

[1312] This invention combines a robot equipped with generative AI that provides information through dialogue with the user with an emotion engine that recognizes the user's emotions. This system performs voice input, voice recognition, text analysis, response generation using generative AI, emotion recognition, and voice synthesis, and ultimately provides a voice response to the user. The specific form is shown below.

[1313] composition

[1314] 1. Terminal

[1315] Microphone: Captures audio input from the user.

[1316] Speech recognition engine: Converts captured voice input into text data.

[1317] Emotion engine: Analyzes user emotions from voice input.

[1318] Communication module: Sends text data and emotion information to the server.

[1319] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[1320] Speaker: Provides audio data to the user.

[1321] 2. Server

[1322] Analysis module: Analyzes the received text data.

[1323] Generative AI: Generates appropriate responses based on analysis results.

[1324] Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state.

[1325] Program processing and explanation

[1326] Receiving voice input

[1327] Terminal

[1328] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[1329] Speech data text conversion and sentiment analysis

[1330] Terminal

[1331] The captured voice data is sent to a speech recognition engine, which converts the voice into text data such as "Please tell me where to find a nearby restaurant." At the same time, an emotion engine analyzes the voice data and identifies the user's emotional state (e.g., happy, worried, etc.).

[1332] Sending text data and emotional information

[1333] Communication Module

[1334] The converted text data and emotion information are sent to the server via the communication module. The text data and emotion information are sent as an HTTP request to the server's API endpoint.

[1335] Analysis of text data and response generation using generative AI

[1336] server

[1337] The server uses an analysis module to analyze the received text data and determine the user's intent (whether they want to know about nearby restaurants). Next, the generative AI generates a response text based on the analysis results, such as "I will list restaurants within 1 km of here."

[1338] Response adjustment based on emotional information

[1339] server

[1340] The generated response text is passed to an emotion adjustment module, which adjusts it based on the user's emotional state. For example, if the user is recognized as distressed, the response might be adjusted to "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1341] Sending adjusted response data

[1342] server

[1343] The adjusted response text is sent to the terminal as an HTTP response.

[1344] Converting response data to audio

[1345] Terminal

[1346] The received text response is sent to a speech synthesis engine, which converts it into speech, generating a voice that says, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1347] Voice response to the user

[1348] Terminal

[1349] Finally, the speaker provides the user with a voice response saying, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[1350] Specific examples

[1351] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[1352] 1. Voice Input

[1353] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[1354] 2. Speech-to-text and emotion recognition

[1355] The device converts the speech into text and simultaneously recognizes the user's emotions. The text "Please tell me about nearby restaurants" and the emotional information "I'm in trouble" are generated.

[1356] 3. Data transmission

[1357] The converted text data and emotion information are sent to the server.

[1358] 4. Parsing and Response Generation

[1359] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[1360] 5. Response Adjustment

[1361] The emotion regulation module adjusts the response to, "I'll list restaurants within 1km of here. Let me know if I can help you."

[1362] 6. Transcribing and providing responses

[1363] The adjusted response is sent to the device, where a speech synthesis engine converts it into audio and provides it to the user. The robot then provides a voice guide, saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[1364] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to their emotions.

[1365] The processing flow will be explained below.

[1366] Step 1:

[1367] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[1368] Step 2:

[1369] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[1370] Step 3:

[1371] The device sends the text data to the emotion engine, which analyzes the voice data and identifies the user's emotional state as "troubled."

[1372] Step 4:

[1373] The device sends the converted text data and emotion information to the server via the communication module. The text data and emotion information are sent to the server's API endpoint as an HTTP request.

[1374] Step 5:

[1375] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[1376] Step 6:

[1377] The server issues instructions to the generative AI based on the analysis results, and the generative AI generates a response text such as "I will list restaurants within 1 km of here."

[1378] Step 7:

[1379] The server passes the generated response text to the emotion adjustment module, which adjusts the response based on the user's emotional state ("I'm in trouble"). For example, it generates an adjusted response text such as "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1380] Step 8:

[1381] The server sends the adjusted response text to the terminal as an HTTP response. The adjusted text data is returned to the terminal.

[1382] Step 9:

[1383] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech synthesis engine generates a voice saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[1384] Step 10:

[1385] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1386] Example 2

[1387] 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."

[1388] Conventional dialogue systems have difficulty providing responses that take into account not only the user's intentions but also their emotional state. As a result, users are often dissatisfied with the responses from the system, and there is a demand for more natural dialogue and accurate information provision.

[1389] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for analyzing text data using generative artificial intelligence and generating an appropriate response, a means for adjusting the generated response based on the user's emotions, and a means for receiving the adjusted analysis result. This makes it possible to provide natural and accurate information according to the user's emotional state.

[1390] "Voice input" is the process of capturing a user's spoken words or instructions with a device such as a microphone and recognizing them as digital signals.

[1391] "Means for converting into text data" refers to a technique or device that converts captured voice input into character string data.

[1392] "Means for recognizing emotions" refers to technologies or devices that analyze and determine a user's emotional state from voice or other input data.

[1393] "Transmitting means" refers to the communications technology or equipment that sends the converted and analyzed data from one point to another.

[1394] "Generative AI" is an AI technology that generates appropriate responses based on input data.

[1395] The "means for adjusting the response" refers to a technique or device that modifies the generated response to an appropriate form depending on the emotional state of the user.

[1396] The "means for receiving the analysis results" refers to the technology or device that inputs the analyzed data and generated responses sent from the server into the terminal.

[1397] "Means for converting into audio data" refers to technology or devices that convert character data into audio signals that can be heard by the user.

[1398] The "means for providing" refers to a technique or device for transmitting the generated audio data to the user through an output device such as a speaker.

[1399] This invention provides a system that enables natural dialogue between a user and a robot. The main components of the system include voice input, voice recognition, emotion recognition, generative artificial intelligence (AI), emotion regulation, voice synthesis, and a communication module for processing and transmitting the respective data.

[1400] Hardware and software used

[1401] Terminal

[1402] Microphone: Captures audio input from the user.

[1403] Speech recognition engine: Converts captured voice input into text data, for example, using the Google Cloud Speech-to-Text API.

[1404] Emotion engine: Analyzes user emotions from voice input, for example using IBM Watson Tone Analyzer.

[1405] Communication module: Sends text data and emotion information to the server.

[1406] Speech synthesis engine: Converts the analysis results received from the server into voice data. For example, use Amazon Polly.

[1407] Speaker: Provides audio data to the user.

[1408] server

[1409] Analysis module: Analyzes the received text data.

[1410] Generative artificial intelligence (AI): Generates appropriate responses based on analysis results, for example, using OpenAI GPT-3.

[1411] Emotion adjustment module: adjusts the generated response based on the user's emotional state.

[1412] Specific examples of processing

[1413] The system operates as follows.

[1414] 1. The user speaks to the robot: "Tell me about a nearby restaurant." The device's microphone captures this voice.

[1415] 2. Text conversion and emotion recognition: The captured voice data is sent to a speech recognition engine and converted into text data: "Please tell me where to find a nearby restaurant." At the same time, the emotion engine analyzes the voice data and identifies the emotion information, "I'm in trouble."

[1416] 3. Sending data to the server: The converted text data and emotion information are sent to the server through the communication module.

[1417] 4. Analysis of text data and generation of a response using generative AI: The analysis module on the server analyzes the received text data and determines the user's intent (want to know about nearby restaurants). Based on the analysis results, the generative AI generates a response text such as "I will list restaurants within 1 km of here." The following prompt sentence is used:

[1418] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[1419] 5. Response adjustment based on emotion information: The generated response text is passed to the emotion adjustment module, which adjusts the response based on the emotion tag “in need” to “I’ll list restaurants within 1km of here. Let me know if I can help you.”

[1420] 6. Sending adjusted response data: The adjusted response text is sent to the terminal again through the communication module.

[1421] 7. Speech synthesis and provision: The received response text data is sent to a speech synthesis engine, which converts it into voice data saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you." Finally, the voice is played over the speaker and provided to the user.

[1422] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to emotions.

[1423] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1424] Program processing flow

[1425] Step 1: Capturing Audio Input

[1426] Terminal

[1427] Input: The user says, "What restaurants are nearby?"

[1428] Processing: A microphone captures this audio and stores it as digital data.

[1429] Output: The user's voice is stored as digital audio data.

[1430] Specific actions

[1431] When a user speaks to the robot, their voice is converted into data by a microphone, and the data is temporarily stored in a buffer on the device.

[1432] Step 2: Converting speech data to text and analyzing sentiment

[1433] Terminal

[1434] Input: Captured digital audio data.

[1435] Processing: The speech recognition engine converts the voice data into text data, and the emotion engine analyzes the voice data to recognize the emotional state.

[1436] Output: Text data "Please tell me a nearby restaurant" and emotion information "I'm in trouble."

[1437] Specific actions

[1438] A speech recognition engine (e.g., Google Cloud Speech-to-Text API) converts the speech into text, generating the string "Please tell me where to find a nearby restaurant." At the same time, an emotion engine (e.g., IBM Watson Tone Analyzer) analyzes the speech data and generates an emotion tag such as "I'm in trouble."

[1439] Step 3: Sending text data and emotion information to the server

[1440] Communication Module

[1441] Input: Text data and emotion information.

[1442] Processing: Sends an HTTP POST request to the server.

[1443] Output: Text data and emotion information sent to the server.

[1444] Specific actions

[1445] The communication module generates an HTTP POST request containing text data and emotion information in the payload and sends it to the server's API endpoint.

[1446] POST / api / v1 / parse

[1447] Host: example.com

[1448] Content-Type: application / json

[1449] {

[1450] "text": "Can you tell me nearby restaurants?",

[1451] "emotion": "troubled"

[1452] }

[1453] Step 4: Analyzing text data and generating responses using AI

[1454] server

[1455] Input: Text data and emotion information.

[1456] Processing: The analytics module analyzes the text data to determine the user's intent, and the generative AI model generates an appropriate response.

[1457] Output: Response text "Listing restaurants within 1km of here."

[1458] Specific actions

[1459] The analysis module analyzes the text data and extracts the user's intent, which is "I want to know about nearby restaurants." The following prompt sentence is used to generate the response:

[1460] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[1461] The AI ​​model (e.g., OpenAI GPT-3) generates a response text such as "I will list restaurants within 1 km of here."

[1462] Step 5: Adjusting responses based on emotional information

[1463] server

[1464] Input: Response text and sentiment information.

[1465] Processing: The emotion regulation module adjusts the response based on the user's emotions.

[1466] Output: Tailored response text: "I'll list restaurants within 1km of here. Let me know if I can help you."

[1467] Specific actions

[1468] The emotion adjustment module adjusts the response based on the emotion tag "distressed," resulting in the final response: "I'll list restaurants within 1km of here. Let me know if I can help you."

[1469] Step 6: Sending adjusted response data to the device

[1470] server

[1471] Input: The adjusted response text.

[1472] Processing: Send to the terminal as an HTTP response.

[1473] Output: The adjusted response text sent to the terminal.

[1474] Specific actions

[1475] The server generates an HTTP response and sends it back to the device, including the adjusted response text.

[1476] HTTP / 1.1 200 OK

[1477] Content-Type: application / json

[1478] {

[1479] "response": "I'll list restaurants within 1km of here. Let me know if I can help you."

[1480] }

[1481] Step 7: Transcribing response data

[1482] Terminal

[1483] Input: The adjusted response text.

[1484] Processing: The speech synthesis engine converts the text data into speech data.

[1485] Output: Speech data saying "I'll list restaurants within 1km of here. Let me know if I can help you."

[1486] Specific actions

[1487] The adjusted response text is sent to a speech synthesis engine (e.g., Amazon Polly) to obtain the corresponding voice data, which is then saved as an audio file for playback.

[1488] Step 8: Audible response to the user

[1489] Terminal

[1490] Input: Audio data.

[1491] Processing: Outputs sound through the speaker.

[1492] Output: An audio response that the user can hear.

[1493] Specific actions

[1494] A voice comes through the speaker and tells the user, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[1495] (Application example 2)

[1496] 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."

[1497] Current voice input systems are unable to respond taking into account the user's emotional state, and therefore often provide inappropriate information to users. Furthermore, guidance services in brick-and-mortar stores are also required to understand what the user is looking for and provide appropriate information, but no system exists that can achieve this. This poses the challenge of making it difficult to improve the user experience and provide efficient guidance.

[1498] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means including a generative AI for analyzing text data, means for adjusting a response generated based on the user's emotional information, and means for receiving the analysis results from the server. This makes it possible to provide appropriate information that takes the user's emotions into consideration.

[1499] A "means for receiving voice input" is a device or function that captures voice from a user and converts it into input data.

[1500] The "means for converting speech input into text data" refers to a technology or device that converts received speech input into text data using speech recognition technology.

[1501] "Means for analyzing user emotions" refers to a technology or device that analyzes and identifies the user's emotional state from voice data or text data.

[1502] The "means for transmitting text data and emotion information to a server" is a communication device or function for transmitting the analyzed text data and emotion information to a server.

[1503] "Means including a generative AI on a server for analyzing text data" refers to a device or system that includes a generative AI function on a server for analyzing received text data and generating an appropriate response.

[1504] The "means for adjusting the response generated based on the user's emotional information" refers to a function or device that adjusts the response based on the analysis results in accordance with the user's emotional state.

[1505] The "means for receiving the analysis results from the server" is a communication device or function for receiving the analysis results provided by the server.

[1506] "Means for converting analysis results into audio data" refers to a technology or device that converts text-based responses into audio data.

[1507] The "means for providing audio data to the user" refers to a playback device or function such as a speaker or headphone that allows the user to listen to the converted audio data.

[1508] To implement this invention, a system is required that executes each step of voice input, voice recognition, emotion analysis, response generation using generative AI, response adjustment, and voice synthesis. The system consists of a terminal that interacts with the user and a server that analyzes data and generates responses.

[1509] Hardware and Software Configuration

[1510] Device:

[1511] 1. Microphone: Captures audio input from the user.

[1512] 2. Speech recognition engine: Converts captured voice into text data. This function uses the Google Speech-to-Text API.

[1513] 3. Emotion engine: Analyzes the user's emotions from voice input. This analysis uses the Microsoft Azure Emotion API.

[1514] 4. Communication module: Sends text data and emotion information to the server using HTTP and REST API as the communication protocol.

[1515] 5. Speech synthesis engine: Converts the analysis results received from the server into voice data. This function uses the Google Text-to-Speech API.

[1516] 6. Speaker: Provides the converted audio data to the user.

[1517] server:

[1518] 1. Analysis module: Analyzes the received text data using natural language processing technology.

[1519] 2. Generative AI: Generates appropriate responses based on the analysis results. The OpenAI GPT-3 model is used for generative AI.

[1520] 3. Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state. For example, if the user is in a distressed emotional state, the response will be changed to a more considerate one.

[1521] Data processing and calculation

[1522] The device receives voice data and converts it into text data using a voice recognition engine. At the same time, the emotion engine analyzes the emotional information. This data is then sent to the server via a communication module.

[1523] On the server, the analysis module analyzes the text data, and the generative AI generates a response. The generated response is adjusted by the emotion adjustment module based on the user's emotional state. The final response is sent back to the device.

[1524] The response text data returned to the terminal is converted into voice data by a voice synthesis engine and provided to the user through a speaker.

[1525] Specific examples

[1526] The user asks aloud, "Where is the recommended sweets corner?" This voice is captured by a microphone and converted into text data by a speech recognition engine. At the same time, an emotion engine analyzes the user's emotional state (distress). The text data and emotion information are sent to the server.

[1527] The server's analysis module analyzes the text data, and the generative AI generates an appropriate response. This response is then adjusted by the emotion adjustment module to, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions." This adjusted response is sent to the device, where it is converted into voice data by the voice synthesis engine. It is then provided to the user via the speaker.

[1528] Prompt Sentence Examples

[1529] "A customer is asking, 'Where's the recommended snack section?' Distressed emotion detected. Suggest an appropriate response."

[1530] The above is an embodiment of the present invention.

[1531] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1532] Step 1:

[1533] Receiving voice input

[1534] The user speaks into the microphone on their smartphone or smart glasses, asking, "Where is the recommended sweets corner?"

[1535] Input: User's voice data.

[1536] What it does: The microphone captures sound and stores it as audio data.

[1537] Output: The captured audio data.

[1538] Step 2:

[1539] Speech data text conversion and sentiment analysis

[1540] The device sends the captured voice data to a speech recognition engine to convert it into text data, while an emotion engine analyzes the voice data to identify the user's emotional state.

[1541] Input: Captured audio data.

[1542] How it works: The speech recognition engine converts the speech into text data such as "Where is the recommended snack corner?" The emotion engine analyzes the speech data and generates emotion information such as "I'm in trouble."

[1543] Output: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[1544] Step 3:

[1545] Sending text data and emotional information

[1546] The terminal transmits the converted text data and emotion information to the server via the communication module.

[1547] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[1548] Operation: The communication module sends text data and emotion information as an HTTP request to the server's API endpoint.

[1549] Output: Text data and emotion information sent to the server.

[1550] Step 4:

[1551] Analysis of text data and response generation using generative AI

[1552] The server uses an analysis module to analyze the received text data and determine the user's intent. Next, a generative AI generates an appropriate response based on the analysis results.

[1553] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[1554] How it works: The analysis module analyzes the text data and determines the user's intent (the user wants to know where the candy corner is). The generative AI generates a response text such as "The candy corner is in the center of the third floor."

[1555] Output: Response text "The candy corner is in the center of the third floor."

[1556] Step 5:

[1557] Response adjustment based on emotional information

[1558] The server passes the generated response text to an emotion adjustment module, which adjusts the response based on the user's emotional state.

[1559] Input: Emotion information "I'm in trouble" and response text "The candy corner is in the center of the third floor."

[1560] How it works: The emotion adjustment module adjusts the response text depending on the emotion, resulting in a final response of "The candy corner is in the center of the third floor. Let me know if you need any further directions."

[1561] Output: Adjusted response text: "The candy section is in the center of the third floor. Let me know if you need any further directions."

[1562] Step 6:

[1563] Sending adjusted response data

[1564] The server sends the adjusted response text to the terminal as an HTTP response.

[1565] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[1566] Operation: The server sends the response text to the terminal as an HTTP response.

[1567] Output: The adjusted response text sent to the terminal.

[1568] Step 7:

[1569] Converting response data to audio

[1570] The terminal sends the received response text to a speech synthesis engine and converts it into voice data.

[1571] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[1572] How it works: A speech synthesis engine converts text data into speech.

[1573] Output: Speech data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[1574] Step 8:

[1575] Voice response to the user

[1576] The terminal ultimately provides a voice response to the user through a speaker.

[1577] Input: Voice data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[1578] Action: The speaker plays the audio data and provides a response to the user.

[1579] Output: A voice description that says, "The sweets corner is in the center of the third floor. Let me know if you need any further directions."

[1580] 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.

[1581] 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.

[1582] 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.

[1583] [Fourth embodiment]

[1584] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1585] 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.

[1586] 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).

[1587] 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.

[1588] 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.

[1589] 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).

[1590] 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.

[1591] 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.

[1592] 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.

[1593] 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.

[1594] 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.

[1595] 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.

[1596] 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."

[1597] This invention relates to a robot equipped with generative AI that provides information through dialogue with a user. This system receives voice input from a user, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, and ultimately provides a voice response to the user.

[1598] Specifically, it is implemented in the following manner.

[1599] composition

[1600] 1. Terminal

[1601] Microphone: Captures audio input from the user.

[1602] Speech recognition engine: Converts captured voice input into text data.

[1603] Communication module: Transmits the converted text data to the server.

[1604] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[1605] Speaker: Provides audio data to the user.

[1606] 2. Server

[1607] Analysis module: Analyzes the received text data.

[1608] Generative AI: Generates appropriate responses based on analysis results.

[1609] Program processing and explanation

[1610] Receiving voice input

[1611] Terminal

[1612] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[1613] Converting audio data to text

[1614] Terminal

[1615] The captured voice data is sent to a voice recognition engine, which converts the voice into text data such as "Please tell me about nearby restaurants."

[1616] Sending text data

[1617] Communication Module

[1618] The converted text data is sent to the server through the communication module.

[1619] Analysis of text data and response generation using generative AI

[1620] server

[1621] The server uses an analysis module to analyze the received text data and determine that the user is inquiring about nearby restaurants. The generative AI then generates a response based on the analysis results, such as "I will list restaurants within 1 km of here."

[1622] Sending response data

[1623] server

[1624] The generated response is sent from the server to the terminal.

[1625] Converting response data to audio

[1626] Terminal

[1627] The received text response is sent to a speech synthesis engine and converted into voice data.

[1628] Voice response to the user

[1629] Terminal

[1630] Finally, the speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[1631] Specific examples

[1632] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[1633] 1. Voice Input

[1634] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[1635] 2. Speech-to-text

[1636] The device converts the speech to text, generating the text "Can you tell me where to find a restaurant nearby?"

[1637] 3. Sending text data

[1638] The converted text data is sent to the server.

[1639] 4. Parsing and Response Generation

[1640] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[1641] 5. Transcribing and delivering responses

[1642] The server sends a response to the device, which converts it into voice and provides it to the user. The robot then provides voice guidance, saying, "I'll list restaurants within 1 km of here."

[1643] Real-world usage scenarios

[1644] This system is expected to be used in shopping malls, airports, hotels, etc. For example, if it is installed as a guide robot in a shopping mall, when a visitor asks, "Where is the restroom?", the system will provide quick and accurate guidance.

[1645] The present invention allows users to quickly obtain accurate information through natural voice dialogue, contributing to improved user satisfaction.

[1646] The processing flow will be explained below.

[1647] Step 1:

[1648] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[1649] Step 2:

[1650] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[1651] Step 3:

[1652] The device sends the converted text data to the server via the communication module, and the text data is sent as an HTTP request to the server's API endpoint.

[1653] Step 4:

[1654] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[1655] Step 5:

[1656] The server issues instructions to the generative AI based on the analysis results, and the generative AI uses the analysis results to generate a response text such as "I will list restaurants within 1 km of here."

[1657] Step 6:

[1658] The server sends the generated response text to the terminal as an HTTP response, and the generated text data is returned to the terminal via the communication module.

[1659] Step 7:

[1660] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech generated is, "I'll list restaurants within 1 km of here."

[1661] Step 8:

[1662] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here."

[1663] Example 1

[1664] 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."

[1665] Conventional voice dialogue systems have insufficient speech recognition accuracy and response generation quality, making it difficult for users to have smooth and natural dialogue. In addition, the overall system processing speed is slow, making it difficult to provide instant responses. This has led to problems that degrade the user experience.

[1666] 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.

[1667] In this invention, the server includes means for analyzing text data using generative artificial intelligence to generate a response, means for transmitting the generated response data to the terminal, and means for the terminal to convert the response into voice data and provide the user with the response. This makes it possible to streamline the process from voice input to response generation and provide the user with a quick and accurate response.

[1668] The "means for receiving voice input" is a means for capturing voice from a user using a device such as a microphone.

[1669] The "means for converting speech input into text data" refers to a means including a speech recognition engine that analyzes received speech and converts it into a corresponding text format.

[1670] The "means for transmitting text data to a server" refers to means including a communication module for transmitting the converted text data to a server via a network.

[1671] "Means for analyzing text data and generating responses using generative artificial intelligence" refers to means for analyzing text data on a server using natural language processing technology and generating appropriate responses using generative artificial intelligence.

[1672] The "means for receiving response data from the server" refers to means including a communication module for receiving the generated response data from the server to the terminal.

[1673] The "means for converting received response data into voice data" refers to a means for converting received response data in text format into voice data using a voice synthesis engine.

[1674] The "means for providing audio data to the user" refers to a means for reproducing the converted audio data to the user through a speaker.

[1675] The present invention relates to a system equipped with generative artificial intelligence (AI) that provides information through dialogue with a user. Specifically, the system receives voice input, performs speech recognition, text analysis, response generation using generative AI, and speech synthesis, ultimately providing a voice response to the user.

[1676] System configuration

[1677] Terminal

[1678] Microphone: Responsible for capturing audio input from the user.

[1679] Speech recognition engine: Converts the captured voice into text data. Common cloud services can be used as speech recognition engines, such as Google Cloud Speech-to-Text and Microsoft Azure Speech Services.

[1680] Communication module: Transmits the converted text data to the server using an internet connection, including a Wi-Fi module and a 4G / 5G module.

[1681] Speech synthesis engine: Converts the response received from the server into voice data. For voice synthesis, Amazon Polly or Google Cloud Text-to-Speech are used.

[1682] Speaker: Used to provide audio data to the user.

[1683] server

[1684] Analysis module: Used to analyze text data sent from the device, using NLP libraries (e.g., SpaCy or NLTK) to understand the user's intent.

[1685] Generative AI: Generate appropriate responses based on the analysis results. For this purpose, OpenAI's GPT-3 and BERT are used as generative AI.

[1686] Specific operation example

[1687] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the system operates as follows:

[1688] 1. Receiving voice input

[1689] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by a microphone.

[1690] 2. Converting voice data to text

[1691] The captured voice data is sent to a voice recognition engine (for example, Google Cloud Speech-to-Text) and converted into text data such as "Please tell me where to find a nearby restaurant."

[1692] 3. Sending text data

[1693] The converted text data is sent to the server via a communication module, which uses an internet connection.

[1694] 4. Text Data Analysis and Response Generation

[1695] The server analyzes the received text data using an analysis module (e.g., SpaCy) to determine that the user is inquiring about nearby restaurants. Next, a generative AI (e.g., GPT-3) generates a response based on the analysis results, such as "I'll list restaurants within 1 km of here."

[1696] 5. Sending response data

[1697] The generated response is sent from the server to the terminal, typically using HTTPS as the communication protocol.

[1698] 6. Speech conversion of response data

[1699] The device sends the received text response to a speech synthesis engine (e.g., Amazon Polly), which converts it into voice data.

[1700] 7. Audio response to the user

[1701] The speaker provides the user with a voice response saying, "I'll list restaurants within 1 km of here."

[1702] Prompt Sentence Examples

[1703] "User said: 'What restaurants are nearby?'"

[1704] In this way, the system enables users to quickly and accurately obtain information through natural voice interaction.

[1705] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1706] Step 1:

[1707] Receiving voice input

[1708] Subject: Terminal

[1709] What it does: The device's microphone captures audio input from the user.

[1710] Input: User speaks to the robot, "What restaurants are nearby?"

[1711] Output: The captured audio data.

[1712] What it does: The microphone converts your voice into an electronic signal and stores that data in temporary memory.

[1713] Step 2:

[1714] Converting audio data to text

[1715] Subject: Terminal

[1716] How it works: Captured voice data is sent to a speech recognition engine, which analyzes the voice data and converts it into corresponding text.

[1717] Input: The captured audio data.

[1718] Output: Text data: "Please tell me where to find a restaurant nearby."

[1719] How it works: The speech recognition engine converts the audio signal into the time-frequency domain and parses it into text. Google Cloud Speech-to-Text can be used as the speech recognition engine.

[1720] Step 3:

[1721] Sending text data

[1722] Subject: Terminal

[1723] Operation: The converted text data is sent to the server through the communication module.

[1724] Input: The converted text data.

[1725] Output: The text data sent to the server.

[1726] Specific operation: Uses Wi-Fi module and 4G / 5G module to send data to the server via HTTPS protocol.

[1727] Step 4:

[1728] Analysis of text data and response generation using generative AI

[1729] Subject: Server

[1730] How it works: The server analyzes the received text data using the analysis module. Based on the analysis results, the generative AI generates an appropriate response.

[1731] Input: Text data sent to the server.

[1732] Output: The response "List of restaurants within 1km of here."

[1733] How it works: The server uses SpaCy, an NLP library, to parse the text data, then uses a generative AI model such as OpenAI's GPT-3 to generate a response.

[1734] Step 5:

[1735] Sending response data

[1736] Subject: Server

[1737] Operation: The generated response data is sent from the server to the device.

[1738] Input: The text data generated as a response.

[1739] Output: The response data sent to the device.

[1740] Specific operation: The server uses the HTTPS protocol to send the generated response data to the terminal.

[1741] Step 6:

[1742] Converting response data to audio

[1743] Subject: Terminal

[1744] How it works: The received text response is sent to a speech synthesis engine and converted into audio data.

[1745] Input: The text format of the response data received from the server.

[1746] Output: Audio data.

[1747] Specific operation: Uses Amazon Polly as a speech synthesis engine to convert text to speech.

[1748] Step 7:

[1749] Voice response to the user

[1750] Subject: Terminal

[1751] Action: A voice response is played through the speaker and provided to the user.

[1752] Input: Speech data generated by the speech synthesis engine.

[1753] Output: The audio response provided to the user.

[1754] Specific operation: The speaker plays the audio data and provides it to the user as sound.

[1755] (Application example 1)

[1756] 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."

[1757] Conventional information guidance systems in brick-and-mortar stores have the problem of being unable to provide immediate and accurate responses to specific product or location information that users are looking for. Particularly in large stores with many products, users often experience inconvenience due to being unable to quickly obtain information. To solve this problem, an efficient system that provides real-time information based on the user's voice input is needed.

[1758] 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.

[1759] In this invention, the server includes means for receiving voice input, means for converting the voice input into text data, means for including in the server a generative AI for analyzing the text data, means for receiving analysis results from the server, means for converting the analysis results into voice data, means for providing the voice data to the user, and means for providing voice guidance about store information based on user questions in the physical store. This enables users to obtain information they desire in the physical store, such as product locations and sale information, in real time, improving the user experience.

[1760] A "means for receiving voice input" is a device or sensor for detecting a user's voice, including a microphone and related hardware.

[1761] "Means for converting voice input into text data" refers to a voice recognition engine that converts voice into text, thereby converting voice data into character data.

[1762] "Means including a server with a generative AI for analyzing text data" refers to a server equipped with artificial intelligence that receives text data, analyzes its contents, and generates an appropriate response.

[1763] "Means for receiving analysis results from the server" refers to a communication module or related software for receiving the analysis results generated by the server.

[1764] "Means for converting the analysis results into voice data" refers to a voice synthesis engine that converts the text data of the analysis results into voice data.

[1765] The "means for providing audio data to the user" refers to an output device such as a speaker for reproducing the converted audio data and providing it to the user.

[1766] "Means for providing voice guidance on in-store information based on questions posed by users in a physical store" refers to a system that includes generative AI and communication means for providing voice guidance on in-store information, such as product locations and sale information, in response to voice questions posed by users in a physical store.

[1767] The present invention relates to a system that provides information desired by a user in a physical store through voice dialogue. This system performs voice input, voice recognition, text analysis, response generation using generative AI, and voice synthesis to provide a voice response. An embodiment of the present invention will be described in detail below.

[1768] Basic configuration

[1769] 1. Voice Input

[1770] The system uses a microphone to receive the user's voice. When the user asks the robot, "Where is the toilet paper counter?", the voice is captured by the microphone.

[1771] 2. Voice Recognition

[1772] The voice data captured by the microphone is sent to the device's speech recognition engine, which analyzes the voice data and converts it into text data. Specifically, the speech recognition engine uses existing technologies such as the Google Speech Recognition API.

[1773] 3. Sending text data

[1774] The converted text data is sent to the server via the communication module, using an HTTP POST request.

[1775] 4. Text Analysis and Response Generation

[1776] The server receives the text data and analyzes it using an analysis module. Based on the analysis results, a generative AI generates an optimal response. The generative AI can use an advanced natural language processing model such as GPT-3.

[1777] 5. Sending response data

[1778] The generated response text is sent to the terminal again through the communication module.

[1779] 6. Speech Synthesis

[1780] The response text sent to the device is converted into voice data by a speech synthesis engine, using technologies such as Google Text-to-Speech (gTTS).

[1781] 7. Audio Provision

[1782] Finally, the voice data is sent to the user through a speaker, and the robot announces, "Toilet paper is sold at Aisle 5."

[1783] Equipment and software used

[1784] Hardware: microphone, speaker, communication module

[1785] software:

[1786] SpeechRecognition (Python library)

[1787] gTTS (Google Text-to-Speech)

[1788] requests (a Python library for sending HTTP requests)

[1789] Advanced natural language processing models (generative AI, e.g. GPT-3)

[1790] Specific examples

[1791] For example, if a user says "Tell me what products are on sale," the following happens:

[1792] 1. The user asks the robot verbally, "Please tell me what products are on sale."

[1793] 2. The device converts the speech to text, generating the text "Tell me what items are on sale."

[1794] 3. This text data is sent to the server.

[1795] 4. The server analyzes the text data, and the generative AI generates a response such as, "Here are the items currently on sale:..."

[1796] 5. The server sends a response to the device, which converts it into audio and provides it to the user.

[1797] Example prompt sentence:

[1798] "Where is the toilet paper section?"

[1799] "Please tell me what products are on sale."

[1800] In this way, the present invention can efficiently and quickly provide information that users desire in a physical store, thereby improving the user experience.

[1801] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1802] Step 1:

[1803] The user provides voice input. The user asks a question to the robot in the physical store, for example, "Where is the toilet paper section?" This voice input is captured by the microphone.

[1804] Input: User's voice

[1805] Output: Audio data captured by the microphone

[1806] Step 2:

[1807] The device sends the captured voice data to a speech recognition engine, which converts the voice data into text using a service such as the Google Speech Recognition API.

[1808] Input: Captured audio data

[1809] Output: Transformed text data (e.g., "Where is the toilet paper counter?")

[1810] Step 3:

[1811] The terminal sends the converted text data to the server via the communication module, using a standard HTTP POST request.

[1812] Input: Converted text data

[1813] Output: Text data sent to the server

[1814] Step 4:

[1815] The server receives the text data and analyzes it using an analysis module. A generative AI, such as GPT-3, generates an appropriate response based on the analysis results, such as "The toilet paper section is on Aisle 5."

[1816] Input: Text data received by the server

[1817] Data processing / data calculation: Text analysis and response generation using generative AI

[1818] Output: The generated response text (e.g., "Toilet paper is available on Aisle 5")

[1819] Step 5:

[1820] The server then sends the generated response text to the terminal via the communication module, again using an HTTP POST request.

[1821] Input: Generated response text

[1822] Output: Response text sent to the terminal

[1823] Step 6:

[1824] The device sends the received response text to a speech synthesis engine, which converts the text data into voice data using a service such as Google Text-to-Speech (gTTS).

[1825] Input: Received response text

[1826] Output: Speech data (e.g. "Toilet paper is on Aisle 5")

[1827] Step 7:

[1828] The terminal provides the user with voice data through a speaker, and ultimately the user can hear voice guidance from the robot.

[1829] Input: Audio data

[1830] Output: A spoken response played through the speaker (e.g., "Toilet paper is on Aisle 5")

[1831] In this way, the data input and output are clearly defined at each step, making it a system that allows users to quickly obtain useful information.

[1832] 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.

[1833] This invention combines a robot equipped with generative AI that provides information through dialogue with the user with an emotion engine that recognizes the user's emotions. This system performs voice input, voice recognition, text analysis, response generation using generative AI, emotion recognition, and voice synthesis, and ultimately provides a voice response to the user. The specific form is shown below.

[1834] composition

[1835] 1. Terminal

[1836] Microphone: Captures audio input from the user.

[1837] Speech recognition engine: Converts captured voice input into text data.

[1838] Emotion engine: Analyzes user emotions from voice input.

[1839] Communication module: Sends text data and emotion information to the server.

[1840] Speech synthesis engine: Converts the analysis results received from the server into voice data.

[1841] Speaker: Provides audio data to the user.

[1842] 2. Server

[1843] Analysis module: Analyzes the received text data.

[1844] Generative AI: Generates appropriate responses based on analysis results.

[1845] Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state.

[1846] Program processing and explanation

[1847] Receiving voice input

[1848] Terminal

[1849] When the user says to the robot, "Tell me where the nearest restaurant is," the voice is captured by the microphone.

[1850] Speech data text conversion and sentiment analysis

[1851] Terminal

[1852] The captured voice data is sent to a speech recognition engine, which converts the voice into text data such as "Please tell me where to find a nearby restaurant." At the same time, an emotion engine analyzes the voice data and identifies the user's emotional state (e.g., happy, worried, etc.).

[1853] Sending text data and emotional information

[1854] Communication Module

[1855] The converted text data and emotion information are sent to the server via the communication module. The text data and emotion information are sent as an HTTP request to the server's API endpoint.

[1856] Analysis of text data and response generation using generative AI

[1857] server

[1858] The server uses an analysis module to analyze the received text data and determine the user's intent (whether they want to know about nearby restaurants). Next, the generative AI generates a response text based on the analysis results, such as "I will list restaurants within 1 km of here."

[1859] Response adjustment based on emotional information

[1860] server

[1861] The generated response text is passed to an emotion adjustment module, which adjusts it based on the user's emotional state. For example, if the user is recognized as distressed, the response might be adjusted to "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1862] Sending adjusted response data

[1863] server

[1864] The adjusted response text is sent to the terminal as an HTTP response.

[1865] Converting response data to audio

[1866] Terminal

[1867] The received text response is sent to a speech synthesis engine, which converts it into speech, generating a voice that says, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1868] Voice response to the user

[1869] Terminal

[1870] Finally, the speaker provides the user with a voice response saying, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[1871] Specific examples

[1872] For example, if a user says to the robot, "Please tell me about a nearby restaurant," the process proceeds as follows:

[1873] 1. Voice Input

[1874] The user asks the robot a question by voice: "Tell me where to find a nearby restaurant."

[1875] 2. Speech-to-text and emotion recognition

[1876] The device converts the speech into text and simultaneously recognizes the user's emotions. The text "Please tell me about nearby restaurants" and the emotional information "I'm in trouble" are generated.

[1877] 3. Data transmission

[1878] The converted text data and emotion information are sent to the server.

[1879] 4. Parsing and Response Generation

[1880] The server analyzes the text data, and the generative AI generates a response such as, "I will list restaurants within 1 km of here."

[1881] 5. Response Adjustment

[1882] The emotion regulation module adjusts the response to, "I'll list restaurants within 1km of here. Let me know if I can help you."

[1883] 6. Transcribing and providing responses

[1884] The adjusted response is sent to the device, where a speech synthesis engine converts it into audio and provides it to the user. The robot then provides a voice guide, saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[1885] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to their emotions.

[1886] The processing flow will be explained below.

[1887] Step 1:

[1888] The user provides voice input: "Tell me about nearby restaurants." The user says to the robot, and the voice is captured by the robot's microphone.

[1889] Step 2:

[1890] The device sends the captured voice data to a voice recognition engine, which analyzes the voice data and converts it into text data such as "Please tell me where to find a nearby restaurant."

[1891] Step 3:

[1892] The device sends the text data to the emotion engine, which analyzes the voice data and identifies the user's emotional state as "troubled."

[1893] Step 4:

[1894] The device sends the converted text data and emotion information to the server via the communication module. The text data and emotion information are sent to the server's API endpoint as an HTTP request.

[1895] Step 5:

[1896] The server passes the received text data to the analysis module, which analyzes the text and understands the user's intent (e.g., "I want to know about nearby restaurants").

[1897] Step 6:

[1898] The server issues instructions to the generative AI based on the analysis results, and the generative AI generates a response text such as "I will list restaurants within 1 km of here."

[1899] Step 7:

[1900] The server passes the generated response text to the emotion adjustment module, which adjusts the response based on the user's emotional state ("I'm in trouble"). For example, it generates an adjusted response text such as "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1901] Step 8:

[1902] The server sends the adjusted response text to the terminal as an HTTP response. The adjusted text data is returned to the terminal.

[1903] Step 9:

[1904] The device sends the received response text data to the speech synthesis engine, which converts the text data into speech data. The speech synthesis engine generates a voice saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you."

[1905] Step 10:

[1906] The device then provides the generated voice data to the user through the speaker, and the user hears a voice response saying, "I'll list restaurants within 1 km of here. Let me know if I can help you."

[1907] Example 2

[1908] 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."

[1909] Conventional dialogue systems have difficulty providing responses that take into account not only the user's intentions but also their emotional state. As a result, users are often dissatisfied with the responses from the system, and there is a demand for more natural dialogue and accurate information provision.

[1910] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for analyzing text data using generative artificial intelligence and generating an appropriate response, a means for adjusting the generated response based on the user's emotions, and a means for receiving the adjusted analysis result. This makes it possible to provide natural and accurate information according to the user's emotional state.

[1911] "Voice input" is the process of capturing a user's spoken words or instructions with a device such as a microphone and recognizing them as digital signals.

[1912] "Means for converting into text data" refers to a technique or device that converts captured voice input into character string data.

[1913] "Means for recognizing emotions" refers to technologies or devices that analyze and determine a user's emotional state from voice or other input data.

[1914] "Transmitting means" refers to the communications technology or equipment that sends the converted and analyzed data from one point to another.

[1915] "Generative AI" is an AI technology that generates appropriate responses based on input data.

[1916] The "means for adjusting the response" refers to a technique or device that modifies the generated response to an appropriate form depending on the emotional state of the user.

[1917] The "means for receiving the analysis results" refers to the technology or device that inputs the analyzed data and generated responses sent from the server into the terminal.

[1918] "Means for converting into audio data" refers to technology or devices that convert character data into audio signals that can be heard by the user.

[1919] The "means for providing" refers to a technique or device for transmitting the generated audio data to the user through an output device such as a speaker.

[1920] This invention provides a system that enables natural dialogue between a user and a robot. The main components of the system include voice input, voice recognition, emotion recognition, generative artificial intelligence (AI), emotion regulation, voice synthesis, and a communication module for processing and transmitting the respective data.

[1921] Hardware and software used

[1922] Terminal

[1923] Microphone: Captures audio input from the user.

[1924] Speech recognition engine: Converts captured voice input into text data, for example, using the Google Cloud Speech-to-Text API.

[1925] Emotion engine: Analyzes user emotions from voice input, for example using IBM Watson Tone Analyzer.

[1926] Communication module: Sends text data and emotion information to the server.

[1927] Speech synthesis engine: Converts the analysis results received from the server into voice data. For example, use Amazon Polly.

[1928] Speaker: Provides audio data to the user.

[1929] server

[1930] Analysis module: Analyzes the received text data.

[1931] Generative artificial intelligence (AI): Generates appropriate responses based on analysis results, for example, using OpenAI GPT-3.

[1932] Emotion adjustment module: adjusts the generated response based on the user's emotional state.

[1933] Specific examples of processing

[1934] The system operates as follows.

[1935] 1. The user speaks to the robot: "Tell me about a nearby restaurant." The device's microphone captures this voice.

[1936] 2. Text conversion and emotion recognition: The captured voice data is sent to a speech recognition engine and converted into text data: "Please tell me where to find a nearby restaurant." At the same time, the emotion engine analyzes the voice data and identifies the emotion information, "I'm in trouble."

[1937] 3. Sending data to the server: The converted text data and emotion information are sent to the server through the communication module.

[1938] 4. Analysis of text data and generation of a response using generative AI: The analysis module on the server analyzes the received text data and determines the user's intent (want to know about nearby restaurants). Based on the analysis results, the generative AI generates a response text such as "I will list restaurants within 1 km of here." The following prompt sentence is used:

[1939] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[1940] 5. Response adjustment based on emotion information: The generated response text is passed to the emotion adjustment module, which adjusts the response based on the emotion tag “in need” to “I’ll list restaurants within 1km of here. Let me know if I can help you.”

[1941] 6. Sending adjusted response data: The adjusted response text is sent to the terminal again through the communication module.

[1942] 7. Speech synthesis and provision: The received response text data is sent to a speech synthesis engine, which converts it into voice data saying, "I'll list restaurants within 1 km of here. Please let me know if I can help you." Finally, the voice is played over the speaker and provided to the user.

[1943] According to the present invention, the user can not only obtain quick and accurate information through natural voice dialogue, but also enjoy appropriate responses according to emotions.

[1944] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1945] Program processing flow

[1946] Step 1: Capturing Audio Input

[1947] Terminal

[1948] Input: The user says, "What restaurants are nearby?"

[1949] Processing: A microphone captures this audio and stores it as digital data.

[1950] Output: The user's voice is stored as digital audio data.

[1951] Specific actions

[1952] When a user speaks to the robot, their voice is converted into data by a microphone, and the data is temporarily stored in a buffer on the device.

[1953] Step 2: Converting speech data to text and analyzing sentiment

[1954] Terminal

[1955] Input: Captured digital audio data.

[1956] Processing: The speech recognition engine converts the voice data into text data, and the emotion engine analyzes the voice data to recognize the emotional state.

[1957] Output: Text data "Please tell me a nearby restaurant" and emotion information "I'm in trouble."

[1958] Specific actions

[1959] A speech recognition engine (e.g., Google Cloud Speech-to-Text API) converts the speech into text, generating the string "Please tell me where to find a nearby restaurant." At the same time, an emotion engine (e.g., IBM Watson Tone Analyzer) analyzes the speech data and generates an emotion tag such as "I'm in trouble."

[1960] Step 3: Sending text data and emotion information to the server

[1961] Communication Module

[1962] Input: Text data and emotion information.

[1963] Processing: Sends an HTTP POST request to the server.

[1964] Output: Text data and emotion information sent to the server.

[1965] Specific actions

[1966] The communication module generates an HTTP POST request containing text data and emotion information in the payload and sends it to the server's API endpoint.

[1967] POST / api / v1 / parse

[1968] Host: example.com

[1969] Content-Type: application / json

[1970] {

[1971] "text": "Can you tell me nearby restaurants?",

[1972] "emotion": "troubled"

[1973] }

[1974] Step 4: Analyzing text data and generating responses using AI

[1975] server

[1976] Input: Text data and emotion information.

[1977] Processing: The analytics module analyzes the text data to determine the user's intent, and the generative AI model generates an appropriate response.

[1978] Output: Response text "Listing restaurants within 1km of here."

[1979] Specific actions

[1980] The analysis module analyzes the text data and extracts the user's intent, which is "I want to know about nearby restaurants." The following prompt sentence is used to generate the response:

[1981] The user says, "What restaurants are nearby?" Generate an appropriate response for this user.

[1982] The AI ​​model (e.g., OpenAI GPT-3) generates a response text such as "I will list restaurants within 1 km of here."

[1983] Step 5: Adjusting responses based on emotional information

[1984] server

[1985] Input: Response text and sentiment information.

[1986] Processing: The emotion regulation module adjusts the response based on the user's emotions.

[1987] Output: Tailored response text: "I'll list restaurants within 1km of here. Let me know if I can help you."

[1988] Specific actions

[1989] The emotion adjustment module adjusts the response based on the emotion tag "distressed," resulting in the final response: "I'll list restaurants within 1km of here. Let me know if I can help you."

[1990] Step 6: Sending adjusted response data to the device

[1991] server

[1992] Input: The adjusted response text.

[1993] Processing: Send to the terminal as an HTTP response.

[1994] Output: The adjusted response text sent to the terminal.

[1995] Specific actions

[1996] The server generates an HTTP response and sends it back to the device, including the adjusted response text.

[1997] HTTP / 1.1 200 OK

[1998] Content-Type: application / json

[1999] {

[2000] "response": "I'll list restaurants within 1km of here. Let me know if I can help you."

[2001] }

[2002] Step 7: Transcribing response data

[2003] Terminal

[2004] Input: The adjusted response text.

[2005] Processing: The speech synthesis engine converts the text data into speech data.

[2006] Output: Speech data saying "I'll list restaurants within 1km of here. Let me know if I can help you."

[2007] Specific actions

[2008] The adjusted response text is sent to a speech synthesis engine (e.g., Amazon Polly) to obtain the corresponding voice data, which is then saved as an audio file for playback.

[2009] Step 8: Audible response to the user

[2010] Terminal

[2011] Input: Audio data.

[2012] Processing: Outputs sound through the speaker.

[2013] Output: An audio response that the user can hear.

[2014] Specific actions

[2015] A voice comes through the speaker and tells the user, "We'll list restaurants within 1 km of here. Let us know if we can help you."

[2016] (Application example 2)

[2017] 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."

[2018] Current voice input systems are unable to respond taking into account the user's emotional state, and therefore often provide inappropriate information to users. Furthermore, guidance services in brick-and-mortar stores are also required to understand what the user is looking for and provide appropriate information, but no system exists that can achieve this. This poses the challenge of making it difficult to improve the user experience and provide efficient guidance.

[2019] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means including a generative AI for analyzing text data, means for adjusting a response generated based on the user's emotional information, and means for receiving the analysis results from the server. This makes it possible to provide appropriate information that takes the user's emotions into consideration.

[2020] A "means for receiving voice input" is a device or function that captures voice from a user and converts it into input data.

[2021] The "means for converting speech input into text data" refers to a technology or device that converts received speech input into text data using speech recognition technology.

[2022] "Means for analyzing user emotions" refers to a technology or device that analyzes and identifies the user's emotional state from voice data or text data.

[2023] The "means for transmitting text data and emotion information to a server" is a communication device or function for transmitting the analyzed text data and emotion information to a server.

[2024] "Means including a generative AI on a server for analyzing text data" refers to a device or system that includes a generative AI function on a server for analyzing received text data and generating an appropriate response.

[2025] The "means for adjusting the response generated based on the user's emotional information" refers to a function or device that adjusts the response based on the analysis results in accordance with the user's emotional state.

[2026] The "means for receiving the analysis results from the server" is a communication device or function for receiving the analysis results provided by the server.

[2027] "Means for converting analysis results into audio data" refers to a technology or device that converts text-based responses into audio data.

[2028] The "means for providing audio data to the user" refers to a playback device or function such as a speaker or headphone that allows the user to listen to the converted audio data.

[2029] To implement this invention, a system is required that executes each step of voice input, voice recognition, emotion analysis, response generation using generative AI, response adjustment, and voice synthesis. The system consists of a terminal that interacts with the user and a server that analyzes data and generates responses.

[2030] Hardware and Software Configuration

[2031] Device:

[2032] 1. Microphone: Captures audio input from the user.

[2033] 2. Speech recognition engine: Converts captured voice into text data. This function uses the Google Speech-to-Text API.

[2034] 3. Emotion engine: Analyzes the user's emotions from voice input. This analysis uses the Microsoft Azure Emotion API.

[2035] 4. Communication module: Sends text data and emotion information to the server using HTTP and REST API as the communication protocol.

[2036] 5. Speech synthesis engine: Converts the analysis results received from the server into voice data. This function uses the Google Text-to-Speech API.

[2037] 6. Speaker: Provides the converted audio data to the user.

[2038] server:

[2039] 1. Analysis module: Analyzes the received text data using natural language processing technology.

[2040] 2. Generative AI: Generates appropriate responses based on the analysis results. The OpenAI GPT-3 model is used for generative AI.

[2041] 3. Emotion Adjustment Module: Adjusts the generated response based on the user's emotional state. For example, if the user is in a distressed emotional state, the response will be changed to a more considerate one.

[2042] Data processing and calculation

[2043] The device receives voice data and converts it into text data using a voice recognition engine. At the same time, the emotion engine analyzes the emotional information. This data is then sent to the server via a communication module.

[2044] On the server, the analysis module analyzes the text data, and the generative AI generates a response. The generated response is adjusted by the emotion adjustment module based on the user's emotional state. The final response is sent back to the device.

[2045] The response text data returned to the terminal is converted into voice data by a voice synthesis engine and provided to the user through a speaker.

[2046] Specific examples

[2047] The user asks aloud, "Where is the recommended sweets corner?" This voice is captured by a microphone and converted into text data by a speech recognition engine. At the same time, an emotion engine analyzes the user's emotional state (distress). The text data and emotion information are sent to the server.

[2048] The server's analysis module analyzes the text data, and the generative AI generates an appropriate response. This response is then adjusted by the emotion adjustment module to, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions." This adjusted response is sent to the device, where it is converted into voice data by the voice synthesis engine. It is then provided to the user via the speaker.

[2049] Prompt Sentence Examples

[2050] "A customer is asking, 'Where's the recommended snack section?' Distressed emotion detected. Suggest an appropriate response."

[2051] The above is an embodiment of the present invention.

[2052] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2053] Step 1:

[2054] Receiving voice input

[2055] The user speaks into the microphone on their smartphone or smart glasses, asking, "Where is the recommended sweets corner?"

[2056] Input: User's voice data.

[2057] What it does: The microphone captures sound and stores it as audio data.

[2058] Output: The captured audio data.

[2059] Step 2:

[2060] Speech data text conversion and sentiment analysis

[2061] The device sends the captured voice data to a speech recognition engine to convert it into text data, while an emotion engine analyzes the voice data to identify the user's emotional state.

[2062] Input: Captured audio data.

[2063] How it works: The speech recognition engine converts the speech into text data such as "Where is the recommended snack corner?" The emotion engine analyzes the speech data and generates emotion information such as "I'm in trouble."

[2064] Output: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[2065] Step 3:

[2066] Sending text data and emotional information

[2067] The terminal transmits the converted text data and emotion information to the server via the communication module.

[2068] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[2069] Operation: The communication module sends text data and emotion information as an HTTP request to the server's API endpoint.

[2070] Output: Text data and emotion information sent to the server.

[2071] Step 4:

[2072] Analysis of text data and response generation using generative AI

[2073] The server uses an analysis module to analyze the received text data and determine the user's intent. Next, a generative AI generates an appropriate response based on the analysis results.

[2074] Input: Text data "Where is the recommended sweets section?" and emotional information "I'm in trouble."

[2075] How it works: The analysis module analyzes the text data and determines the user's intent (the user wants to know where the candy corner is). The generative AI generates a response text such as "The candy corner is in the center of the third floor."

[2076] Output: Response text "The candy corner is in the center of the third floor."

[2077] Step 5:

[2078] Response adjustment based on emotional information

[2079] The server passes the generated response text to an emotion adjustment module, which adjusts the response based on the user's emotional state.

[2080] Input: Emotion information "I'm in trouble" and response text "The candy corner is in the center of the third floor."

[2081] How it works: The emotion adjustment module adjusts the response text depending on the emotion, resulting in a final response of "The candy corner is in the center of the third floor. Let me know if you need any further directions."

[2082] Output: Adjusted response text: "The candy section is in the center of the third floor. Let me know if you need any further directions."

[2083] Step 6:

[2084] Sending adjusted response data

[2085] The server sends the adjusted response text to the terminal as an HTTP response.

[2086] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[2087] Operation: The server sends the response text to the terminal as an HTTP response.

[2088] Output: The adjusted response text sent to the terminal.

[2089] Step 7:

[2090] Converting response data to audio

[2091] The terminal sends the received response text to a speech synthesis engine and converts it into voice data.

[2092] Input: Adjusted response text: "The candy section is located in the center of the third floor. Let me know if you need any further directions."

[2093] How it works: A speech synthesis engine converts text data into speech.

[2094] Output: Speech data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[2095] Step 8:

[2096] Voice response to the user

[2097] The terminal ultimately provides a voice response to the user through a speaker.

[2098] Input: Voice data saying, "The sweets corner is in the center of the third floor. Please let me know if you need any further directions."

[2099] Action: The speaker plays the audio data and provides a response to the user.

[2100] Output: A voice description that says, "The sweets corner is in the center of the third floor. Let me know if you need any further directions."

[2101] 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.

[2102] 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.

[2103] 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.

[2104] 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.

[2105] FIG. 9 illustrates 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 behaviors 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.

[2106] 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.

[2107] 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).

[2108] 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.

[2109] 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."

[2110] 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.

[2111] 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).

[2112] 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.

[2113] 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.

[2114] 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.

[2115] 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.

[2116] 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.

[2117] 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.

[2118] 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.

[2119] 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.

[2120] 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.

[2121] 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.

[2122] The following is further disclosed regarding the above embodiment.

[2123] (Claim 1)

[2124] means for receiving audio input;

[2125] means for converting voice input into text data;

[2126] A means for including a generative AI for analyzing text data in a server;

[2127] A means for receiving the analysis results from the server;

[2128] A means for converting the analysis results into audio data;

[2129] means for providing audio data to a user;

[2130] A system including:

[2131] (Claim 2)

[2132] 2. The system of claim 1, wherein the generative AI includes a server for providing information based on a user's voice instructions.

[2133] (Claim 3)

[2134] 10. The system of claim 1, including a terminal equipped with a speech recognition engine and a speech synthesis engine.

[2135] "Example 1"

[2136] (Claim 1)

[2137] means for receiving audio input;

[2138] means for converting voice input into text data;

[2139] means for transmitting text data to a server;

[2140] A means for analyzing text data and generating a response using generative artificial intelligence;

[2141] means for receiving response data from the server;

[2142] means for converting the received response data into voice data;

[2143] means for providing audio data to a user;

[2144] A system including:

[2145] (Claim 2)

[2146] 2. The system according to claim 1, including a server in which the generative artificial intelligence analyzes text data and provides information generated based on user requests.

[2147] (Claim 3)

[2148] 10. The system of claim 1, including a terminal equipped with a speech recognition engine and a speech synthesis engine.

[2149] "Application Example 1"

[2150] (Claim 1)

[2151] means for receiving audio input;

[2152] means for converting voice input into text data;

[2153] A means for including a generative AI for analyzing text data in a server;

[2154] A means for receiving the analysis results from the server;

[2155] A means for converting the analysis results into audio data;

[2156] means for providing audio data to a user;

[2157] A means for providing audio guidance of in-store information based on a question from a user in a physical store;

[2158] A system including:

[2159] (Claim 2)

[2160] The system of claim 1, including a server for the generative AI to provide information based on a user's voice instructions in a physical store.

[2161] (Claim 3)

[2162] 10. The system of claim 1, including a terminal equipped with a speech recognition engine and a speech synthesis engine.

[2163] "Example 2: Combining Emotion Engines"

[2164] (Claim 1)

[2165] means for receiving audio input;

[2166] means for converting voice input into text data;

[2167] means for analyzing voice data and recognizing user emotions;

[2168] means for transmitting text data and emotion information;

[2169] a server including means for analyzing the text data using generative artificial intelligence to generate an appropriate response;

[2170] means for adjusting the generated response based on the user's emotions;

[2171] a means for receiving the adjusted analysis results;

[2172] A means for converting the analysis results into audio data;

[2173] means for providing audio data to a user;

[2174] A system including:

[2175] (Claim 2)

[2176] 2. The system of claim 1, wherein the generative artificial intelligence includes a server for providing information based on a user's voice instructions and emotional information.

[2177] (Claim 3)

[2178] 10. The system of claim 1, comprising a terminal equipped with a speech recognition engine, an emotion recognition engine, and a speech synthesis engine.

[2179] "Application example 2 when combining emotion engines"

[2180] (Claim 1)

[2181] means for receiving audio input;

[2182] means for converting voice input into text data;

[2183] means for analyzing user emotions;

[2184] means for transmitting text data and emotion information to a server;

[2185] A means for including a generative AI for analyzing text data in a server;

[2186] means for adjusting the generated response based on the user's emotional information;

[2187] A means for receiving the analysis results from the server;

[2188] A means for converting the analysis results into audio data;

[2189] means for providing audio data to a user;

[2190] A system including:

[2191] (Claim 2)

[2192] 2. The system of claim 1, including a server for the generative AI to provide information based on a user's voice instructions and emotional information.

[2193] (Claim 3)

[2194] 10. The system of claim 1, including a terminal equipped with a speech recognition engine and a speech synthesis engine. [Explanation of symbols]

[2195] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving audio input; means for converting voice input into text data; A means for including a generative AI for analyzing text data in a server; A means for receiving the analysis results from the server; A means for converting the analysis results into audio data; means for providing audio data to a user; A system including:

2. 2. The system of claim 1, wherein the generative AI includes a server for providing information based on a user's voice instructions.

3. 10. The system of claim 1, including a terminal equipped with a speech recognition engine and a speech synthesis engine.

Images