System

Abstract

To provide an advanced multilingual conversation system corresponding to various needs.SOLUTION: The system includes means for receiving voice input from a user, means for converting the voice input into text form, means for analyzing the user's intention from the text-form data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data by voice synthesis from the translated text, and means for transmitting the voice data to a terminal of the user and causing the terminal to reproduce the voice data.SELECTED DRAWING: None

Description

[Technical Field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In modern society, smooth communication between people who speak different languages ​​is extremely important. However, current technology lacks multilingual voice conversation systems, making it particularly difficult to conduct natural conversations across a wide range of topics, from everyday conversations to specialized content. Furthermore, there are only a limited number of systems that can provide natural conversations for elderly people living alone. There is a need to solve these problems and provide an advanced multilingual conversation system that can meet diverse needs. [Means for solving the problem]

[0005] This invention solves these problems with a system that includes a means for receiving voice input from a user and converting it into text format, a means for analyzing the user's intent from the text data and generating an appropriate response, a means for translating the generated response into another language, a means for generating the translated text as voice data using voice synthesis, and a means for transmitting the voice data to the user's device and playing it back. This system receives voice input using a microphone on the user's device and converts the voice to text using a cloud-based voice recognition service. It also uses a natural language processing engine to analyze the user's intent and generate an appropriate response, covering a range of conversations from everyday to specialized. Using a translation engine and a voice synthesis engine, the system provides responses in the user's preferred language with native pronunciation, allowing for natural conversation even for elderly people living alone.

[0006] "User" refers to a person who provides voice input to the system and receives voice responses from the system.

[0007] "Terminal" refers to a device that allows a user to input speech and plays back a synthesized speech response. Examples include smartphones and tablets.

[0008] "Server" refers to a remote computer system that performs back-end processing such as voice data processing, text conversion, natural language processing, translation, and speech synthesis.

[0009] "Voice input" refers to voice data spoken by a user into a terminal.

[0010] A "microphone" refers to a device that captures the user's voice and supplies it to a terminal as voice data.

[0011] A "speech recognition engine" refers to software or a service that converts voice data into text. A speech recognition engine analyzes voice input and generates corresponding text data.

[0012] "Text format" refers to a data format expressed as a string of characters converted by a speech recognition engine.

[0013] A "natural language processing engine" refers to software or algorithms that analyze user intent from text data and generate appropriate responses.

[0014] A "translation engine" refers to software or services that convert text data into different languages.

[0015] A "speech synthesis engine" refers to software or a service that converts text data into voice data and generates voice files.

[0016] "Audio Data" refers to digital audio files generated by a speech synthesis engine for playback to a user.

[0017] "Audio File" means a digital file format in which sound data is stored, including, for example, WAV and MP3 formats.

[0018] "Natural language processing" is a general term for technologies that analyze text data and generate responses, and refers to technologies for understanding and processing human language.

[0019] "Cloud-based services" refers to services that are delivered on remote servers over the internet, including cloud-based speech recognition and translation services. [Brief explanation of the drawings]

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

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

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

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

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

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

[0026] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

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

[0028] [First embodiment]

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

[0030] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

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

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

[0033] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

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

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

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

[0037] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

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

[0039] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

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

[0041] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[0042] Functionality Overview

[0043] 1. Voice Input and Speech Recognition

[0044] When a user speaks into a device, the device's microphone captures this voice, and the captured voice data is saved as a digital audio file (e.g., WAV format).

[0045] The device sends this audio file to the server, where necessary metadata (e.g., user ID, timestamp, etc.) is added to the audio data.

[0046] 2. Voice Recognition

[0047] The server analyzes the received voice data using a cloud-based speech recognition service, such as the Google® Speech-to-Text API or a similar service, to convert the voice data into text.

[0048] 3. Natural Language Processing and Response Generation

[0049] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[0050] The server generates an appropriate response based on the analysis results. For example, if the user says hello, the server generates a text response such as "Hello!"

[0051] 4. Language Conversion

[0052] The server translates the generated response text into the language specified by the user. This process uses a translation engine (e.g., Google Translate API or DeepL) to translate the text data.

[0053] The translated text data is saved and the next step, speech synthesis, is carried out.

[0054] 5. Speech Synthesis

[0055] The server generates the translated text as audio data using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API.

[0056] The generated audio data is saved as a file in MP3 format or similar.

[0057] 6. Sending and Playing Audio Data

[0058] The server transmits the generated voice data to the terminal.

[0059] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[0060] Specific examples

[0061] Examples of everyday conversation

[0062] User: "Hello"

[0063] The device captures the audio, stores it and sends it to a server.

[0064] The server converts the speech to text "Hello" and uses natural language processing to recognize it as a greeting. It generates the appropriate response: "Hello!"

[0065] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[0066] The device receives and plays the audio data. The user hears "Hello!"

[0067] Examples of professional conversations

[0068] User: "Tell me about the advances in pacemakers."

[0069] The device captures the audio, stores it and sends it to a server.

[0070] The server converts the speech to text and uses a natural language processing engine with medical expertise to generate a response, such as, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life."

[0071] The server translates the response into the specified language (e.g., English) and generates audio data using a speech synthesis engine.

[0072] The terminal receives and plays the audio data, and the user hears the audio containing the appropriate expertise.

[0073] Thus, the system of the present invention provides a comprehensive means for facilitating multilingual conversations, allowing users to receive responses in a natural and understandable manner.

[0074] The processing flow will be explained below.

[0075] Step 1:

[0076] The user speaks to the terminal. Specifically, the user says "hello."

[0077] Step 2:

[0078] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[0079] Step 3:

[0080] The device sends the captured audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[0081] Step 4:

[0082] The server receives the voice data and passes it to a cloud-based voice recognition service.

[0083] Step 5:

[0084] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format, such as "Hello."

[0085] Step 6:

[0086] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent.

[0087] Step 7:

[0088] The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the response generated is "hello!"

[0089] Step 8:

[0090] The server passes the generated response text to a translation engine, which translates the response text into the specified language using a translation engine (e.g., Google Translate API or DeepL). For example, "Hello!" is translated to "Hello!"

[0091] Step 9:

[0092] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[0093] Step 10:

[0094] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[0095] Step 11:

[0096] The server sends the audio file to the device, encoded in the appropriate data format.

[0097] Step 12:

[0098] The device decodes the received audio file and plays it through the speaker. The user hears "Hello!" being played from the device.

[0099] This series of steps creates a system that allows users to input voice and receive natural-sounding responses in different languages.

[0100] Example 1

[0101] 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."

[0102] In today's global society, smooth communication across multiple languages ​​is important. However, due to the lack of means for users to easily converse in different language environments, there is a demand for a multilingual system based on real-time voice input.

[0103] 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.

[0104] In this invention, the server includes means for receiving voice input from a user, means for saving the voice input as a digital voice file and transmitting it to the server, means for converting the voice data into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into multiple languages, means for generating voice data from the translated text by speech synthesis, and means for transmitting the voice data to the user's terminal and playing it back, thereby enabling users to smoothly engage in real-time voice conversations in different languages.

[0105] "Voice input" refers to capturing voice information spoken by a user using an input device such as a microphone on a device.

[0106] A "digital audio file" is a file that has been converted from analog audio data into a digital format and saved, and is usually saved in a format such as WAV or MP3.

[0107] A "server" is a computer system that processes, stores, and transfers data over a network.

[0108] "Speech Recognition Service" means a cloud-based service for converting voice data into text data using certain algorithms and machine learning models.

[0109] "Text format" is a format in which voice data and other data are expressed as character string information.

[0110] A "natural language processing engine" is an algorithm or machine learning model that understands and processes text data and is used to analyze user intent.

[0111] "Multilingual translation" is the process of converting text written in one language into another language.

[0112] "Speech synthesis" is a technology that converts text data into voice data, and the synthesized voice is reproduced as a voice that resembles a human voice.

[0113] "Playback" means decoding the stored audio data and outputting it as audio again through an output device such as a speaker or headphones.

[0114] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[0115] First, when a user speaks into a terminal, the microphone, which is an input device of the terminal, captures this voice. Specifically, consider the case where the user asks, "What's the weather going to be like tomorrow?" The captured voice data is saved as a digital voice file (e.g., WAV format). The terminal then sends this voice file to the server. Metadata such as the user ID and timestamp are added to the voice data, allowing accurate processing on the server side.

[0116] The server converts the received voice data into text using a cloud-based speech recognition service (e.g., a general speech recognition API). The converted text data is stored on the server in the form of "What's the weather going to be like tomorrow?" The server then passes this text data to a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. Through this analysis process, the server understands that the user wants to know the weather for the next day.

[0117] Based on the understanding, the server generates a response text of "The weather tomorrow is sunny." The response text may be generated by retrieving the latest weather information from a pre-installed weather data API. This response can then be translated into multiple languages, so the server uses a translation engine (e.g., a common translation API) to convert the text to "The weather tomorrow is sunny."

[0118] Once the translation is complete, the server passes the translated text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data. This audio data is saved in MP3 format and temporarily stored on the server. In the final step, the server sends the generated audio data to the device. The device decodes the received audio data and plays it through the speaker.

[0119] This system allows users to smoothly converse in real time across different languages.

[0120] As a concrete example, consider the case where a user says, "Tell me about the advances in pacemakers." The device captures the speech, saves it, and sends it to the server. The server converts the speech to text and uses a natural language processing engine to generate a response using medical expertise. The generated text might be something like, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life." The server translates the response into a specified language (e.g., English) and generates audio data using a speech synthesis engine. The device receives the audio data and plays it back. The user can listen to audio containing expert knowledge.

[0121] In this way, the system of the present invention allows users to ask questions in their own language and receive responses in multiple languages, which is extremely convenient as users can obtain information in a natural conversational format and smoothly carry out conversations in multiple languages.

[0122] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0123] Step 1:

[0124] Capture and transmit audio input

[0125] The user speaks into the device, for example asking a question like "What's the weather like tomorrow?"

[0126] The device's microphone captures the user's voice and stores it as a digital audio file (WAV format).

[0127] The device adds metadata such as the user ID and timestamp to the saved audio file and sends it to the server. This data package is encrypted using the SSL / TLS protocol.

[0128] Input: User's voice

[0129] Output: Digital audio file (with metadata) sent to a server

[0130] Step 2:

[0131] Speech recognition and text conversion

[0132] The server receives the received audio data package and analyzes and identifies the metadata.

[0133] The server sends the voice data to a speech recognition service (a general speech recognition API) and converts it into text. The converted text data is saved in the form of "What's the weather like tomorrow?"

[0134] Input: Digital audio file (with metadata)

[0135] Output: Speech content in text format (e.g., "What's the weather like tomorrow?")

[0136] Step 3:

[0137] Natural language processing of text data

[0138] The server then inputs the converted text data into a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. BERT analyzes the meaning of the sentence, and the generative AI model generates an appropriate response.

[0139] The natural language processing engine understands the intent of "I want to know tomorrow's weather" and provides analysis results as a preliminary step to generating an appropriate response.

[0140] Input: Speech content in text form (e.g., "What's the weather like tomorrow?")

[0141] Output: User intent analysis results

[0142] Step 4:

[0143] Response Generation and Translation

[0144] Based on the analysis results, the server generates a response text "Tomorrow's weather will be sunny." This response may also retrieve the latest weather information from a pre-installed weather data API.

[0145] The server uses a common translation API to translate the response text into the specified language. For example, the response text "Tomorrow's weather is sunny" is translated to "The weather tomorrow is sunny."

[0146] Input: User intent analysis results

[0147] Output: The translated response text (e.g., "The weather tomorrow is sunny")

[0148] Step 5:

[0149] Speech synthesis

[0150] The server passes the translated response text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data, which is saved in MP3 format.

[0151] The server temporarily stores the generated voice data and proceeds to the next transmission step.

[0152] Input: Translated response text (e.g., "The weather tomorrow is sunny")

[0153] Output: Audio data (e.g. "The weather tomorrow is sunny" in MP3 format)

[0154] Step 6:

[0155] Sending and playing audio data

[0156] The server then sends the generated audio data to the device, which is also encrypted using the SSL / TLS protocol.

[0157] The device decodes the received audio data and plays it through the speaker at a volume and quality that is easy for the user to hear.

[0158] Input: Audio data (MP3 format)

[0159] Output: The audio played (e.g. "The weather tomorrow is sunny")

[0160] (Application example 1)

[0161] 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."

[0162] Conventional food delivery systems often pose obstacles when ordering from multilingual users. In particular, language differences can lead to misunderstandings and delays in ordering, impairing the user experience. This has created a need for systems that can accept orders quickly and accurately in regions and situations where multilingual support is difficult.

[0163] 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.

[0164] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's smartphone and playing it back, and means for realizing a dialogue for placing a food delivery order in multiple languages, thereby enabling multilingual users to smoothly place a food delivery order.

[0165] A "user" is an individual or corporation that uses this system to input voice information.

[0166] "Voice input" refers to the voice uttered by the user through a microphone, and is the voice data that the system analyzes.

[0167] "Text format" refers to text information that is the result of converting voice input.

[0168] "User intent" refers to the user's needs and goals expressed in speech input.

[0169] A "response" is a reply or response generated based on the results of analyzing the user's intent.

[0170] "Translation" is the act of converting text expressed in a particular language into another language.

[0171] "Speech synthesis" is a technology that converts text data into voice data.

[0172] "Audio data" refers to a digital audio file generated by voice synthesis.

[0173] A "smartphone" is a type of mobile phone, a device with multiple functions such as voice input, audio playback, and internet connection.

[0174] "Food delivery" is a service that allows users to order food and drinks and have them delivered to a specified location.

[0175] "Dialogue" refers to mutual communication between a user and a system, including voice input and voice output.

[0176] To implement this invention, it is necessary to build a multilingual food delivery system that integrates voice input, voice recognition, natural language processing, language conversion, voice synthesis, and voice playback. Specifically, this system is configured as follows:

[0177] System Configuration

[0178] 1. User Device and Voice Input

[0179] A user places an order by voice using a smartphone, which has a built-in microphone that serves to receive voice input. This voice data is captured and saved as a digital audio file.

[0180] 2. Voice Recognition

[0181] The device sends the captured voice data to a server, which converts it into text using a cloud-based speech recognition service (e.g., Google Speech-to-Text API). This step also adds metadata (e.g., user ID, timestamp, etc.) for processing.

[0182] 3. Natural Language Processing

[0183] The server then passes the converted text data to a natural language processing engine, which uses technologies such as BERT or GPT, to analyze the user's intent. Based on the analysis results, the server generates an appropriate response.

[0184] 4. Language Conversion

[0185] After the response is generated, the server translates it into another language using a translation engine (e.g., Google Translate API or DeepL). The translated text data is then passed to the next step, speech synthesis.

[0186] 5. Speech Synthesis

[0187] The server generates audio data from the translated text using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, and saves the audio data as a file in MP3 format or similar.

[0188] 6. Sending and Playing Audio Data

[0189] The server sends the generated audio data to the user's smartphone, which then decodes the data and plays it through the speaker.

[0190] Specific use cases

[0191] The series of steps in this system will be explained using the following specific example.

[0192] Usage example

[0193] Example prompt: "Hello, can I order? I'd like a pizza and a Coke."

[0194] 1. The user says, "Hello, I'd like to place an order. I'd like a pizza and a Coke."

[0195] 2. The smartphone's microphone captures the audio and sends it to the server as audio data.

[0196] 3. The server performs speech recognition and converts the speech into text: "Hello, can I place my order? I'd like a pizza and a Coke."

[0197] 4. The natural language processing engine analyzes the user's order and generates the appropriate response: "Your order for one pizza and one coke has been accepted."

[0198] 5. The response text is translated into another language (e.g., English), resulting in the text "Your order for one pizza and one coke has been received."

[0199] 6. The translated text is synthesized into speech to create the following audio data: "Your order for one pizza and one coke has been received."

[0200] 7. The audio is sent to your smartphone and played back as audio.

[0201] In this way, multilingual users can smoothly place food delivery orders.

[0202] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0203] Step 1:

[0204] The user inputs voice through the smartphone's microphone. The input voice might be something like, "Hello, can I order? I'd like a pizza and a Coke." The smartphone saves this voice data as a digital audio file in WAV format or similar. The input here is the user's voice, and the output is a digital audio file.

[0205] Step 2:

[0206] The device sends the saved audio data to the server. The data sent includes not only the audio data but also metadata such as the user ID and timestamp. The input here is the digital audio file and metadata, and the output is data sent to the server.

[0207] Step 3:

[0208] The server receives the audio data and converts it into text using a cloud-based speech recognition service, such as the Google Speech-to-Text API. The input is a digital audio file, and the output is the converted text data.

[0209] Step 4:

[0210] The server passes the text data to a natural language processing engine to analyze the user's intent. Generative AI models such as BERT and the GPT series are used to read the user's request from the text data. For example, it recognizes that the user wants to place an order. The input is the text data, and the output is the analyzed user's intent and the generated response.

[0211] Step 5:

[0212] The server generates a response based on the user's intent. For example, if the user says, "I would like to order a pizza and a Coke," the server generates a response such as, "Your order for one pizza and one Coke has been accepted." The input is the parsed user intent, and the output is the generated response text.

[0213] Step 6:

[0214] The server translates the generated response text into the specified language. It uses translation services such as Google Translate API and DeepL to convert the text into multiple languages. For example, the Japanese response "Your order for one pizza and one coke has been received" is translated into English and output as "Your order for one pizza and one coke has been received." The input is the generated response text, and the output is the translated text.

[0215] Step 7:

[0216] The server generates audio data from the translated text using a speech synthesis engine. It uses Amazon Polly or the Google Text-to-Speech API to create an audio file (e.g., MP3 format) from the text. The input is the translated text, and the output is the generated audio data.

[0217] Step 8:

[0218] The server sends the generated audio data to the user's smartphone, which then decodes the received audio data and plays it back through the speaker. The input here is the generated audio data, and the output is the audio playback.

[0219] In this way, by performing specific data processing and calculations at each step, multilingual users can smoothly place food delivery orders.

[0220] 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.

[0221] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[0222] Functionality Overview

[0223] 1. Voice Input and Emotion Recognition

[0224] When a user speaks into a device, the device's microphone captures this sound, which is then saved as a digital audio file (e.g., WAV format).

[0225] The device sends the saved audio file to the server, where necessary metadata (e.g., user ID, timestamp) is added to the audio data.

[0226] 2. Speech Recognition and Emotion Analysis

[0227] The server then analyzes the received voice data using a cloud-based speech recognition service, such as the Google Speech-to-Text API or a similar service, to convert the voice data into text.

[0228] At the same time, the server uses an emotion engine to analyze the pitch, rate, and volume of the voice data to recognize the user's emotions, which identify emotional states such as joy, sadness, anger, surprise, etc.

[0229] 3. Natural Language Processing and Response Generation

[0230] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[0231] The server generates an appropriate response based on the analysis, for example, if the user says "hello", the response generated is "hello!"

[0232] 4. Emotional regulation of responses

[0233] The server adjusts the response based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response will be adjusted to be kinder.

[0234] The adjusted response text is passed to a translation engine for translation into the specified language.

[0235] 5. Language Conversion and Speech Synthesis

[0236] The server translates the generated response text using a translation engine, such as Google Translate API or DeepL.

[0237] The server passes the translated text to a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, to generate audio data.

[0238] The generated audio data is saved as a file in MP3 format or similar.

[0239] 6. Sending and Playing Audio Data

[0240] The server transmits the generated voice data to the terminal.

[0241] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[0242] Specific examples

[0243] Examples of everyday conversation

[0244] User: "Hello"

[0245] The device captures the audio, stores it and sends it to a server.

[0246] The server converts the speech to text "Hello," then uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize that it is a greeting and then generates the appropriate response: "Hello!"

[0247] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[0248] The device plays "Hello!", which the user hears.

[0249] Examples of professional conversations

[0250] User: "Tell me about the advances in pacemakers."

[0251] The device captures the audio, stores it and sends it to a server.

[0252] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to generate a response using medical expertise, such as "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[0253] The server translates the generated response into the specified language (e.g., English) and generates audio data using a speech synthesis engine, adjusting the response to a gentle tone to reduce tension.

[0254] The device plays audio data such as, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased," and the user listens to it.

[0255] In this way, the system of the present invention can recognize the user's emotions and adjust responses with an appropriate tone, providing a more natural and understandable multilingual conversation.

[0256] The processing flow will be explained below.

[0257] Step 1:

[0258] The user speaks to the terminal. Specifically, the user utters, "What's the weather like this week?"

[0259] Step 2:

[0260] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[0261] Step 3:

[0262] The device sends the saved audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[0263] Step 4:

[0264] The server receives the voice data and passes it to a cloud-based voice recognition service.

[0265] Step 5:

[0266] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format. The converted text might be, for example, "What's the weather like this week?"

[0267] Step 6:

[0268] At the same time, the server passes the voice data to the emotion engine, which analyzes the pitch, speed, volume, etc. to recognize the user's emotion. For example, it may recognize that the user is excited.

[0269] Step 7:

[0270] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent. In this case, it determines that the user is looking for weather information.

[0271] Step 8:

[0272] The server generates an appropriate response based on the analysis results, for example, a text response such as "The weather this week is going to be rainy."

[0273] Step 9:

[0274] The server adjusts the generated response based on the emotion recognition results from the emotion engine, adjusting the tone of the response to calm an agitated user.

[0275] Step 10:

[0276] The server passes the adjusted response text to a translation engine (e.g., Google Translate API or DeepL), which translates the response text into the specified language. In this case, it is translated as "The weather for this week is mostly rainy."

[0277] Step 11:

[0278] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[0279] Step 12:

[0280] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[0281] Step 13:

[0282] The server sends the audio file to the device, encoded in the appropriate data format.

[0283] Step 14:

[0284] The device decodes the received audio file and plays it through the speaker. The user hears the audio being played from the device: "The weather for this week is mostly rainy."

[0285] This series of steps builds a system that allows users to receive natural responses in different languages ​​that take emotions into account in response to their input voice.

[0286] Example 2

[0287] 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."

[0288] Conventional voice conversation systems generate responses without considering the user's emotions, resulting in a poor user experience. They also struggle to achieve natural multilingual conversations, which can lead to a lack of accurate understanding of the user's intent. Furthermore, processing voice data is complex, making it difficult to respond in real time. Technology to resolve these issues is needed.

[0289] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0290] In this invention, the server includes means for receiving voice input and saving it as a digital voice file, means for transmitting the voice data to the server, means for converting the voice input into text format using a voice recognition service, means for analyzing the voice data using an emotion engine to recognize the user's emotional state, means including a natural language processing engine for analyzing the user's intention and generating an appropriate response, means for adjusting the tone of the generated response based on the emotion recognition result, means for translating the adjusted response into another language, means for generating the translated text as voice data using a voice synthesis engine, and means for transmitting the generated voice data to a terminal for playback. This enables natural, easy-to-understand multilingual conversations that take the user's emotions into consideration.

[0291] "Voice input" refers to the voice that a user utters into a terminal, and is the subject of processing by voice recognition technology.

[0292] A "digital audio file" refers to data that has been converted from analog audio into digital format and saved in formats such as WAV or MP3.

[0293] "Server" means a central processing unit that receives, processes, and generates responses from voice data, including cloud-based services.

[0294] A "speech recognition service" is a technology or service that converts voice data into text, typically using a cloud-based API.

[0295] "Text format" refers to a format in which voice data is expressed as text information, and is used in natural language processing and translation.

[0296] An "emotion engine" refers to a technology or system that analyzes and recognizes a user's emotional state from the pitch, speed, volume, etc. of voice data.

[0297] A "natural language processing engine" is a technology or system that analyzes user intent from text data and generates appropriate responses.

[0298] "Response tone" refers to the emotional and expressive modulation of the generated response voice, which changes depending on the user's emotional state.

[0299] "Translation" refers to the process of converting text data expressed in one language into another, supporting multilingual conversations.

[0300] A "speech synthesis engine" is a technology or system that converts text data into voice data, and is used to generate natural-sounding voices.

[0301] A "terminal" is a device used by a user, which has the functions of capturing audio, transmitting and receiving audio data, and playing audio.

[0302] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes user emotions. The system program uses specific hardware and software to efficiently process user voice input and provide natural, easy-to-understand multilingual responses.

[0303] First, when a user speaks into the device, the device's microphone captures the voice. This voice data is saved as a digital audio file (e.g., WAV format). The device then sends the saved audio file to the server. During this process, metadata (e.g., user ID, timestamp) is added to the voice data.

[0304] The server analyzes the received voice data using a cloud-based voice recognition service. Specific services include commonly used voice recognition APIs (e.g., Speech-to-Text APIs). This service converts the voice data into text format. At the same time, the server uses an emotion engine to analyze the pitch, speed, and volume of the voice data to recognize the user's emotional state. This emotion engine identifies emotional states such as joy, sadness, anger, and surprise.

[0305] The server then passes the converted text data to a natural language processing engine to analyze the user's intent. The natural language processing engine used here may utilize technologies from the BERT or GPT series. The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the server generates the response "hello!"

[0306] The generated response is adjusted based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response is adjusted to be gentler. This adjusted response text is passed to a translation engine and translated into the specified language. Services such as Google Translate API and DeepL are used as translation engines.

[0307] The server then passes the translated text to a speech synthesis engine, typically Amazon Polly or the Text-to-Speech API, to generate audio data, which is then saved as an audio file in MP3 format or similar.

[0308] Finally, the server sends the generated voice data to the device, which then decodes the data and plays it back through the speaker, allowing the user to hear a natural-sounding response.

[0309] Specific examples

[0310] Examples of everyday conversation

[0311] User: "Hello"

[0312] The device captures the audio, stores it and sends it to a server.

[0313] The server converts the speech to text "Hello," and uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize it as a greeting and generates the appropriate response: "Hello!"

[0314] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[0315] The device plays "Hello!", which the user hears.

[0316] Examples of professional conversations

[0317] User: "Tell me about the advances in pacemakers."

[0318] The device captures the audio, stores it and sends it to a server.

[0319] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to reference medical expertise to generate a response such as, "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[0320] The server translates the generated text into the specified language (e.g., English) and generates voice data using a speech synthesis engine, adjusting the tone to be gentler to reduce tension.

[0321] The device plays audio data to the user that says, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased."

[0322] As described above, the present invention can provide multilingual conversations in a more natural and understandable manner by recognizing the user's emotions and adjusting responses with an appropriate tone.

[0323] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0324] Step 1:

[0325] The user starts speaking into the device. The device's microphone captures this voice and saves it as a digital audio file (e.g., WAV format). The input is the user's voice data, and the output is the saved digital audio file.

[0326] Step 2:

[0327] The device sends the stored audio file and metadata (e.g., user ID, timestamp) to the server. The input is the digital audio file and metadata, and the output is the data sent to the server. Communication is secure using the HTTP protocol.

[0328] Step 3:

[0329] The server sends the received audio file to a cloud-based speech recognition service (e.g., speech recognition API) to convert the audio into text. The input is the audio file, and the output is the converted text data. The server calls the speech recognition API, and the audio data is converted into text data.

[0330] Step 4:

[0331] The server passes the converted text data to an emotion engine, which analyzes the pitch, speed, and volume of the audio data to recognize the user's emotional state. The input is the text data and audio data properties, and the output is the recognized emotional state. The server uses the emotion engine to perform polarity analysis and acoustic feature analysis.

[0332] Step 5:

[0333] The server passes the emotion recognition results and converted text data to a natural language processing engine (e.g., a generative AI model) to analyze the user's intent. The input is text data and emotion recognition results, and the output is a response text based on the user's intent. The natural language processing engine performs calculations based on the prompt sentence and generates an appropriate response.

[0334] Step 6:

[0335] The server adjusts the generated response text based on the emotion recognition results. The input is the generated response text and the emotion recognition results, and the output is the adjusted response text. The server adjusts the tone and style according to the emotion.

[0336] Step 7:

[0337] The server passes the adjusted response text to a translation engine to translate it into the specified language. The input is the adjusted response text and the output is the translated text. The translation engine (e.g., translation API) enables multilingual support.

[0338] Step 8:

[0339] The server passes the translated text to a speech synthesis engine, which generates the audio data. The input is the translated text, and the output is the generated audio data (e.g., MP3 format). The audio data is generated using a speech synthesis engine (e.g., a speech synthesis API).

[0340] Step 9:

[0341] The server sends the generated audio data to the device. The input is the generated audio data, and the output is the audio data sent to the device. The audio data is securely sent to the device using the HTTP protocol.

[0342] Step 10:

[0343] The device decodes the received audio data and plays it through the speaker. The input is the audio data sent to the device, and the output is the audio the user hears. The device decodes the audio data and plays it at the appropriate volume and quality.

[0344] (Application example 2)

[0345] 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."

[0346] Multilingual voice conversation systems often lack the ability to recognize users' emotions and adjust responses appropriately, which can lead to a decline in the quality of communication. Furthermore, security services require the ability to recognize visitors' emotions and provide accurate warnings to users, but current systems lack such advanced emotion recognition capabilities, which creates a problem of insufficient safety.

[0347] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0348] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's terminal and playing it back, means for recognizing the user's emotion and appropriately adjusting the response, and means for transmitting the emotion recognition result and the voice data for security notification. This enables a natural response that takes the user's emotion into consideration, making it possible to realize an advanced emotion-recognition security assistant that contributes to improved security.

[0349] The "means for receiving voice input" is a function for capturing the user's spoken voice with a device such as a microphone and storing it as digital data.

[0350] The "means for converting into text format" is a technology for converting voice input into text data, and is realized using a voice recognition engine.

[0351] "Means for analyzing the user's intent and generating an appropriate response" refers to a function that analyzes the content of the user's speech using natural language processing, understands the intent, and then creates an appropriate response.

[0352] A "means for translating into another language" refers to a translation engine used to translate the generated response text into another specified language.

[0353] "Means for generating voice data by voice synthesis" refers to a technology that uses a voice synthesis engine to convert text data into voice data.

[0354] The "means for transmitting audio data to a user's terminal and playing it back" is a function for transferring the generated audio data to a user's device and playing it back through a speaker or headphones.

[0355] "Means for recognizing the user's emotions and adjusting the response appropriately" refers to a system that analyzes the user's emotional state from voice data and appropriately changes the tone and content of the response based on the results.

[0356] "Means for sending emotion recognition results and voice data for security notifications" is a function for notifying the user of the results of analysis by the emotion recognition engine and related voice data in real time to their smartphone, etc.

[0357] The present invention is a security assistant system that processes user voice input in multiple languages, recognizes emotions, and generates appropriate responses. This system is realized by a combination of hardware and software.

[0358] The server receives the user's voice input using a microphone installed on the user's device. The voice data captured by the microphone is saved in digital format and then converted into text format. A cloud-based speech recognition service is used as the speech recognition engine for the voice conversion. Specifically, the Google Speech-to-Text API is used.

[0359] After the voice input is converted to text, the server uses a natural language processing engine (e.g., OpenAI's GPT-3) to analyze the user's intent and generate an appropriate response. During this analysis step, a generative AI model operates based on a prompt sentence. For example, the following prompt sentence is used:

[0360] The visitor seems angry. Generate an appropriate response.

[0361] Once the analysis is complete, the generated response is translated into another language using the Google Translate API or other translation engines. The translated text data is then converted into audio data using a speech synthesis engine (e.g., Google Text-to-Speech API). This audio data is sent to the user's device and played through the speaker.

[0362] Regarding emotion recognition, an emotion engine (e.g., IBM Watson® Tone Analyzer) is used to recognize emotions from the user's voice, and the tone and content of the generated response are adjusted according to the user's emotional state.

[0363] Furthermore, emotion recognition results and voice data are sent to the user's smartphone for security notifications. For example, if a visitor is recognized as angry, a warning message will be displayed on the user's smartphone saying, "The visitor is angry. Please be careful." This function allows the user to understand the visitor's emotional state and respond safely.

[0364] For example, if a visitor says "Open quickly!" through the doorbell, the system captures the speech and converts it into text using the Google Speech-to-Text API. The speech is then analyzed by OpenAI's GPT-3 and the emotion is recognized as "anger" using IBM Watson Tone Analyzer. The response generated is "Please wait a moment, we'll respond shortly," which is translated using the Google Translate API and then converted into speech using the Google Text-to-Speech API. Finally, the speech data and emotion recognition results are sent to the user's smartphone, prompting them to take an appropriate action.

[0365] In this way, the present invention not only recognizes the user's intention and emotion from their voice input and responds appropriately in multiple languages, but also has a function to notify the emotion recognition results to enhance security.

[0366] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0367] Step 1:

[0368] The user speaks into the terminal.

[0369] Input: User's voice

[0370] Output: Audio data (WAV format)

[0371] Specific operation: The microphone on the device captures the user's voice and saves it as a WAV audio file, which is then used for further processing.

[0372] Step 2:

[0373] The terminal transmits the voice data to the server.

[0374] Input: Audio data (WAV format)

[0375] Output: None (forwarded to server)

[0376] Specific operation: The device sends the captured audio data to the server, along with the necessary metadata (user ID, timestamp, etc.).

[0377] Step 3:

[0378] The server converts the audio data into text format.

[0379] Input: Audio data (WAV format)

[0380] Output: Text data

[0381] What happens: The server uses the Google Speech-to-Text API to convert the audio data into text, which is then used in the next analysis step.

[0382] Step 4:

[0383] The server performs emotion recognition.

[0384] Input: Audio data (WAV format)

[0385] Output: Sentiment analysis results (e.g. anger, joy, sadness, etc.)

[0386] How it works: The server uses IBM Watson Tone Analyzer to analyze the pitch, rate, and volume of the audio data to identify the user's emotions.

[0387] Step 5:

[0388] The server analyzes the user's intent from the text data and generates an appropriate response.

[0389] Input: Text data

[0390] Output: Response text

[0391] How it works: The server uses OpenAI's GPT-3 model to analyze the text data and understand the user's intent. Based on the analysis results, it generates an appropriate response in text format.

[0392] Step 6:

[0393] The server translates the generated response text into another language.

[0394] Input: Response text

[0395] Output: Translated text

[0396] Specific behavior: The server uses a translation engine such as the Google Translate API to translate the generated response text into another specified language.

[0397] Step 7:

[0398] The server generates the translated text as audio data.

[0399] Input: Translated text

[0400] Output: Audio data (MP3 format)

[0401] How it works: The server uses the Google Text-to-Speech API to convert the translated text into audio data, which is then saved in MP3 format.

[0402] Step 8:

[0403] The server sends the audio data to the terminal and plays it back.

[0404] Input: Audio data (MP3 format)

[0405] Output: None (transfer to terminal)

[0406] Specific operation: The server sends the generated audio data to the terminal, and the terminal decodes the received audio data and plays it through the speaker.

[0407] Step 9:

[0408] The server sends the emotion recognition results and voice data to the user's device for security notification.

[0409] Input: Emotion recognition results, audio data

[0410] Output: Security Notice

[0411] Specific operation: The server sends the emotion recognition results and related voice data to the user's smartphone, and displays warnings and notifications based on the user's emotional state.

[0412] 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.

[0413] 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.

[0414] 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.

[0415] [Second embodiment]

[0416] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0417] 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.

[0418] 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).

[0419] 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.

[0420] 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.

[0421] 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).

[0422] 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.

[0423] 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.

[0424] 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.

[0425] 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.

[0426] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0427] 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."

[0428] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[0429] Functionality Overview

[0430] 1. Voice Input and Speech Recognition

[0431] When a user speaks into a device, the device's microphone captures this voice, and the captured voice data is saved as a digital audio file (e.g., WAV format).

[0432] The device sends this audio file to the server, where necessary metadata (e.g., user ID, timestamp, etc.) is added to the audio data.

[0433] 2. Voice Recognition

[0434] The server then analyzes the received voice data using a cloud-based speech recognition service, such as the Google Speech-to-Text API or a similar service, to convert the voice data into text.

[0435] 3. Natural Language Processing and Response Generation

[0436] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[0437] The server generates an appropriate response based on the analysis results. For example, if the user says hello, the server generates a text response such as "Hello!"

[0438] 4. Language Conversion

[0439] The server translates the generated response text into the language specified by the user. This process uses a translation engine (e.g., Google Translate API or DeepL) to translate the text data.

[0440] The translated text data is saved and the next step, speech synthesis, is carried out.

[0441] 5. Speech Synthesis

[0442] The server generates the translated text as audio data using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API.

[0443] The generated audio data is saved as a file in MP3 format or similar.

[0444] 6. Sending and Playing Audio Data

[0445] The server transmits the generated voice data to the terminal.

[0446] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[0447] Specific examples

[0448] Examples of everyday conversation

[0449] User: "Hello"

[0450] The device captures the audio, stores it and sends it to a server.

[0451] The server converts the speech to text "Hello" and uses natural language processing to recognize it as a greeting. It generates the appropriate response: "Hello!"

[0452] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[0453] The device receives and plays the audio data. The user hears "Hello!"

[0454] Examples of professional conversations

[0455] User: "Tell me about the advances in pacemakers."

[0456] The device captures the audio, stores it and sends it to a server.

[0457] The server converts the speech to text and uses a natural language processing engine with medical expertise to generate a response, such as, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life."

[0458] The server translates the response into the specified language (e.g., English) and generates audio data using a speech synthesis engine.

[0459] The terminal receives and plays the audio data, and the user hears the audio containing the appropriate expertise.

[0460] Thus, the system of the present invention provides a comprehensive means for facilitating multilingual conversations, allowing users to receive responses in a natural and understandable manner.

[0461] The processing flow will be explained below.

[0462] Step 1:

[0463] The user speaks to the terminal. Specifically, the user says "hello."

[0464] Step 2:

[0465] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[0466] Step 3:

[0467] The device sends the captured audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[0468] Step 4:

[0469] The server receives the voice data and passes it to a cloud-based voice recognition service.

[0470] Step 5:

[0471] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format, such as "Hello."

[0472] Step 6:

[0473] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent.

[0474] Step 7:

[0475] The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the response generated is "hello!"

[0476] Step 8:

[0477] The server passes the generated response text to a translation engine, which translates the response text into the specified language using a translation engine (e.g., Google Translate API or DeepL). For example, "Hello!" is translated to "Hello!"

[0478] Step 9:

[0479] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[0480] Step 10:

[0481] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[0482] Step 11:

[0483] The server sends the audio file to the device, encoded in the appropriate data format.

[0484] Step 12:

[0485] The device decodes the received audio file and plays it through the speaker. The user hears "Hello!" being played from the device.

[0486] This series of steps creates a system that allows users to input voice and receive natural-sounding responses in different languages.

[0487] Example 1

[0488] 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."

[0489] In today's global society, smooth communication across multiple languages ​​is important. However, due to the lack of means for users to easily converse in different language environments, there is a demand for a multilingual system based on real-time voice input.

[0490] 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.

[0491] In this invention, the server includes means for receiving voice input from a user, means for saving the voice input as a digital voice file and transmitting it to the server, means for converting the voice data into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into multiple languages, means for generating voice data from the translated text by speech synthesis, and means for transmitting the voice data to the user's terminal and playing it back, thereby enabling users to smoothly engage in real-time voice conversations in different languages.

[0492] "Voice input" refers to capturing voice information spoken by a user using an input device such as a microphone on a device.

[0493] A "digital audio file" is a file that has been converted from analog audio data into a digital format and saved, and is usually saved in a format such as WAV or MP3.

[0494] A "server" is a computer system that processes, stores, and transfers data over a network.

[0495] "Speech Recognition Service" means a cloud-based service for converting voice data into text data using certain algorithms and machine learning models.

[0496] "Text format" is a format in which voice data and other data are expressed as character string information.

[0497] A "natural language processing engine" is an algorithm or machine learning model that understands and processes text data and is used to analyze user intent.

[0498] "Multilingual translation" is the process of converting text written in one language into another language.

[0499] "Speech synthesis" is a technology that converts text data into voice data, and the synthesized voice is reproduced as a voice that resembles a human voice.

[0500] "Playback" means decoding the stored audio data and outputting it as audio again through an output device such as a speaker or headphones.

[0501] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[0502] First, when a user speaks into a terminal, the microphone, which is an input device of the terminal, captures this voice. Specifically, consider the case where the user asks, "What's the weather going to be like tomorrow?" The captured voice data is saved as a digital voice file (e.g., WAV format). The terminal then sends this voice file to the server. Metadata such as the user ID and timestamp are added to the voice data, allowing accurate processing on the server side.

[0503] The server converts the received voice data into text using a cloud-based speech recognition service (e.g., a general speech recognition API). The converted text data is stored on the server in the form of "What's the weather going to be like tomorrow?" The server then passes this text data to a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. Through this analysis process, the server understands that the user wants to know the weather for the next day.

[0504] Based on the understanding, the server generates a response text of "The weather tomorrow is sunny." The response text may be generated by retrieving the latest weather information from a pre-installed weather data API. This response can then be translated into multiple languages, so the server uses a translation engine (e.g., a common translation API) to convert the text to "The weather tomorrow is sunny."

[0505] Once the translation is complete, the server passes the translated text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data. This audio data is saved in MP3 format and temporarily stored on the server. In the final step, the server sends the generated audio data to the device. The device decodes the received audio data and plays it through the speaker.

[0506] This system allows users to smoothly converse in real time across different languages.

[0507] As a concrete example, consider the case where a user says, "Tell me about the advances in pacemakers." The device captures the speech, saves it, and sends it to the server. The server converts the speech to text and uses a natural language processing engine to generate a response using medical expertise. The generated text might be something like, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life." The server translates the response into a specified language (e.g., English) and generates audio data using a speech synthesis engine. The device receives the audio data and plays it back. The user can listen to audio containing expert knowledge.

[0508] In this way, the system of the present invention allows users to ask questions in their own language and receive responses in multiple languages, which is extremely convenient as users can obtain information in a natural conversational format and smoothly carry out conversations in multiple languages.

[0509] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0510] Step 1:

[0511] Capture and transmit audio input

[0512] The user speaks into the device, for example asking a question like "What's the weather like tomorrow?"

[0513] The device's microphone captures the user's voice and stores it as a digital audio file (WAV format).

[0514] The device adds metadata such as the user ID and timestamp to the saved audio file and sends it to the server. This data package is encrypted using the SSL / TLS protocol.

[0515] Input: User's voice

[0516] Output: Digital audio file (with metadata) sent to a server

[0517] Step 2:

[0518] Speech recognition and text conversion

[0519] The server receives the received audio data package and analyzes and identifies the metadata.

[0520] The server sends the voice data to a speech recognition service (a general speech recognition API) and converts it into text. The converted text data is saved in the form of "What's the weather like tomorrow?"

[0521] Input: Digital audio file (with metadata)

[0522] Output: Speech content in text format (e.g., "What's the weather like tomorrow?")

[0523] Step 3:

[0524] Natural language processing of text data

[0525] The server then inputs the converted text data into a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. BERT analyzes the meaning of the sentence, and the generative AI model generates an appropriate response.

[0526] The natural language processing engine understands the intent of "I want to know tomorrow's weather" and provides analysis results as a preliminary step to generating an appropriate response.

[0527] Input: Speech content in text form (e.g., "What's the weather like tomorrow?")

[0528] Output: User intent analysis results

[0529] Step 4:

[0530] Response Generation and Translation

[0531] Based on the analysis results, the server generates a response text "Tomorrow's weather will be sunny." This response may also retrieve the latest weather information from a pre-installed weather data API.

[0532] The server uses a common translation API to translate the response text into the specified language. For example, the response text "Tomorrow's weather is sunny" is translated to "The weather tomorrow is sunny."

[0533] Input: User intent analysis results

[0534] Output: The translated response text (e.g., "The weather tomorrow is sunny")

[0535] Step 5:

[0536] Speech synthesis

[0537] The server passes the translated response text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data, which is saved in MP3 format.

[0538] The server temporarily stores the generated voice data and proceeds to the next transmission step.

[0539] Input: Translated response text (e.g., "The weather tomorrow is sunny")

[0540] Output: Audio data (e.g. "The weather tomorrow is sunny" in MP3 format)

[0541] Step 6:

[0542] Sending and playing audio data

[0543] The server then sends the generated audio data to the device, which is also encrypted using the SSL / TLS protocol.

[0544] The device decodes the received audio data and plays it through the speaker at a volume and quality that is easy for the user to hear.

[0545] Input: Audio data (MP3 format)

[0546] Output: The audio played (e.g. "The weather tomorrow is sunny")

[0547] (Application example 1)

[0548] 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."

[0549] Conventional food delivery systems often pose obstacles when ordering from multilingual users. In particular, language differences can lead to misunderstandings and delays in ordering, impairing the user experience. This has created a need for systems that can accept orders quickly and accurately in regions and situations where multilingual support is difficult.

[0550] 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.

[0551] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's smartphone and playing it back, and means for realizing a dialogue for placing a food delivery order in multiple languages, thereby enabling multilingual users to smoothly place a food delivery order.

[0552] A "user" is an individual or corporation that uses this system to input voice information.

[0553] "Voice input" refers to the voice uttered by the user through a microphone, and is the voice data that the system analyzes.

[0554] "Text format" refers to text information that is the result of converting voice input.

[0555] "User intent" refers to the user's needs and goals expressed in speech input.

[0556] A "response" is a reply or response generated based on the results of analyzing the user's intent.

[0557] "Translation" is the act of converting text expressed in a particular language into another language.

[0558] "Speech synthesis" is a technology that converts text data into voice data.

[0559] "Audio data" refers to a digital audio file generated by voice synthesis.

[0560] A "smartphone" is a type of mobile phone, a device with multiple functions such as voice input, audio playback, and internet connection.

[0561] "Food delivery" is a service that allows users to order food and drinks and have them delivered to a specified location.

[0562] "Dialogue" refers to mutual communication between a user and a system, including voice input and voice output.

[0563] To implement this invention, it is necessary to build a multilingual food delivery system that integrates voice input, voice recognition, natural language processing, language conversion, voice synthesis, and voice playback. Specifically, this system is configured as follows:

[0564] System Configuration

[0565] 1. User Device and Voice Input

[0566] A user places an order by voice using a smartphone, which has a built-in microphone that serves to receive voice input. This voice data is captured and saved as a digital audio file.

[0567] 2. Voice Recognition

[0568] The device sends the captured voice data to a server, which converts it into text using a cloud-based speech recognition service (e.g., Google Speech-to-Text API). This step also adds metadata (e.g., user ID, timestamp, etc.) for processing.

[0569] 3. Natural Language Processing

[0570] The server then passes the converted text data to a natural language processing engine, which uses technologies such as BERT or GPT, to analyze the user's intent. Based on the analysis results, the server generates an appropriate response.

[0571] 4. Language Conversion

[0572] After the response is generated, the server translates it into another language using a translation engine (e.g., Google Translate API or DeepL). The translated text data is then passed to the next step, speech synthesis.

[0573] 5. Speech Synthesis

[0574] The server generates audio data from the translated text using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, and saves the audio data as a file in MP3 format or similar.

[0575] 6. Sending and Playing Audio Data

[0576] The server sends the generated audio data to the user's smartphone, which then decodes the data and plays it through the speaker.

[0577] Specific use cases

[0578] The series of steps in this system will be explained using the following specific example.

[0579] Usage example

[0580] Example prompt: "Hello, can I order? I'd like a pizza and a Coke."

[0581] 1. The user says, "Hello, I'd like to place an order. I'd like a pizza and a Coke."

[0582] 2. The smartphone's microphone captures the audio and sends it to the server as audio data.

[0583] 3. The server performs speech recognition and converts the speech into text: "Hello, can I place my order? I'd like a pizza and a Coke."

[0584] 4. The natural language processing engine analyzes the user's order and generates the appropriate response: "Your order for one pizza and one coke has been accepted."

[0585] 5. The response text is translated into another language (e.g., English), resulting in the text "Your order for one pizza and one coke has been received."

[0586] 6. The translated text is synthesized into speech to create the following audio data: "Your order for one pizza and one coke has been received."

[0587] 7. The audio is sent to your smartphone and played back as audio.

[0588] In this way, multilingual users can smoothly place food delivery orders.

[0589] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0590] Step 1:

[0591] The user inputs voice through the smartphone's microphone. The input voice might be something like, "Hello, can I order? I'd like a pizza and a Coke." The smartphone saves this voice data as a digital audio file in WAV format or similar. The input here is the user's voice, and the output is a digital audio file.

[0592] Step 2:

[0593] The device sends the saved audio data to the server. The data sent includes not only the audio data but also metadata such as the user ID and timestamp. The input here is the digital audio file and metadata, and the output is data sent to the server.

[0594] Step 3:

[0595] The server receives the audio data and converts it into text using a cloud-based speech recognition service, such as the Google Speech-to-Text API. The input is a digital audio file, and the output is the converted text data.

[0596] Step 4:

[0597] The server passes the text data to a natural language processing engine to analyze the user's intent. Generative AI models such as BERT and the GPT series are used to read the user's request from the text data. For example, it recognizes that the user wants to place an order. The input is the text data, and the output is the analyzed user's intent and the generated response.

[0598] Step 5:

[0599] The server generates a response based on the user's intent. For example, if the user says, "I would like to order a pizza and a Coke," the server generates a response such as, "Your order for one pizza and one Coke has been accepted." The input is the parsed user intent, and the output is the generated response text.

[0600] Step 6:

[0601] The server translates the generated response text into the specified language. It uses translation services such as Google Translate API and DeepL to convert the text into multiple languages. For example, the Japanese response "Your order for one pizza and one coke has been received" is translated into English and output as "Your order for one pizza and one coke has been received." The input is the generated response text, and the output is the translated text.

[0602] Step 7:

[0603] The server generates audio data from the translated text using a speech synthesis engine. It uses Amazon Polly or the Google Text-to-Speech API to create an audio file (e.g., MP3 format) from the text. The input is the translated text, and the output is the generated audio data.

[0604] Step 8:

[0605] The server sends the generated audio data to the user's smartphone, which then decodes the received audio data and plays it back through the speaker. The input here is the generated audio data, and the output is the audio playback.

[0606] In this way, by performing specific data processing and calculations at each step, multilingual users can smoothly place food delivery orders.

[0607] 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.

[0608] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[0609] Functionality Overview

[0610] 1. Voice Input and Emotion Recognition

[0611] When a user speaks into a device, the device's microphone captures this sound, which is then saved as a digital audio file (e.g., WAV format).

[0612] The device sends the saved audio file to the server, where necessary metadata (e.g., user ID, timestamp) is added to the audio data.

[0613] 2. Speech Recognition and Emotion Analysis

[0614] The server then analyzes the received voice data using a cloud-based speech recognition service, such as the Google Speech-to-Text API or a similar service, to convert the voice data into text.

[0615] At the same time, the server uses an emotion engine to analyze the pitch, rate, and volume of the voice data to recognize the user's emotions, which identify emotional states such as joy, sadness, anger, surprise, etc.

[0616] 3. Natural Language Processing and Response Generation

[0617] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[0618] The server generates an appropriate response based on the analysis, for example, if the user says "hello", the response generated is "hello!"

[0619] 4. Emotional regulation of responses

[0620] The server adjusts the response based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response will be adjusted to be kinder.

[0621] The adjusted response text is passed to a translation engine for translation into the specified language.

[0622] 5. Language Conversion and Speech Synthesis

[0623] The server translates the generated response text using a translation engine, such as Google Translate API or DeepL.

[0624] The server passes the translated text to a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, to generate audio data.

[0625] The generated audio data is saved as a file in MP3 format or similar.

[0626] 6. Sending and Playing Audio Data

[0627] The server transmits the generated voice data to the terminal.

[0628] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[0629] Specific examples

[0630] Examples of everyday conversation

[0631] User: "Hello"

[0632] The device captures the audio, stores it and sends it to a server.

[0633] The server converts the speech to text "Hello," then uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize that it is a greeting and then generates the appropriate response: "Hello!"

[0634] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[0635] The device plays "Hello!", which the user hears.

[0636] Examples of professional conversations

[0637] User: "Tell me about the advances in pacemakers."

[0638] The device captures the audio, stores it and sends it to a server.

[0639] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to generate a response using medical expertise, such as "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[0640] The server translates the generated response into the specified language (e.g., English) and generates audio data using a speech synthesis engine, adjusting the response to a gentle tone to reduce tension.

[0641] The device plays audio data such as, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased," and the user listens to it.

[0642] In this way, the system of the present invention can recognize the user's emotions and adjust responses with an appropriate tone, providing a more natural and understandable multilingual conversation.

[0643] The processing flow will be explained below.

[0644] Step 1:

[0645] The user speaks to the terminal. Specifically, the user utters, "What's the weather like this week?"

[0646] Step 2:

[0647] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[0648] Step 3:

[0649] The device sends the saved audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[0650] Step 4:

[0651] The server receives the voice data and passes it to a cloud-based voice recognition service.

[0652] Step 5:

[0653] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format. The converted text might be, for example, "What's the weather like this week?"

[0654] Step 6:

[0655] At the same time, the server passes the voice data to the emotion engine, which analyzes the pitch, speed, volume, etc. to recognize the user's emotion. For example, it may recognize that the user is excited.

[0656] Step 7:

[0657] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent. In this case, it determines that the user is looking for weather information.

[0658] Step 8:

[0659] The server generates an appropriate response based on the analysis results, for example, a text response such as "The weather this week is going to be rainy."

[0660] Step 9:

[0661] The server adjusts the generated response based on the emotion recognition results from the emotion engine, adjusting the tone of the response to calm an agitated user.

[0662] Step 10:

[0663] The server passes the adjusted response text to a translation engine (e.g., Google Translate API or DeepL), which translates the response text into the specified language. In this case, it is translated as "The weather for this week is mostly rainy."

[0664] Step 11:

[0665] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[0666] Step 12:

[0667] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[0668] Step 13:

[0669] The server sends the audio file to the device, encoded in the appropriate data format.

[0670] Step 14:

[0671] The device decodes the received audio file and plays it through the speaker. The user hears the audio being played from the device: "The weather for this week is mostly rainy."

[0672] This series of steps builds a system that allows users to receive natural responses in different languages ​​that take emotions into account in response to their input voice.

[0673] Example 2

[0674] 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."

[0675] Conventional voice conversation systems generate responses without considering the user's emotions, resulting in a poor user experience. They also struggle to achieve natural multilingual conversations, which can lead to a lack of accurate understanding of the user's intent. Furthermore, processing voice data is complex, making it difficult to respond in real time. Technology to resolve these issues is needed.

[0676] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0677] In this invention, the server includes means for receiving voice input and saving it as a digital voice file, means for transmitting the voice data to the server, means for converting the voice input into text format using a voice recognition service, means for analyzing the voice data using an emotion engine to recognize the user's emotional state, means including a natural language processing engine for analyzing the user's intention and generating an appropriate response, means for adjusting the tone of the generated response based on the emotion recognition result, means for translating the adjusted response into another language, means for generating the translated text as voice data using a voice synthesis engine, and means for transmitting the generated voice data to a terminal for playback. This enables natural, easy-to-understand multilingual conversations that take the user's emotions into consideration.

[0678] "Voice input" refers to the voice that a user utters into a terminal, and is the subject of processing by voice recognition technology.

[0679] A "digital audio file" refers to data that has been converted from analog audio into digital format and saved in formats such as WAV or MP3.

[0680] "Server" means a central processing unit that receives, processes, and generates responses from voice data, including cloud-based services.

[0681] A "speech recognition service" is a technology or service that converts voice data into text, typically using a cloud-based API.

[0682] "Text format" refers to a format in which voice data is expressed as text information, and is used in natural language processing and translation.

[0683] An "emotion engine" refers to a technology or system that analyzes and recognizes a user's emotional state from the pitch, speed, volume, etc. of voice data.

[0684] A "natural language processing engine" is a technology or system that analyzes user intent from text data and generates appropriate responses.

[0685] "Response tone" refers to the emotional and expressive modulation of the generated response voice, which changes depending on the user's emotional state.

[0686] "Translation" refers to the process of converting text data expressed in one language into another, supporting multilingual conversations.

[0687] A "speech synthesis engine" is a technology or system that converts text data into voice data, and is used to generate natural-sounding voices.

[0688] A "terminal" is a device used by a user, which has the functions of capturing audio, transmitting and receiving audio data, and playing audio.

[0689] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes user emotions. The system program uses specific hardware and software to efficiently process user voice input and provide natural, easy-to-understand multilingual responses.

[0690] First, when a user speaks into the device, the device's microphone captures the voice. This voice data is saved as a digital audio file (e.g., WAV format). The device then sends the saved audio file to the server. During this process, metadata (e.g., user ID, timestamp) is added to the voice data.

[0691] The server analyzes the received voice data using a cloud-based voice recognition service. Specific services include commonly used voice recognition APIs (e.g., Speech-to-Text APIs). This service converts the voice data into text format. At the same time, the server uses an emotion engine to analyze the pitch, speed, and volume of the voice data to recognize the user's emotional state. This emotion engine identifies emotional states such as joy, sadness, anger, and surprise.

[0692] The server then passes the converted text data to a natural language processing engine to analyze the user's intent. The natural language processing engine used here may utilize technologies from the BERT or GPT series. The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the server generates the response "hello!"

[0693] The generated response is adjusted based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response is adjusted to be gentler. This adjusted response text is passed to a translation engine and translated into the specified language. Services such as Google Translate API and DeepL are used as translation engines.

[0694] The server then passes the translated text to a speech synthesis engine, typically Amazon Polly or the Text-to-Speech API, to generate audio data, which is then saved as an audio file in MP3 format or similar.

[0695] Finally, the server sends the generated voice data to the device, which then decodes the data and plays it back through the speaker, allowing the user to hear a natural-sounding response.

[0696] Specific examples

[0697] Examples of everyday conversation

[0698] User: "Hello"

[0699] The device captures the audio, stores it and sends it to a server.

[0700] The server converts the speech to text "Hello," and uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize it as a greeting and generates the appropriate response: "Hello!"

[0701] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[0702] The device plays "Hello!", which the user hears.

[0703] Examples of professional conversations

[0704] User: "Tell me about the advances in pacemakers."

[0705] The device captures the audio, stores it and sends it to a server.

[0706] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to reference medical expertise to generate a response such as, "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[0707] The server translates the generated text into the specified language (e.g., English) and generates voice data using a speech synthesis engine, adjusting the tone to be gentler to reduce tension.

[0708] The device plays audio data to the user that says, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased."

[0709] As described above, the present invention can provide multilingual conversations in a more natural and understandable manner by recognizing the user's emotions and adjusting responses with an appropriate tone.

[0710] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0711] Step 1:

[0712] The user starts speaking into the device. The device's microphone captures this voice and saves it as a digital audio file (e.g., WAV format). The input is the user's voice data, and the output is the saved digital audio file.

[0713] Step 2:

[0714] The device sends the stored audio file and metadata (e.g., user ID, timestamp) to the server. The input is the digital audio file and metadata, and the output is the data sent to the server. Communication is secure using the HTTP protocol.

[0715] Step 3:

[0716] The server sends the received audio file to a cloud-based speech recognition service (e.g., speech recognition API) to convert the audio into text. The input is the audio file, and the output is the converted text data. The server calls the speech recognition API, and the audio data is converted into text data.

[0717] Step 4:

[0718] The server passes the converted text data to an emotion engine, which analyzes the pitch, speed, and volume of the audio data to recognize the user's emotional state. The input is the text data and audio data properties, and the output is the recognized emotional state. The server uses the emotion engine to perform polarity analysis and acoustic feature analysis.

[0719] Step 5:

[0720] The server passes the emotion recognition results and converted text data to a natural language processing engine (e.g., a generative AI model) to analyze the user's intent. The input is text data and emotion recognition results, and the output is a response text based on the user's intent. The natural language processing engine performs calculations based on the prompt sentence and generates an appropriate response.

[0721] Step 6:

[0722] The server adjusts the generated response text based on the emotion recognition results. The input is the generated response text and the emotion recognition results, and the output is the adjusted response text. The server adjusts the tone and style according to the emotion.

[0723] Step 7:

[0724] The server passes the adjusted response text to a translation engine to translate it into the specified language. The input is the adjusted response text and the output is the translated text. The translation engine (e.g., translation API) enables multilingual support.

[0725] Step 8:

[0726] The server passes the translated text to a speech synthesis engine, which generates the audio data. The input is the translated text, and the output is the generated audio data (e.g., MP3 format). The audio data is generated using a speech synthesis engine (e.g., a speech synthesis API).

[0727] Step 9:

[0728] The server sends the generated audio data to the device. The input is the generated audio data, and the output is the audio data sent to the device. The audio data is securely sent to the device using the HTTP protocol.

[0729] Step 10:

[0730] The device decodes the received audio data and plays it through the speaker. The input is the audio data sent to the device, and the output is the audio the user hears. The device decodes the audio data and plays it at the appropriate volume and quality.

[0731] (Application example 2)

[0732] 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."

[0733] Multilingual voice conversation systems often lack the ability to recognize users' emotions and adjust responses appropriately, which can lead to a decline in the quality of communication. Furthermore, security services require the ability to recognize visitors' emotions and provide accurate warnings to users, but current systems lack such advanced emotion recognition capabilities, which creates a problem of insufficient safety.

[0734] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0735] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's terminal and playing it back, means for recognizing the user's emotion and appropriately adjusting the response, and means for transmitting the emotion recognition result and the voice data for security notification. This enables a natural response that takes the user's emotion into consideration, making it possible to realize an advanced emotion-recognition security assistant that contributes to improved security.

[0736] The "means for receiving voice input" is a function for capturing the user's spoken voice with a device such as a microphone and storing it as digital data.

[0737] The "means for converting into text format" is a technology for converting voice input into text data, and is realized using a voice recognition engine.

[0738] "Means for analyzing the user's intent and generating an appropriate response" refers to a function that analyzes the content of the user's speech using natural language processing, understands the intent, and then creates an appropriate response.

[0739] A "means for translating into another language" refers to a translation engine used to translate the generated response text into another specified language.

[0740] "Means for generating voice data by voice synthesis" refers to a technology that uses a voice synthesis engine to convert text data into voice data.

[0741] The "means for transmitting audio data to a user's terminal and playing it back" is a function for transferring the generated audio data to a user's device and playing it back through a speaker or headphones.

[0742] "Means for recognizing the user's emotions and adjusting the response appropriately" refers to a system that analyzes the user's emotional state from voice data and appropriately changes the tone and content of the response based on the results.

[0743] "Means for sending emotion recognition results and voice data for security notifications" is a function for notifying the user of the results of analysis by the emotion recognition engine and related voice data in real time to their smartphone, etc.

[0744] The present invention is a security assistant system that processes user voice input in multiple languages, recognizes emotions, and generates appropriate responses. This system is realized by a combination of hardware and software.

[0745] The server receives the user's voice input using a microphone installed on the user's device. The voice data captured by the microphone is saved in digital format and then converted into text format. A cloud-based speech recognition service is used as the speech recognition engine for the voice conversion. Specifically, the Google Speech-to-Text API is used.

[0746] After the voice input is converted to text, the server uses a natural language processing engine (e.g., OpenAI's GPT-3) to analyze the user's intent and generate an appropriate response. During this analysis step, a generative AI model operates based on a prompt sentence. For example, the following prompt sentence is used:

[0747] The visitor seems angry. Generate an appropriate response.

[0748] Once the analysis is complete, the generated response is translated into another language using the Google Translate API or other translation engines. The translated text data is then converted into audio data using a speech synthesis engine (e.g., Google Text-to-Speech API). This audio data is sent to the user's device and played through the speaker.

[0749] Regarding emotion recognition, an emotion engine (e.g., IBM Watson Tone Analyzer) is used to recognize emotions from the user's voice, and the tone and content of the generated response are adjusted according to the user's emotional state.

[0750] Furthermore, emotion recognition results and voice data are sent to the user's smartphone for security notifications. For example, if a visitor is recognized as angry, a warning message will be displayed on the user's smartphone saying, "The visitor is angry. Please be careful." This function allows the user to understand the visitor's emotional state and respond safely.

[0751] For example, if a visitor says "Open quickly!" through the doorbell, the system captures the speech and converts it into text using the Google Speech-to-Text API. The speech is then analyzed by OpenAI's GPT-3 and the emotion is recognized as "anger" using IBM Watson Tone Analyzer. The response generated is "Please wait a moment, we'll respond shortly," which is translated using the Google Translate API and then converted into speech using the Google Text-to-Speech API. Finally, the speech data and emotion recognition results are sent to the user's smartphone, prompting them to take an appropriate action.

[0752] In this way, the present invention not only recognizes the user's intention and emotion from their voice input and responds appropriately in multiple languages, but also has a function to notify the emotion recognition results to enhance security.

[0753] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0754] Step 1:

[0755] The user speaks into the terminal.

[0756] Input: User's voice

[0757] Output: Audio data (WAV format)

[0758] Specific operation: The microphone on the device captures the user's voice and saves it as a WAV audio file, which is then used for further processing.

[0759] Step 2:

[0760] The terminal transmits the voice data to the server.

[0761] Input: Audio data (WAV format)

[0762] Output: None (forwarded to server)

[0763] Specific operation: The device sends the captured audio data to the server, along with the necessary metadata (user ID, timestamp, etc.).

[0764] Step 3:

[0765] The server converts the audio data into text format.

[0766] Input: Audio data (WAV format)

[0767] Output: Text data

[0768] What happens: The server uses the Google Speech-to-Text API to convert the audio data into text, which is then used in the next analysis step.

[0769] Step 4:

[0770] The server performs emotion recognition.

[0771] Input: Audio data (WAV format)

[0772] Output: Sentiment analysis results (e.g. anger, joy, sadness, etc.)

[0773] How it works: The server uses IBM Watson Tone Analyzer to analyze the pitch, rate, and volume of the audio data to identify the user's emotions.

[0774] Step 5:

[0775] The server analyzes the user's intent from the text data and generates an appropriate response.

[0776] Input: Text data

[0777] Output: Response text

[0778] How it works: The server uses OpenAI's GPT-3 model to analyze the text data and understand the user's intent. Based on the analysis results, it generates an appropriate response in text format.

[0779] Step 6:

[0780] The server translates the generated response text into another language.

[0781] Input: Response text

[0782] Output: Translated text

[0783] Specific behavior: The server uses a translation engine such as the Google Translate API to translate the generated response text into another specified language.

[0784] Step 7:

[0785] The server generates the translated text as audio data.

[0786] Input: Translated text

[0787] Output: Audio data (MP3 format)

[0788] How it works: The server uses the Google Text-to-Speech API to convert the translated text into audio data, which is then saved in MP3 format.

[0789] Step 8:

[0790] The server sends the audio data to the terminal and plays it back.

[0791] Input: Audio data (MP3 format)

[0792] Output: None (transfer to terminal)

[0793] Specific operation: The server sends the generated audio data to the terminal, and the terminal decodes the received audio data and plays it through the speaker.

[0794] Step 9:

[0795] The server sends the emotion recognition results and voice data to the user's device for security notification.

[0796] Input: Emotion recognition results, audio data

[0797] Output: Security Notice

[0798] Specific operation: The server sends the emotion recognition results and related voice data to the user's smartphone, and displays warnings and notifications based on the user's emotional state.

[0799] 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.

[0800] 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.

[0801] 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.

[0802] [Third embodiment]

[0803] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0804] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0805] 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).

[0806] 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.

[0807] 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.

[0808] 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).

[0809] 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.

[0810] 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.

[0811] 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.

[0812] 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.

[0813] 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.

[0814] 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."

[0815] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[0816] Functionality Overview

[0817] 1. Voice Input and Speech Recognition

[0818] When a user speaks into a device, the device's microphone captures this voice, and the captured voice data is saved as a digital audio file (e.g., WAV format).

[0819] The device sends this audio file to the server, where necessary metadata (e.g., user ID, timestamp, etc.) is added to the audio data.

[0820] 2. Voice Recognition

[0821] The server then analyzes the received voice data using a cloud-based speech recognition service, such as the Google Speech-to-Text API or a similar service, to convert the voice data into text.

[0822] 3. Natural Language Processing and Response Generation

[0823] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[0824] The server generates an appropriate response based on the analysis results. For example, if the user says hello, the server generates a text response such as "Hello!"

[0825] 4. Language Conversion

[0826] The server translates the generated response text into the language specified by the user. This process uses a translation engine (e.g., Google Translate API or DeepL) to translate the text data.

[0827] The translated text data is saved and the next step, speech synthesis, is carried out.

[0828] 5. Speech Synthesis

[0829] The server generates the translated text as audio data using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API.

[0830] The generated audio data is saved as a file in MP3 format or similar.

[0831] 6. Sending and Playing Audio Data

[0832] The server transmits the generated voice data to the terminal.

[0833] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[0834] Specific examples

[0835] Examples of everyday conversation

[0836] User: "Hello"

[0837] The device captures the audio, stores it and sends it to a server.

[0838] The server converts the speech to text "Hello" and uses natural language processing to recognize it as a greeting. It generates the appropriate response: "Hello!"

[0839] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[0840] The device receives and plays the audio data. The user hears "Hello!"

[0841] Examples of professional conversations

[0842] User: "Tell me about the advances in pacemakers."

[0843] The device captures the audio, stores it and sends it to a server.

[0844] The server converts the speech to text and uses a natural language processing engine with medical expertise to generate a response, such as, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life."

[0845] The server translates the response into the specified language (e.g., English) and generates audio data using a speech synthesis engine.

[0846] The terminal receives and plays the audio data, and the user hears the audio containing the appropriate expertise.

[0847] Thus, the system of the present invention provides a comprehensive means for facilitating multilingual conversations, allowing users to receive responses in a natural and understandable manner.

[0848] The processing flow will be explained below.

[0849] Step 1:

[0850] The user speaks to the terminal. Specifically, the user says "hello."

[0851] Step 2:

[0852] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[0853] Step 3:

[0854] The device sends the captured audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[0855] Step 4:

[0856] The server receives the voice data and passes it to a cloud-based voice recognition service.

[0857] Step 5:

[0858] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format, such as "Hello."

[0859] Step 6:

[0860] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent.

[0861] Step 7:

[0862] The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the response generated is "hello!"

[0863] Step 8:

[0864] The server passes the generated response text to a translation engine, which translates the response text into the specified language using a translation engine (e.g., Google Translate API or DeepL). For example, "Hello!" is translated to "Hello!"

[0865] Step 9:

[0866] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[0867] Step 10:

[0868] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[0869] Step 11:

[0870] The server sends the audio file to the device, encoded in the appropriate data format.

[0871] Step 12:

[0872] The device decodes the received audio file and plays it through the speaker. The user hears "Hello!" being played from the device.

[0873] This series of steps creates a system that allows users to input voice and receive natural-sounding responses in different languages.

[0874] Example 1

[0875] 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."

[0876] In today's global society, smooth communication across multiple languages ​​is important. However, due to the lack of means for users to easily converse in different language environments, there is a demand for a multilingual system based on real-time voice input.

[0877] 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.

[0878] In this invention, the server includes means for receiving voice input from a user, means for saving the voice input as a digital voice file and transmitting it to the server, means for converting the voice data into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into multiple languages, means for generating voice data from the translated text by speech synthesis, and means for transmitting the voice data to the user's terminal and playing it back, thereby enabling users to smoothly engage in real-time voice conversations in different languages.

[0879] "Voice input" refers to capturing voice information spoken by a user using an input device such as a microphone on a device.

[0880] A "digital audio file" is a file that has been converted from analog audio data into a digital format and saved, and is usually saved in a format such as WAV or MP3.

[0881] A "server" is a computer system that processes, stores, and transfers data over a network.

[0882] "Speech Recognition Service" means a cloud-based service for converting voice data into text data using certain algorithms and machine learning models.

[0883] "Text format" is a format in which voice data and other data are expressed as character string information.

[0884] A "natural language processing engine" is an algorithm or machine learning model that understands and processes text data and is used to analyze user intent.

[0885] "Multilingual translation" is the process of converting text written in one language into another language.

[0886] "Speech synthesis" is a technology that converts text data into voice data, and the synthesized voice is reproduced as a voice that resembles a human voice.

[0887] "Playback" means decoding the stored audio data and outputting it as audio again through an output device such as a speaker or headphones.

[0888] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[0889] First, when a user speaks into a terminal, the microphone, which is an input device of the terminal, captures this voice. Specifically, consider the case where the user asks, "What's the weather going to be like tomorrow?" The captured voice data is saved as a digital voice file (e.g., WAV format). The terminal then sends this voice file to the server. Metadata such as the user ID and timestamp are added to the voice data, allowing accurate processing on the server side.

[0890] The server converts the received voice data into text using a cloud-based speech recognition service (e.g., a general speech recognition API). The converted text data is stored on the server in the form of "What's the weather going to be like tomorrow?" The server then passes this text data to a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. Through this analysis process, the server understands that the user wants to know the weather for the next day.

[0891] Based on the understanding, the server generates a response text of "The weather tomorrow is sunny." The response text may be generated by retrieving the latest weather information from a pre-installed weather data API. This response can then be translated into multiple languages, so the server uses a translation engine (e.g., a common translation API) to convert the text to "The weather tomorrow is sunny."

[0892] Once the translation is complete, the server passes the translated text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data. This audio data is saved in MP3 format and temporarily stored on the server. In the final step, the server sends the generated audio data to the device. The device decodes the received audio data and plays it through the speaker.

[0893] This system allows users to smoothly converse in real time across different languages.

[0894] As a concrete example, consider the case where a user says, "Tell me about the advances in pacemakers." The device captures the speech, saves it, and sends it to the server. The server converts the speech to text and uses a natural language processing engine to generate a response using medical expertise. The generated text might be something like, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life." The server translates the response into a specified language (e.g., English) and generates audio data using a speech synthesis engine. The device receives the audio data and plays it back. The user can listen to audio containing expert knowledge.

[0895] In this way, the system of the present invention allows users to ask questions in their own language and receive responses in multiple languages, which is extremely convenient as users can obtain information in a natural conversational format and smoothly carry out conversations in multiple languages.

[0896] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0897] Step 1:

[0898] Capture and transmit audio input

[0899] The user speaks into the device, for example asking a question like "What's the weather like tomorrow?"

[0900] The device's microphone captures the user's voice and stores it as a digital audio file (WAV format).

[0901] The device adds metadata such as the user ID and timestamp to the saved audio file and sends it to the server. This data package is encrypted using the SSL / TLS protocol.

[0902] Input: User's voice

[0903] Output: Digital audio file (with metadata) sent to a server

[0904] Step 2:

[0905] Speech recognition and text conversion

[0906] The server receives the received audio data package and analyzes and identifies the metadata.

[0907] The server sends the voice data to a speech recognition service (a general speech recognition API) and converts it into text. The converted text data is saved in the form of "What's the weather like tomorrow?"

[0908] Input: Digital audio file (with metadata)

[0909] Output: Speech content in text format (e.g., "What's the weather like tomorrow?")

[0910] Step 3:

[0911] Natural language processing of text data

[0912] The server then inputs the converted text data into a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. BERT analyzes the meaning of the sentence, and the generative AI model generates an appropriate response.

[0913] The natural language processing engine understands the intent of "I want to know tomorrow's weather" and provides analysis results as a preliminary step to generating an appropriate response.

[0914] Input: Speech content in text form (e.g., "What's the weather like tomorrow?")

[0915] Output: User intent analysis results

[0916] Step 4:

[0917] Response Generation and Translation

[0918] Based on the analysis results, the server generates a response text "Tomorrow's weather will be sunny." This response may also retrieve the latest weather information from a pre-installed weather data API.

[0919] The server uses a common translation API to translate the response text into the specified language. For example, the response text "Tomorrow's weather is sunny" is translated to "The weather tomorrow is sunny."

[0920] Input: User intent analysis results

[0921] Output: The translated response text (e.g., "The weather tomorrow is sunny")

[0922] Step 5:

[0923] Speech synthesis

[0924] The server passes the translated response text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data, which is saved in MP3 format.

[0925] The server temporarily stores the generated voice data and proceeds to the next transmission step.

[0926] Input: Translated response text (e.g., "The weather tomorrow is sunny")

[0927] Output: Audio data (e.g. "The weather tomorrow is sunny" in MP3 format)

[0928] Step 6:

[0929] Sending and playing audio data

[0930] The server then sends the generated audio data to the device, which is also encrypted using the SSL / TLS protocol.

[0931] The device decodes the received audio data and plays it through the speaker at a volume and quality that is easy for the user to hear.

[0932] Input: Audio data (MP3 format)

[0933] Output: The audio played (e.g. "The weather tomorrow is sunny")

[0934] (Application example 1)

[0935] 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."

[0936] Conventional food delivery systems often pose obstacles when ordering from multilingual users. In particular, language differences can lead to misunderstandings and delays in ordering, impairing the user experience. This has created a need for systems that can accept orders quickly and accurately in regions and situations where multilingual support is difficult.

[0937] 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.

[0938] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's smartphone and playing it back, and means for realizing a dialogue for placing a food delivery order in multiple languages, thereby enabling multilingual users to smoothly place a food delivery order.

[0939] A "user" is an individual or corporation that uses this system to input voice information.

[0940] "Voice input" refers to the voice uttered by the user through a microphone, and is the voice data that the system analyzes.

[0941] "Text format" refers to text information that is the result of converting voice input.

[0942] "User intent" refers to the user's needs and goals expressed in speech input.

[0943] A "response" is a reply or response generated based on the results of analyzing the user's intent.

[0944] "Translation" is the act of converting text expressed in a particular language into another language.

[0945] "Speech synthesis" is a technology that converts text data into voice data.

[0946] "Audio data" refers to a digital audio file generated by voice synthesis.

[0947] A "smartphone" is a type of mobile phone, a device with multiple functions such as voice input, audio playback, and internet connection.

[0948] "Food delivery" is a service that allows users to order food and drinks and have them delivered to a specified location.

[0949] "Dialogue" refers to mutual communication between a user and a system, including voice input and voice output.

[0950] To implement this invention, it is necessary to build a multilingual food delivery system that integrates voice input, voice recognition, natural language processing, language conversion, voice synthesis, and voice playback. Specifically, this system is configured as follows:

[0951] System Configuration

[0952] 1. User Device and Voice Input

[0953] A user places an order by voice using a smartphone, which has a built-in microphone that serves to receive voice input. This voice data is captured and saved as a digital audio file.

[0954] 2. Voice Recognition

[0955] The device sends the captured voice data to a server, which converts it into text using a cloud-based speech recognition service (e.g., Google Speech-to-Text API). This step also adds metadata (e.g., user ID, timestamp, etc.) for processing.

[0956] 3. Natural Language Processing

[0957] The server then passes the converted text data to a natural language processing engine, which uses technologies such as BERT or GPT, to analyze the user's intent. Based on the analysis results, the server generates an appropriate response.

[0958] 4. Language Conversion

[0959] After the response is generated, the server translates it into another language using a translation engine (e.g., Google Translate API or DeepL). The translated text data is then passed to the next step, speech synthesis.

[0960] 5. Speech Synthesis

[0961] The server generates audio data from the translated text using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, and saves the audio data as a file in MP3 format or similar.

[0962] 6. Sending and Playing Audio Data

[0963] The server sends the generated audio data to the user's smartphone, which then decodes the data and plays it through the speaker.

[0964] Specific use cases

[0965] The series of steps in this system will be explained using the following specific example.

[0966] Usage example

[0967] Example prompt: "Hello, can I order? I'd like a pizza and a Coke."

[0968] 1. The user says, "Hello, I'd like to place an order. I'd like a pizza and a Coke."

[0969] 2. The smartphone's microphone captures the audio and sends it to the server as audio data.

[0970] 3. The server performs speech recognition and converts the speech into text: "Hello, can I place my order? I'd like a pizza and a Coke."

[0971] 4. The natural language processing engine analyzes the user's order and generates the appropriate response: "Your order for one pizza and one coke has been accepted."

[0972] 5. The response text is translated into another language (e.g., English), resulting in the text "Your order for one pizza and one coke has been received."

[0973] 6. The translated text is synthesized into speech to create the following audio data: "Your order for one pizza and one coke has been received."

[0974] 7. The audio is sent to your smartphone and played back as audio.

[0975] In this way, multilingual users can smoothly place food delivery orders.

[0976] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0977] Step 1:

[0978] The user inputs voice through the smartphone's microphone. The input voice might be something like, "Hello, can I order? I'd like a pizza and a Coke." The smartphone saves this voice data as a digital audio file in WAV format or similar. The input here is the user's voice, and the output is a digital audio file.

[0979] Step 2:

[0980] The device sends the saved audio data to the server. The data sent includes not only the audio data but also metadata such as the user ID and timestamp. The input here is the digital audio file and metadata, and the output is data sent to the server.

[0981] Step 3:

[0982] The server receives the audio data and converts it into text using a cloud-based speech recognition service, such as the Google Speech-to-Text API. The input is a digital audio file, and the output is the converted text data.

[0983] Step 4:

[0984] The server passes the text data to a natural language processing engine to analyze the user's intent. Generative AI models such as BERT and the GPT series are used to read the user's request from the text data. For example, it recognizes that the user wants to place an order. The input is the text data, and the output is the analyzed user's intent and the generated response.

[0985] Step 5:

[0986] The server generates a response based on the user's intent. For example, if the user says, "I would like to order a pizza and a Coke," the server generates a response such as, "Your order for one pizza and one Coke has been accepted." The input is the parsed user intent, and the output is the generated response text.

[0987] Step 6:

[0988] The server translates the generated response text into the specified language. It uses translation services such as Google Translate API and DeepL to convert the text into multiple languages. For example, the Japanese response "Your order for one pizza and one coke has been received" is translated into English and output as "Your order for one pizza and one coke has been received." The input is the generated response text, and the output is the translated text.

[0989] Step 7:

[0990] The server generates audio data from the translated text using a speech synthesis engine. It uses Amazon Polly or the Google Text-to-Speech API to create an audio file (e.g., MP3 format) from the text. The input is the translated text, and the output is the generated audio data.

[0991] Step 8:

[0992] The server sends the generated audio data to the user's smartphone, which then decodes the received audio data and plays it back through the speaker. The input here is the generated audio data, and the output is the audio playback.

[0993] In this way, by performing specific data processing and calculations at each step, multilingual users can smoothly place food delivery orders.

[0994] 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.

[0995] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[0996] Functionality Overview

[0997] 1. Voice Input and Emotion Recognition

[0998] When a user speaks into a device, the device's microphone captures this sound, which is then saved as a digital audio file (e.g., WAV format).

[0999] The device sends the saved audio file to the server, where necessary metadata (e.g., user ID, timestamp) is added to the audio data.

[1000] 2. Speech Recognition and Emotion Analysis

[1001] The server then analyzes the received voice data using a cloud-based speech recognition service, such as the Google Speech-to-Text API or a similar service, to convert the voice data into text.

[1002] At the same time, the server uses an emotion engine to analyze the pitch, rate, and volume of the voice data to recognize the user's emotions, which identify emotional states such as joy, sadness, anger, surprise, etc.

[1003] 3. Natural Language Processing and Response Generation

[1004] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[1005] The server generates an appropriate response based on the analysis, for example, if the user says "hello", the response generated is "hello!"

[1006] 4. Emotional regulation of responses

[1007] The server adjusts the response based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response will be adjusted to be kinder.

[1008] The adjusted response text is passed to a translation engine for translation into the specified language.

[1009] 5. Language Conversion and Speech Synthesis

[1010] The server translates the generated response text using a translation engine, such as Google Translate API or DeepL.

[1011] The server passes the translated text to a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, to generate audio data.

[1012] The generated audio data is saved as a file in MP3 format or similar.

[1013] 6. Sending and Playing Audio Data

[1014] The server transmits the generated voice data to the terminal.

[1015] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[1016] Specific examples

[1017] Examples of everyday conversation

[1018] User: "Hello"

[1019] The device captures the audio, stores it and sends it to a server.

[1020] The server converts the speech to text "Hello," then uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize that it is a greeting and then generates the appropriate response: "Hello!"

[1021] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[1022] The device plays "Hello!", which the user hears.

[1023] Examples of professional conversations

[1024] User: "Tell me about the advances in pacemakers."

[1025] The device captures the audio, stores it and sends it to a server.

[1026] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to generate a response using medical expertise, such as "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[1027] The server translates the generated response into the specified language (e.g., English) and generates audio data using a speech synthesis engine, adjusting the response to a gentle tone to reduce tension.

[1028] The device plays audio data such as, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased," and the user listens to it.

[1029] In this way, the system of the present invention can recognize the user's emotions and adjust responses with an appropriate tone, providing a more natural and understandable multilingual conversation.

[1030] The processing flow will be explained below.

[1031] Step 1:

[1032] The user speaks to the terminal. Specifically, the user utters, "What's the weather like this week?"

[1033] Step 2:

[1034] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[1035] Step 3:

[1036] The device sends the saved audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[1037] Step 4:

[1038] The server receives the voice data and passes it to a cloud-based voice recognition service.

[1039] Step 5:

[1040] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format. The converted text might be, for example, "What's the weather like this week?"

[1041] Step 6:

[1042] At the same time, the server passes the voice data to the emotion engine, which analyzes the pitch, speed, volume, etc. to recognize the user's emotion. For example, it may recognize that the user is excited.

[1043] Step 7:

[1044] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent. In this case, it determines that the user is looking for weather information.

[1045] Step 8:

[1046] The server generates an appropriate response based on the analysis results, for example, a text response such as "The weather this week is going to be rainy."

[1047] Step 9:

[1048] The server adjusts the generated response based on the emotion recognition results from the emotion engine, adjusting the tone of the response to calm an agitated user.

[1049] Step 10:

[1050] The server passes the adjusted response text to a translation engine (e.g., Google Translate API or DeepL), which translates the response text into the specified language. In this case, it is translated as "The weather for this week is mostly rainy."

[1051] Step 11:

[1052] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[1053] Step 12:

[1054] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[1055] Step 13:

[1056] The server sends the audio file to the device, encoded in the appropriate data format.

[1057] Step 14:

[1058] The device decodes the received audio file and plays it through the speaker. The user hears the audio being played from the device: "The weather for this week is mostly rainy."

[1059] This series of steps builds a system that allows users to receive natural responses in different languages ​​that take emotions into account in response to their input voice.

[1060] Example 2

[1061] 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."

[1062] Conventional voice conversation systems generate responses without considering the user's emotions, resulting in a poor user experience. They also struggle to achieve natural multilingual conversations, which can lead to a lack of accurate understanding of the user's intent. Furthermore, processing voice data is complex, making it difficult to respond in real time. Technology to resolve these issues is needed.

[1063] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1064] In this invention, the server includes means for receiving voice input and saving it as a digital voice file, means for transmitting the voice data to the server, means for converting the voice input into text format using a voice recognition service, means for analyzing the voice data using an emotion engine to recognize the user's emotional state, means including a natural language processing engine for analyzing the user's intention and generating an appropriate response, means for adjusting the tone of the generated response based on the emotion recognition result, means for translating the adjusted response into another language, means for generating the translated text as voice data using a voice synthesis engine, and means for transmitting the generated voice data to a terminal for playback. This enables natural, easy-to-understand multilingual conversations that take the user's emotions into consideration.

[1065] "Voice input" refers to the voice that a user utters into a terminal, and is the subject of processing by voice recognition technology.

[1066] A "digital audio file" refers to data that has been converted from analog audio into digital format and saved in formats such as WAV or MP3.

[1067] "Server" means a central processing unit that receives, processes, and generates responses from voice data, including cloud-based services.

[1068] A "speech recognition service" is a technology or service that converts voice data into text, typically using a cloud-based API.

[1069] "Text format" refers to a format in which voice data is expressed as text information, and is used in natural language processing and translation.

[1070] An "emotion engine" refers to a technology or system that analyzes and recognizes a user's emotional state from the pitch, speed, volume, etc. of voice data.

[1071] A "natural language processing engine" is a technology or system that analyzes user intent from text data and generates appropriate responses.

[1072] "Response tone" refers to the emotional and expressive modulation of the generated response voice, which changes depending on the user's emotional state.

[1073] "Translation" refers to the process of converting text data expressed in one language into another, supporting multilingual conversations.

[1074] A "speech synthesis engine" is a technology or system that converts text data into voice data, and is used to generate natural-sounding voices.

[1075] A "terminal" is a device used by a user, which has the functions of capturing audio, transmitting and receiving audio data, and playing audio.

[1076] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes user emotions. The system program uses specific hardware and software to efficiently process user voice input and provide natural, easy-to-understand multilingual responses.

[1077] First, when a user speaks into the device, the device's microphone captures the voice. This voice data is saved as a digital audio file (e.g., WAV format). The device then sends the saved audio file to the server. During this process, metadata (e.g., user ID, timestamp) is added to the voice data.

[1078] The server analyzes the received voice data using a cloud-based voice recognition service. Specific services include commonly used voice recognition APIs (e.g., Speech-to-Text APIs). This service converts the voice data into text format. At the same time, the server uses an emotion engine to analyze the pitch, speed, and volume of the voice data to recognize the user's emotional state. This emotion engine identifies emotional states such as joy, sadness, anger, and surprise.

[1079] The server then passes the converted text data to a natural language processing engine to analyze the user's intent. The natural language processing engine used here may utilize technologies from the BERT or GPT series. The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the server generates the response "hello!"

[1080] The generated response is adjusted based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response is adjusted to be gentler. This adjusted response text is passed to a translation engine and translated into the specified language. Services such as Google Translate API and DeepL are used as translation engines.

[1081] The server then passes the translated text to a speech synthesis engine, typically Amazon Polly or the Text-to-Speech API, to generate audio data, which is then saved as an audio file in MP3 format or similar.

[1082] Finally, the server sends the generated voice data to the device, which then decodes the data and plays it back through the speaker, allowing the user to hear a natural-sounding response.

[1083] Specific examples

[1084] Examples of everyday conversation

[1085] User: "Hello"

[1086] The device captures the audio, stores it and sends it to a server.

[1087] The server converts the speech to text "Hello," and uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize it as a greeting and generates the appropriate response: "Hello!"

[1088] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[1089] The device plays "Hello!", which the user hears.

[1090] Examples of professional conversations

[1091] User: "Tell me about the advances in pacemakers."

[1092] The device captures the audio, stores it and sends it to a server.

[1093] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to reference medical expertise to generate a response such as, "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[1094] The server translates the generated text into the specified language (e.g., English) and generates voice data using a speech synthesis engine, adjusting the tone to be gentler to reduce tension.

[1095] The device plays audio data to the user that says, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased."

[1096] As described above, the present invention can provide multilingual conversations in a more natural and understandable manner by recognizing the user's emotions and adjusting responses with an appropriate tone.

[1097] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1098] Step 1:

[1099] The user starts speaking into the device. The device's microphone captures this voice and saves it as a digital audio file (e.g., WAV format). The input is the user's voice data, and the output is the saved digital audio file.

[1100] Step 2:

[1101] The device sends the stored audio file and metadata (e.g., user ID, timestamp) to the server. The input is the digital audio file and metadata, and the output is the data sent to the server. Communication is secure using the HTTP protocol.

[1102] Step 3:

[1103] The server sends the received audio file to a cloud-based speech recognition service (e.g., speech recognition API) to convert the audio into text. The input is the audio file, and the output is the converted text data. The server calls the speech recognition API, and the audio data is converted into text data.

[1104] Step 4:

[1105] The server passes the converted text data to an emotion engine, which analyzes the pitch, speed, and volume of the audio data to recognize the user's emotional state. The input is the text data and audio data properties, and the output is the recognized emotional state. The server uses the emotion engine to perform polarity analysis and acoustic feature analysis.

[1106] Step 5:

[1107] The server passes the emotion recognition results and converted text data to a natural language processing engine (e.g., a generative AI model) to analyze the user's intent. The input is text data and emotion recognition results, and the output is a response text based on the user's intent. The natural language processing engine performs calculations based on the prompt sentence and generates an appropriate response.

[1108] Step 6:

[1109] The server adjusts the generated response text based on the emotion recognition results. The input is the generated response text and the emotion recognition results, and the output is the adjusted response text. The server adjusts the tone and style according to the emotion.

[1110] Step 7:

[1111] The server passes the adjusted response text to a translation engine to translate it into the specified language. The input is the adjusted response text and the output is the translated text. The translation engine (e.g., translation API) enables multilingual support.

[1112] Step 8:

[1113] The server passes the translated text to a speech synthesis engine, which generates the audio data. The input is the translated text, and the output is the generated audio data (e.g., MP3 format). The audio data is generated using a speech synthesis engine (e.g., a speech synthesis API).

[1114] Step 9:

[1115] The server sends the generated audio data to the device. The input is the generated audio data, and the output is the audio data sent to the device. The audio data is securely sent to the device using the HTTP protocol.

[1116] Step 10:

[1117] The device decodes the received audio data and plays it through the speaker. The input is the audio data sent to the device, and the output is the audio the user hears. The device decodes the audio data and plays it at the appropriate volume and quality.

[1118] (Application example 2)

[1119] 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."

[1120] Multilingual voice conversation systems often lack the ability to recognize users' emotions and adjust responses appropriately, which can lead to a decline in the quality of communication. Furthermore, security services require the ability to recognize visitors' emotions and provide accurate warnings to users, but current systems lack such advanced emotion recognition capabilities, which creates a problem of insufficient safety.

[1121] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1122] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's terminal and playing it back, means for recognizing the user's emotion and appropriately adjusting the response, and means for transmitting the emotion recognition result and the voice data for security notification. This enables a natural response that takes the user's emotion into consideration, making it possible to realize an advanced emotion-recognition security assistant that contributes to improved security.

[1123] The "means for receiving voice input" is a function for capturing the user's spoken voice with a device such as a microphone and storing it as digital data.

[1124] The "means for converting into text format" is a technology for converting voice input into text data, and is realized using a voice recognition engine.

[1125] "Means for analyzing the user's intent and generating an appropriate response" refers to a function that analyzes the content of the user's speech using natural language processing, understands the intent, and then creates an appropriate response.

[1126] A "means for translating into another language" refers to a translation engine used to translate the generated response text into another specified language.

[1127] "Means for generating voice data by voice synthesis" refers to a technology that uses a voice synthesis engine to convert text data into voice data.

[1128] The "means for transmitting audio data to a user's terminal and playing it back" is a function for transferring the generated audio data to a user's device and playing it back through a speaker or headphones.

[1129] "Means for recognizing the user's emotions and adjusting the response appropriately" refers to a system that analyzes the user's emotional state from voice data and appropriately changes the tone and content of the response based on the results.

[1130] "Means for sending emotion recognition results and voice data for security notifications" is a function for notifying the user of the results of analysis by the emotion recognition engine and related voice data in real time to their smartphone, etc.

[1131] The present invention is a security assistant system that processes user voice input in multiple languages, recognizes emotions, and generates appropriate responses. This system is realized by a combination of hardware and software.

[1132] The server receives the user's voice input using a microphone installed on the user's device. The voice data captured by the microphone is saved in digital format and then converted into text format. A cloud-based speech recognition service is used as the speech recognition engine for the voice conversion. Specifically, the Google Speech-to-Text API is used.

[1133] After the voice input is converted to text, the server uses a natural language processing engine (e.g., OpenAI's GPT-3) to analyze the user's intent and generate an appropriate response. During this analysis step, a generative AI model operates based on a prompt sentence. For example, the following prompt sentence is used:

[1134] The visitor seems angry. Generate an appropriate response.

[1135] Once the analysis is complete, the generated response is translated into another language using the Google Translate API or other translation engines. The translated text data is then converted into audio data using a speech synthesis engine (e.g., Google Text-to-Speech API). This audio data is sent to the user's device and played through the speaker.

[1136] Regarding emotion recognition, an emotion engine (e.g., IBM Watson Tone Analyzer) is used to recognize emotions from the user's voice, and the tone and content of the generated response are adjusted according to the user's emotional state.

[1137] Furthermore, emotion recognition results and voice data are sent to the user's smartphone for security notifications. For example, if a visitor is recognized as angry, a warning message will be displayed on the user's smartphone saying, "The visitor is angry. Please be careful." This function allows the user to understand the visitor's emotional state and respond safely.

[1138] For example, if a visitor says "Open quickly!" through the doorbell, the system captures the speech and converts it into text using the Google Speech-to-Text API. The speech is then analyzed by OpenAI's GPT-3 and the emotion is recognized as "anger" using IBM Watson Tone Analyzer. The response generated is "Please wait a moment, we'll respond shortly," which is translated using the Google Translate API and then converted into speech using the Google Text-to-Speech API. Finally, the speech data and emotion recognition results are sent to the user's smartphone, prompting them to take an appropriate action.

[1139] In this way, the present invention not only recognizes the user's intention and emotion from their voice input and responds appropriately in multiple languages, but also has a function to notify the emotion recognition results to enhance security.

[1140] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1141] Step 1:

[1142] The user speaks into the terminal.

[1143] Input: User's voice

[1144] Output: Audio data (WAV format)

[1145] Specific operation: The microphone on the device captures the user's voice and saves it as a WAV audio file, which is then used for further processing.

[1146] Step 2:

[1147] The terminal transmits the voice data to the server.

[1148] Input: Audio data (WAV format)

[1149] Output: None (forwarded to server)

[1150] Specific operation: The device sends the captured audio data to the server, along with the necessary metadata (user ID, timestamp, etc.).

[1151] Step 3:

[1152] The server converts the audio data into text format.

[1153] Input: Audio data (WAV format)

[1154] Output: Text data

[1155] What happens: The server uses the Google Speech-to-Text API to convert the audio data into text, which is then used in the next analysis step.

[1156] Step 4:

[1157] The server performs emotion recognition.

[1158] Input: Audio data (WAV format)

[1159] Output: Sentiment analysis results (e.g. anger, joy, sadness, etc.)

[1160] How it works: The server uses IBM Watson Tone Analyzer to analyze the pitch, rate, and volume of the audio data to identify the user's emotions.

[1161] Step 5:

[1162] The server analyzes the user's intent from the text data and generates an appropriate response.

[1163] Input: Text data

[1164] Output: Response text

[1165] How it works: The server uses OpenAI's GPT-3 model to analyze the text data and understand the user's intent. Based on the analysis results, it generates an appropriate response in text format.

[1166] Step 6:

[1167] The server translates the generated response text into another language.

[1168] Input: Response text

[1169] Output: Translated text

[1170] Specific behavior: The server uses a translation engine such as the Google Translate API to translate the generated response text into another specified language.

[1171] Step 7:

[1172] The server generates the translated text as audio data.

[1173] Input: Translated text

[1174] Output: Audio data (MP3 format)

[1175] How it works: The server uses the Google Text-to-Speech API to convert the translated text into audio data, which is then saved in MP3 format.

[1176] Step 8:

[1177] The server sends the audio data to the terminal and plays it back.

[1178] Input: Audio data (MP3 format)

[1179] Output: None (transfer to terminal)

[1180] Specific operation: The server sends the generated audio data to the terminal, and the terminal decodes the received audio data and plays it through the speaker.

[1181] Step 9:

[1182] The server sends the emotion recognition results and voice data to the user's device for security notification.

[1183] Input: Emotion recognition results, audio data

[1184] Output: Security Notice

[1185] Specific operation: The server sends the emotion recognition results and related voice data to the user's smartphone, and displays warnings and notifications based on the user's emotional state.

[1186] 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.

[1187] 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.

[1188] 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.

[1189] [Fourth embodiment]

[1190] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1191] 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.

[1192] 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).

[1193] 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.

[1194] 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.

[1195] 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).

[1196] 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.

[1197] 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.

[1198] 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.

[1199] 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.

[1200] 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.

[1201] 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.

[1202] 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."

[1203] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[1204] Functionality Overview

[1205] 1. Voice Input and Speech Recognition

[1206] When a user speaks into a device, the device's microphone captures this voice, and the captured voice data is saved as a digital audio file (e.g., WAV format).

[1207] The device sends this audio file to the server, where necessary metadata (e.g., user ID, timestamp, etc.) is added to the audio data.

[1208] 2. Voice Recognition

[1209] The server then analyzes the received voice data using a cloud-based speech recognition service, such as the Google Speech-to-Text API or a similar service, to convert the voice data into text.

[1210] 3. Natural Language Processing and Response Generation

[1211] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[1212] The server generates an appropriate response based on the analysis results. For example, if the user says hello, the server generates a text response such as "Hello!"

[1213] 4. Language Conversion

[1214] The server translates the generated response text into the language specified by the user. This process uses a translation engine (e.g., Google Translate API or DeepL) to translate the text data.

[1215] The translated text data is saved and the next step, speech synthesis, is carried out.

[1216] 5. Speech Synthesis

[1217] The server generates the translated text as audio data using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API.

[1218] The generated audio data is saved as a file in MP3 format or similar.

[1219] 6. Sending and Playing Audio Data

[1220] The server transmits the generated voice data to the terminal.

[1221] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[1222] Specific examples

[1223] Examples of everyday conversation

[1224] User: "Hello"

[1225] The device captures the audio, stores it and sends it to a server.

[1226] The server converts the speech to text "Hello" and uses natural language processing to recognize it as a greeting. It generates the appropriate response: "Hello!"

[1227] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[1228] The device receives and plays the audio data. The user hears "Hello!"

[1229] Examples of professional conversations

[1230] User: "Tell me about the advances in pacemakers."

[1231] The device captures the audio, stores it and sends it to a server.

[1232] The server converts the speech to text and uses a natural language processing engine with medical expertise to generate a response, such as, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life."

[1233] The server translates the response into the specified language (e.g., English) and generates audio data using a speech synthesis engine.

[1234] The terminal receives and plays the audio data, and the user hears the audio containing the appropriate expertise.

[1235] Thus, the system of the present invention provides a comprehensive means for facilitating multilingual conversations, allowing users to receive responses in a natural and understandable manner.

[1236] The processing flow will be explained below.

[1237] Step 1:

[1238] The user speaks to the terminal. Specifically, the user says "hello."

[1239] Step 2:

[1240] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[1241] Step 3:

[1242] The device sends the captured audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[1243] Step 4:

[1244] The server receives the voice data and passes it to a cloud-based voice recognition service.

[1245] Step 5:

[1246] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format, such as "Hello."

[1247] Step 6:

[1248] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent.

[1249] Step 7:

[1250] The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the response generated is "hello!"

[1251] Step 8:

[1252] The server passes the generated response text to a translation engine, which translates the response text into the specified language using a translation engine (e.g., Google Translate API or DeepL). For example, "Hello!" is translated to "Hello!"

[1253] Step 9:

[1254] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[1255] Step 10:

[1256] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[1257] Step 11:

[1258] The server sends the audio file to the device, encoded in the appropriate data format.

[1259] Step 12:

[1260] The device decodes the received audio file and plays it through the speaker. The user hears "Hello!" being played from the device.

[1261] This series of steps creates a system that allows users to input voice and receive natural-sounding responses in different languages.

[1262] Example 1

[1263] 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."

[1264] In today's global society, smooth communication across multiple languages ​​is important. However, due to the lack of means for users to easily converse in different language environments, there is a demand for a multilingual system based on real-time voice input.

[1265] 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.

[1266] In this invention, the server includes means for receiving voice input from a user, means for saving the voice input as a digital voice file and transmitting it to the server, means for converting the voice data into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into multiple languages, means for generating voice data from the translated text by speech synthesis, and means for transmitting the voice data to the user's terminal and playing it back, thereby enabling users to smoothly engage in real-time voice conversations in different languages.

[1267] "Voice input" refers to capturing voice information spoken by a user using an input device such as a microphone on a device.

[1268] A "digital audio file" is a file that has been converted from analog audio data into a digital format and saved, and is usually saved in a format such as WAV or MP3.

[1269] A "server" is a computer system that processes, stores, and transfers data over a network.

[1270] "Speech Recognition Service" means a cloud-based service for converting voice data into text data using certain algorithms and machine learning models.

[1271] "Text format" is a format in which voice data and other data are expressed as character string information.

[1272] A "natural language processing engine" is an algorithm or machine learning model that understands and processes text data and is used to analyze user intent.

[1273] "Multilingual translation" is the process of converting text written in one language into another language.

[1274] "Speech synthesis" is a technology that converts text data into voice data, and the synthesized voice is reproduced as a voice that resembles a human voice.

[1275] "Playback" means decoding the stored audio data and outputting it as audio again through an output device such as a speaker or headphones.

[1276] The system of the present invention aims to understand speech input from a user into a terminal in multiple languages ​​and generate an appropriate response. Specific embodiments for carrying out the present invention will be described below.

[1277] First, when a user speaks into a terminal, the microphone, which is an input device of the terminal, captures this voice. Specifically, consider the case where the user asks, "What's the weather going to be like tomorrow?" The captured voice data is saved as a digital voice file (e.g., WAV format). The terminal then sends this voice file to the server. Metadata such as the user ID and timestamp are added to the voice data, allowing accurate processing on the server side.

[1278] The server converts the received voice data into text using a cloud-based speech recognition service (e.g., a general speech recognition API). The converted text data is stored on the server in the form of "What's the weather going to be like tomorrow?" The server then passes this text data to a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. Through this analysis process, the server understands that the user wants to know the weather for the next day.

[1279] Based on the understanding, the server generates a response text of "The weather tomorrow is sunny." The response text may be generated by retrieving the latest weather information from a pre-installed weather data API. This response can then be translated into multiple languages, so the server uses a translation engine (e.g., a common translation API) to convert the text to "The weather tomorrow is sunny."

[1280] Once the translation is complete, the server passes the translated text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data. This audio data is saved in MP3 format and temporarily stored on the server. In the final step, the server sends the generated audio data to the device. The device decodes the received audio data and plays it through the speaker.

[1281] This system allows users to smoothly converse in real time across different languages.

[1282] As a concrete example, consider the case where a user says, "Tell me about the advances in pacemakers." The device captures the speech, saves it, and sends it to the server. The server converts the speech to text and uses a natural language processing engine to generate a response using medical expertise. The generated text might be something like, "In recent years, pacemakers have become smaller and their lithium batteries have significantly longer battery life." The server translates the response into a specified language (e.g., English) and generates audio data using a speech synthesis engine. The device receives the audio data and plays it back. The user can listen to audio containing expert knowledge.

[1283] In this way, the system of the present invention allows users to ask questions in their own language and receive responses in multiple languages, which is extremely convenient as users can obtain information in a natural conversational format and smoothly carry out conversations in multiple languages.

[1284] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1285] Step 1:

[1286] Capture and transmit audio input

[1287] The user speaks into the device, for example asking a question like "What's the weather like tomorrow?"

[1288] The device's microphone captures the user's voice and stores it as a digital audio file (WAV format).

[1289] The device adds metadata such as the user ID and timestamp to the saved audio file and sends it to the server. This data package is encrypted using the SSL / TLS protocol.

[1290] Input: User's voice

[1291] Output: Digital audio file (with metadata) sent to a server

[1292] Step 2:

[1293] Speech recognition and text conversion

[1294] The server receives the received audio data package and analyzes and identifies the metadata.

[1295] The server sends the voice data to a speech recognition service (a general speech recognition API) and converts it into text. The converted text data is saved in the form of "What's the weather like tomorrow?"

[1296] Input: Digital audio file (with metadata)

[1297] Output: Speech content in text format (e.g., "What's the weather like tomorrow?")

[1298] Step 3:

[1299] Natural language processing of text data

[1300] The server then inputs the converted text data into a natural language processing engine (e.g., BERT or a generative AI model) to analyze the user's intent. BERT analyzes the meaning of the sentence, and the generative AI model generates an appropriate response.

[1301] The natural language processing engine understands the intent of "I want to know tomorrow's weather" and provides analysis results as a preliminary step to generating an appropriate response.

[1302] Input: Speech content in text form (e.g., "What's the weather like tomorrow?")

[1303] Output: User intent analysis results

[1304] Step 4:

[1305] Response Generation and Translation

[1306] Based on the analysis results, the server generates a response text "Tomorrow's weather will be sunny." This response may also retrieve the latest weather information from a pre-installed weather data API.

[1307] The server uses a common translation API to translate the response text into the specified language. For example, the response text "Tomorrow's weather is sunny" is translated to "The weather tomorrow is sunny."

[1308] Input: User intent analysis results

[1309] Output: The translated response text (e.g., "The weather tomorrow is sunny")

[1310] Step 5:

[1311] Speech synthesis

[1312] The server passes the translated response text to a speech synthesis engine (e.g., a publicly known speech synthesis service) to generate audio data, which is saved in MP3 format.

[1313] The server temporarily stores the generated voice data and proceeds to the next transmission step.

[1314] Input: Translated response text (e.g., "The weather tomorrow is sunny")

[1315] Output: Audio data (e.g. "The weather tomorrow is sunny" in MP3 format)

[1316] Step 6:

[1317] Sending and playing audio data

[1318] The server then sends the generated audio data to the device, which is also encrypted using the SSL / TLS protocol.

[1319] The device decodes the received audio data and plays it through the speaker at a volume and quality that is easy for the user to hear.

[1320] Input: Audio data (MP3 format)

[1321] Output: The audio played (e.g. "The weather tomorrow is sunny")

[1322] (Application example 1)

[1323] 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."

[1324] Conventional food delivery systems often pose obstacles when ordering from multilingual users. In particular, language differences can lead to misunderstandings and delays in ordering, impairing the user experience. This has created a need for systems that can accept orders quickly and accurately in regions and situations where multilingual support is difficult.

[1325] 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.

[1326] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's smartphone and playing it back, and means for realizing a dialogue for placing a food delivery order in multiple languages, thereby enabling multilingual users to smoothly place a food delivery order.

[1327] A "user" is an individual or corporation that uses this system to input voice information.

[1328] "Voice input" refers to the voice uttered by the user through a microphone, and is the voice data that the system analyzes.

[1329] "Text format" refers to text information that is the result of converting voice input.

[1330] "User intent" refers to the user's needs and goals expressed in speech input.

[1331] A "response" is a reply or response generated based on the results of analyzing the user's intent.

[1332] "Translation" is the act of converting text expressed in a particular language into another language.

[1333] "Speech synthesis" is a technology that converts text data into voice data.

[1334] "Audio data" refers to a digital audio file generated by voice synthesis.

[1335] A "smartphone" is a type of mobile phone, a device with multiple functions such as voice input, audio playback, and internet connection.

[1336] "Food delivery" is a service that allows users to order food and drinks and have them delivered to a specified location.

[1337] "Dialogue" refers to mutual communication between a user and a system, including voice input and voice output.

[1338] To implement this invention, it is necessary to build a multilingual food delivery system that integrates voice input, voice recognition, natural language processing, language conversion, voice synthesis, and voice playback. Specifically, this system is configured as follows:

[1339] System Configuration

[1340] 1. User Device and Voice Input

[1341] A user places an order by voice using a smartphone, which has a built-in microphone that serves to receive voice input. This voice data is captured and saved as a digital audio file.

[1342] 2. Voice Recognition

[1343] The device sends the captured voice data to a server, which converts it into text using a cloud-based speech recognition service (e.g., Google Speech-to-Text API). This step also adds metadata (e.g., user ID, timestamp, etc.) for processing.

[1344] 3. Natural Language Processing

[1345] The server then passes the converted text data to a natural language processing engine, which uses technologies such as BERT or GPT, to analyze the user's intent. Based on the analysis results, the server generates an appropriate response.

[1346] 4. Language Conversion

[1347] After the response is generated, the server translates it into another language using a translation engine (e.g., Google Translate API or DeepL). The translated text data is then passed to the next step, speech synthesis.

[1348] 5. Speech Synthesis

[1349] The server generates audio data from the translated text using a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, and saves the audio data as a file in MP3 format or similar.

[1350] 6. Sending and Playing Audio Data

[1351] The server sends the generated audio data to the user's smartphone, which then decodes the data and plays it through the speaker.

[1352] Specific use cases

[1353] The series of steps in this system will be explained using the following specific example.

[1354] Usage example

[1355] Example prompt: "Hello, can I order? I'd like a pizza and a Coke."

[1356] 1. The user says, "Hello, I'd like to place an order. I'd like a pizza and a Coke."

[1357] 2. The smartphone's microphone captures the audio and sends it to the server as audio data.

[1358] 3. The server performs speech recognition and converts the speech into text: "Hello, can I place my order? I'd like a pizza and a Coke."

[1359] 4. The natural language processing engine analyzes the user's order and generates the appropriate response: "Your order for one pizza and one coke has been accepted."

[1360] 5. The response text is translated into another language (e.g., English), resulting in the text "Your order for one pizza and one coke has been received."

[1361] 6. The translated text is synthesized into speech to create the following audio data: "Your order for one pizza and one coke has been received."

[1362] 7. The audio is sent to your smartphone and played back as audio.

[1363] In this way, multilingual users can smoothly place food delivery orders.

[1364] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1365] Step 1:

[1366] The user inputs voice through the smartphone's microphone. The input voice might be something like, "Hello, can I order? I'd like a pizza and a Coke." The smartphone saves this voice data as a digital audio file in WAV format or similar. The input here is the user's voice, and the output is a digital audio file.

[1367] Step 2:

[1368] The device sends the saved audio data to the server. The data sent includes not only the audio data but also metadata such as the user ID and timestamp. The input here is the digital audio file and metadata, and the output is data sent to the server.

[1369] Step 3:

[1370] The server receives the audio data and converts it into text using a cloud-based speech recognition service, such as the Google Speech-to-Text API. The input is a digital audio file, and the output is the converted text data.

[1371] Step 4:

[1372] The server passes the text data to a natural language processing engine to analyze the user's intent. Generative AI models such as BERT and the GPT series are used to read the user's request from the text data. For example, it recognizes that the user wants to place an order. The input is the text data, and the output is the analyzed user's intent and the generated response.

[1373] Step 5:

[1374] The server generates a response based on the user's intent. For example, if the user says, "I would like to order a pizza and a Coke," the server generates a response such as, "Your order for one pizza and one Coke has been accepted." The input is the parsed user intent, and the output is the generated response text.

[1375] Step 6:

[1376] The server translates the generated response text into the specified language. It uses translation services such as Google Translate API and DeepL to convert the text into multiple languages. For example, the Japanese response "Your order for one pizza and one coke has been received" is translated into English and output as "Your order for one pizza and one coke has been received." The input is the generated response text, and the output is the translated text.

[1377] Step 7:

[1378] The server generates audio data from the translated text using a speech synthesis engine. It uses Amazon Polly or the Google Text-to-Speech API to create an audio file (e.g., MP3 format) from the text. The input is the translated text, and the output is the generated audio data.

[1379] Step 8:

[1380] The server sends the generated audio data to the user's smartphone, which then decodes the received audio data and plays it back through the speaker. The input here is the generated audio data, and the output is the audio playback.

[1381] In this way, by performing specific data processing and calculations at each step, multilingual users can smoothly place food delivery orders.

[1382] 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.

[1383] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[1384] Functionality Overview

[1385] 1. Voice Input and Emotion Recognition

[1386] When a user speaks into a device, the device's microphone captures this sound, which is then saved as a digital audio file (e.g., WAV format).

[1387] The device sends the saved audio file to the server, where necessary metadata (e.g., user ID, timestamp) is added to the audio data.

[1388] 2. Speech Recognition and Emotion Analysis

[1389] The server then analyzes the received voice data using a cloud-based speech recognition service, such as the Google Speech-to-Text API or a similar service, to convert the voice data into text.

[1390] At the same time, the server uses an emotion engine to analyze the pitch, rate, and volume of the voice data to recognize the user's emotions, which identify emotional states such as joy, sadness, anger, surprise, etc.

[1391] 3. Natural Language Processing and Response Generation

[1392] The server then passes the converted text data to a natural language processing engine, which may use technologies such as BERT or the GPT series, to analyze the user's intent.

[1393] The server generates an appropriate response based on the analysis, for example, if the user says "hello", the response generated is "hello!"

[1394] 4. Emotional regulation of responses

[1395] The server adjusts the response based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response will be adjusted to be kinder.

[1396] The adjusted response text is passed to a translation engine for translation into the specified language.

[1397] 5. Language Conversion and Speech Synthesis

[1398] The server translates the generated response text using a translation engine, such as Google Translate API or DeepL.

[1399] The server passes the translated text to a speech synthesis engine, such as Amazon Polly or the Google Text-to-Speech API, to generate audio data.

[1400] The generated audio data is saved as a file in MP3 format or similar.

[1401] 6. Sending and Playing Audio Data

[1402] The server transmits the generated voice data to the terminal.

[1403] The device decodes the received audio data and plays it through the speaker at an appropriate volume and quality so that the user can easily hear it.

[1404] Specific examples

[1405] Examples of everyday conversation

[1406] User: "Hello"

[1407] The device captures the audio, stores it and sends it to a server.

[1408] The server converts the speech to text "Hello," then uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize that it is a greeting and then generates the appropriate response: "Hello!"

[1409] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[1410] The device plays "Hello!", which the user hears.

[1411] Examples of professional conversations

[1412] User: "Tell me about the advances in pacemakers."

[1413] The device captures the audio, stores it and sends it to a server.

[1414] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to generate a response using medical expertise, such as "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[1415] The server translates the generated response into the specified language (e.g., English) and generates audio data using a speech synthesis engine, adjusting the response to a gentle tone to reduce tension.

[1416] The device plays audio data such as, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased," and the user listens to it.

[1417] In this way, the system of the present invention can recognize the user's emotions and adjust responses with an appropriate tone, providing a more natural and understandable multilingual conversation.

[1418] The processing flow will be explained below.

[1419] Step 1:

[1420] The user speaks to the terminal. Specifically, the user utters, "What's the weather like this week?"

[1421] Step 2:

[1422] The device captures audio through the microphone, and the captured audio data is saved as a digital audio file (e.g., WAV format).

[1423] Step 3:

[1424] The device sends the saved audio file to the server, where metadata (e.g., user ID, timestamp) is added to the audio data.

[1425] Step 4:

[1426] The server receives the voice data and passes it to a cloud-based voice recognition service.

[1427] Step 5:

[1428] The server uses a cloud-based speech recognition service (e.g., Google Speech-to-Text API) to convert the voice data into text format. The converted text might be, for example, "What's the weather like this week?"

[1429] Step 6:

[1430] At the same time, the server passes the voice data to the emotion engine, which analyzes the pitch, speed, volume, etc. to recognize the user's emotion. For example, it may recognize that the user is excited.

[1431] Step 7:

[1432] The server passes the text data to a natural language processing engine (e.g., BERT or GPT series), which analyzes the user's intent. In this case, it determines that the user is looking for weather information.

[1433] Step 8:

[1434] The server generates an appropriate response based on the analysis results, for example, a text response such as "The weather this week is going to be rainy."

[1435] Step 9:

[1436] The server adjusts the generated response based on the emotion recognition results from the emotion engine, adjusting the tone of the response to calm an agitated user.

[1437] Step 10:

[1438] The server passes the adjusted response text to a translation engine (e.g., Google Translate API or DeepL), which translates the response text into the specified language. In this case, it is translated as "The weather for this week is mostly rainy."

[1439] Step 11:

[1440] The server passes the translated text data to a speech synthesis engine (e.g., Amazon Polly or Google Text-to-Speech API), which generates audio from the translated text.

[1441] Step 12:

[1442] The server stores the generated audio data as a digital audio file (e.g., MP3 format).

[1443] Step 13:

[1444] The server sends the audio file to the device, encoded in the appropriate data format.

[1445] Step 14:

[1446] The device decodes the received audio file and plays it through the speaker. The user hears the audio being played from the device: "The weather for this week is mostly rainy."

[1447] This series of steps builds a system that allows users to receive natural responses in different languages ​​that take emotions into account in response to their input voice.

[1448] Example 2

[1449] 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."

[1450] Conventional voice conversation systems generate responses without considering the user's emotions, resulting in a poor user experience. They also struggle to achieve natural multilingual conversations, which can lead to a lack of accurate understanding of the user's intent. Furthermore, processing voice data is complex, making it difficult to respond in real time. Technology to resolve these issues is needed.

[1451] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1452] In this invention, the server includes means for receiving voice input and saving it as a digital voice file, means for transmitting the voice data to the server, means for converting the voice input into text format using a voice recognition service, means for analyzing the voice data using an emotion engine to recognize the user's emotional state, means including a natural language processing engine for analyzing the user's intention and generating an appropriate response, means for adjusting the tone of the generated response based on the emotion recognition result, means for translating the adjusted response into another language, means for generating the translated text as voice data using a voice synthesis engine, and means for transmitting the generated voice data to a terminal for playback. This enables natural, easy-to-understand multilingual conversations that take the user's emotions into consideration.

[1453] "Voice input" refers to the voice that a user utters into a terminal, and is the subject of processing by voice recognition technology.

[1454] A "digital audio file" refers to data that has been converted from analog audio into digital format and saved in formats such as WAV or MP3.

[1455] "Server" means a central processing unit that receives, processes, and generates responses from voice data, including cloud-based services.

[1456] A "speech recognition service" is a technology or service that converts voice data into text, typically using a cloud-based API.

[1457] "Text format" refers to a format in which voice data is expressed as text information, and is used in natural language processing and translation.

[1458] An "emotion engine" refers to a technology or system that analyzes and recognizes a user's emotional state from the pitch, speed, volume, etc. of voice data.

[1459] A "natural language processing engine" is a technology or system that analyzes user intent from text data and generates appropriate responses.

[1460] "Response tone" refers to the emotional and expressive modulation of the generated response voice, which changes depending on the user's emotional state.

[1461] "Translation" refers to the process of converting text data expressed in one language into another, supporting multilingual conversations.

[1462] A "speech synthesis engine" is a technology or system that converts text data into voice data, and is used to generate natural-sounding voices.

[1463] A "terminal" is a device used by a user, which has the functions of capturing audio, transmitting and receiving audio data, and playing audio.

[1464] The present invention is a system that combines a multilingual voice conversation system with an emotion engine that recognizes user emotions. The system program uses specific hardware and software to efficiently process user voice input and provide natural, easy-to-understand multilingual responses.

[1465] First, when a user speaks into the device, the device's microphone captures the voice. This voice data is saved as a digital audio file (e.g., WAV format). The device then sends the saved audio file to the server. During this process, metadata (e.g., user ID, timestamp) is added to the voice data.

[1466] The server analyzes the received voice data using a cloud-based voice recognition service. Specific services include commonly used voice recognition APIs (e.g., Speech-to-Text APIs). This service converts the voice data into text format. At the same time, the server uses an emotion engine to analyze the pitch, speed, and volume of the voice data to recognize the user's emotional state. This emotion engine identifies emotional states such as joy, sadness, anger, and surprise.

[1467] The server then passes the converted text data to a natural language processing engine to analyze the user's intent. The natural language processing engine used here may utilize technologies from the BERT or GPT series. The server generates an appropriate response based on the analysis results. For example, if the user says "hello," the server generates the response "hello!"

[1468] The generated response is adjusted based on the emotion recognition results from the emotion engine. For example, if the user is recognized as sad, the tone of the response is adjusted to be gentler. This adjusted response text is passed to a translation engine and translated into the specified language. Services such as Google Translate API and DeepL are used as translation engines.

[1469] The server then passes the translated text to a speech synthesis engine, typically Amazon Polly or the Text-to-Speech API, to generate audio data, which is then saved as an audio file in MP3 format or similar.

[1470] Finally, the server sends the generated voice data to the device, which then decodes the data and plays it back through the speaker, allowing the user to hear a natural-sounding response.

[1471] Specific examples

[1472] Examples of everyday conversation

[1473] User: "Hello"

[1474] The device captures the audio, stores it and sends it to a server.

[1475] The server converts the speech to text "Hello," and uses an emotion engine to analyze pitch, rate, and volume to recognize the happy emotion. It uses natural language processing to recognize it as a greeting and generates the appropriate response: "Hello!"

[1476] The server translates the response into English "Hello!" and generates voice data using a speech synthesis engine.

[1477] The device plays "Hello!", which the user hears.

[1478] Examples of professional conversations

[1479] User: "Tell me about the advances in pacemakers."

[1480] The device captures the audio, stores it and sends it to a server.

[1481] The server converts the speech into text, uses an emotion engine to recognize tension in the user's voice, and uses a natural language processing engine to reference medical expertise to generate a response such as, "In recent years, pacemakers have become smaller and lithium batteries have significantly longer battery life."

[1482] The server translates the generated text into the specified language (e.g., English) and generates voice data using a speech synthesis engine, adjusting the tone to be gentler to reduce tension.

[1483] The device plays audio data to the user that says, "In recent years, pacemakers have become smaller and the lifespan of lithium batteries has significantly increased."

[1484] As described above, the present invention can provide multilingual conversations in a more natural and understandable manner by recognizing the user's emotions and adjusting responses with an appropriate tone.

[1485] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1486] Step 1:

[1487] The user starts speaking into the device. The device's microphone captures this voice and saves it as a digital audio file (e.g., WAV format). The input is the user's voice data, and the output is the saved digital audio file.

[1488] Step 2:

[1489] The device sends the stored audio file and metadata (e.g., user ID, timestamp) to the server. The input is the digital audio file and metadata, and the output is the data sent to the server. Communication is secure using the HTTP protocol.

[1490] Step 3:

[1491] The server sends the received audio file to a cloud-based speech recognition service (e.g., speech recognition API) to convert the audio into text. The input is the audio file, and the output is the converted text data. The server calls the speech recognition API, and the audio data is converted into text data.

[1492] Step 4:

[1493] The server passes the converted text data to an emotion engine, which analyzes the pitch, speed, and volume of the audio data to recognize the user's emotional state. The input is the text data and audio data properties, and the output is the recognized emotional state. The server uses the emotion engine to perform polarity analysis and acoustic feature analysis.

[1494] Step 5:

[1495] The server passes the emotion recognition results and converted text data to a natural language processing engine (e.g., a generative AI model) to analyze the user's intent. The input is text data and emotion recognition results, and the output is a response text based on the user's intent. The natural language processing engine performs calculations based on the prompt sentence and generates an appropriate response.

[1496] Step 6:

[1497] The server adjusts the generated response text based on the emotion recognition results. The input is the generated response text and the emotion recognition results, and the output is the adjusted response text. The server adjusts the tone and style according to the emotion.

[1498] Step 7:

[1499] The server passes the adjusted response text to a translation engine to translate it into the specified language. The input is the adjusted response text and the output is the translated text. The translation engine (e.g., translation API) enables multilingual support.

[1500] Step 8:

[1501] The server passes the translated text to a speech synthesis engine, which generates the audio data. The input is the translated text, and the output is the generated audio data (e.g., MP3 format). The audio data is generated using a speech synthesis engine (e.g., a speech synthesis API).

[1502] Step 9:

[1503] The server sends the generated audio data to the device. The input is the generated audio data, and the output is the audio data sent to the device. The audio data is securely sent to the device using the HTTP protocol.

[1504] Step 10:

[1505] The device decodes the received audio data and plays it through the speaker. The input is the audio data sent to the device, and the output is the audio the user hears. The device decodes the audio data and plays it at the appropriate volume and quality.

[1506] (Application example 2)

[1507] 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."

[1508] Multilingual voice conversation systems often lack the ability to recognize users' emotions and adjust responses appropriately, which can lead to a decline in the quality of communication. Furthermore, security services require the ability to recognize visitors' emotions and provide accurate warnings to users, but current systems lack such advanced emotion recognition capabilities, which creates a problem of insufficient safety.

[1509] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1510] In this invention, the server includes means for receiving voice input from a user, means for converting the voice input into text format, means for analyzing the user's intention from the text format data and generating an appropriate response, means for translating the generated response into another language, means for generating voice data from the translated text by speech synthesis, means for transmitting the voice data to the user's terminal and playing it back, means for recognizing the user's emotion and appropriately adjusting the response, and means for transmitting the emotion recognition result and the voice data for security notification. This enables a natural response that takes the user's emotion into consideration, making it possible to realize an advanced emotion-recognition security assistant that contributes to improved security.

[1511] The "means for receiving voice input" is a function for capturing the user's spoken voice with a device such as a microphone and storing it as digital data.

[1512] The "means for converting into text format" is a technology for converting voice input into text data, and is realized using a voice recognition engine.

[1513] "Means for analyzing the user's intent and generating an appropriate response" refers to a function that analyzes the content of the user's speech using natural language processing, understands the intent, and then creates an appropriate response.

[1514] A "means for translating into another language" refers to a translation engine used to translate the generated response text into another specified language.

[1515] "Means for generating voice data by voice synthesis" refers to a technology that uses a voice synthesis engine to convert text data into voice data.

[1516] The "means for transmitting audio data to a user's terminal and playing it back" is a function for transferring the generated audio data to a user's device and playing it back through a speaker or headphones.

[1517] "Means for recognizing the user's emotions and adjusting the response appropriately" refers to a system that analyzes the user's emotional state from voice data and appropriately changes the tone and content of the response based on the results.

[1518] "Means for sending emotion recognition results and voice data for security notifications" is a function for notifying the user of the results of analysis by the emotion recognition engine and related voice data in real time to their smartphone, etc.

[1519] The present invention is a security assistant system that processes user voice input in multiple languages, recognizes emotions, and generates appropriate responses. This system is realized by a combination of hardware and software.

[1520] The server receives the user's voice input using a microphone installed on the user's device. The voice data captured by the microphone is saved in digital format and then converted into text format. A cloud-based speech recognition service is used as the speech recognition engine for the voice conversion. Specifically, the Google Speech-to-Text API is used.

[1521] After the voice input is converted to text, the server uses a natural language processing engine (e.g., OpenAI's GPT-3) to analyze the user's intent and generate an appropriate response. During this analysis step, a generative AI model operates based on a prompt sentence. For example, the following prompt sentence is used:

[1522] The visitor seems angry. Generate an appropriate response.

[1523] Once the analysis is complete, the generated response is translated into another language using the Google Translate API or other translation engines. The translated text data is then converted into audio data using a speech synthesis engine (e.g., Google Text-to-Speech API). This audio data is sent to the user's device and played through the speaker.

[1524] Regarding emotion recognition, an emotion engine (e.g., IBM Watson Tone Analyzer) is used to recognize emotions from the user's voice, and the tone and content of the generated response are adjusted according to the user's emotional state.

[1525] Furthermore, emotion recognition results and voice data are sent to the user's smartphone for security notifications. For example, if a visitor is recognized as angry, a warning message will be displayed on the user's smartphone saying, "The visitor is angry. Please be careful." This function allows the user to understand the visitor's emotional state and respond safely.

[1526] For example, if a visitor says "Open quickly!" through the doorbell, the system captures the speech and converts it into text using the Google Speech-to-Text API. The speech is then analyzed by OpenAI's GPT-3 and the emotion is recognized as "anger" using IBM Watson Tone Analyzer. The response generated is "Please wait a moment, we'll respond shortly," which is translated using the Google Translate API and then converted into speech using the Google Text-to-Speech API. Finally, the speech data and emotion recognition results are sent to the user's smartphone, prompting them to take an appropriate action.

[1527] In this way, the present invention not only recognizes the user's intention and emotion from their voice input and responds appropriately in multiple languages, but also has a function to notify the emotion recognition results to enhance security.

[1528] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1529] Step 1:

[1530] The user speaks into the terminal.

[1531] Input: User's voice

[1532] Output: Audio data (WAV format)

[1533] Specific operation: The microphone on the device captures the user's voice and saves it as a WAV audio file, which is then used for further processing.

[1534] Step 2:

[1535] The terminal transmits the voice data to the server.

[1536] Input: Audio data (WAV format)

[1537] Output: None (forwarded to server)

[1538] Specific operation: The device sends the captured audio data to the server, along with the necessary metadata (user ID, timestamp, etc.).

[1539] Step 3:

[1540] The server converts the audio data into text format.

[1541] Input: Audio data (WAV format)

[1542] Output: Text data

[1543] What happens: The server uses the Google Speech-to-Text API to convert the audio data into text, which is then used in the next analysis step.

[1544] Step 4:

[1545] The server performs emotion recognition.

[1546] Input: Audio data (WAV format)

[1547] Output: Sentiment analysis results (e.g. anger, joy, sadness, etc.)

[1548] How it works: The server uses IBM Watson Tone Analyzer to analyze the pitch, rate, and volume of the audio data to identify the user's emotions.

[1549] Step 5:

[1550] The server analyzes the user's intent from the text data and generates an appropriate response.

[1551] Input: Text data

[1552] Output: Response text

[1553] How it works: The server uses OpenAI's GPT-3 model to analyze the text data and understand the user's intent. Based on the analysis results, it generates an appropriate response in text format.

[1554] Step 6:

[1555] The server translates the generated response text into another language.

[1556] Input: Response text

[1557] Output: Translated text

[1558] Specific behavior: The server uses a translation engine such as the Google Translate API to translate the generated response text into another specified language.

[1559] Step 7:

[1560] The server generates the translated text as audio data.

[1561] Input: Translated text

[1562] Output: Audio data (MP3 format)

[1563] How it works: The server uses the Google Text-to-Speech API to convert the translated text into audio data, which is then saved in MP3 format.

[1564] Step 8:

[1565] The server sends the audio data to the terminal and plays it back.

[1566] Input: Audio data (MP3 format)

[1567] Output: None (transfer to terminal)

[1568] Specific operation: The server sends the generated audio data to the terminal, and the terminal decodes the received audio data and plays it through the speaker.

[1569] Step 9:

[1570] The server sends the emotion recognition results and voice data to the user's device for security notification.

[1571] Input: Emotion recognition results, audio data

[1572] Output: Security Notice

[1573] Specific operation: The server sends the emotion recognition results and related voice data to the user's smartphone, and displays warnings and notifications based on the user's emotional state.

[1574] 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.

[1575] 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.

[1576] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1577] 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.

[1578] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1579] 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.

[1580] 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).

[1581] 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.

[1582] 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."

[1583] 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.

[1584] 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).

[1585] 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.

[1586] 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.

[1587] 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.

[1588] 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.

[1589] 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.

[1590] 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.

[1591] 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.

[1592] 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.

[1593] 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.

[1594] 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.

[1595] The following is further disclosed regarding the above embodiment.

[1596] (Claim 1)

[1597] means for receiving voice input from a user;

[1598] means for converting said speech input into a text format;

[1599] means for analyzing a user's intent from the text data and generating an appropriate response;

[1600] means for translating the generated response into another language;

[1601] means for generating voice data from the translated text by speech synthesis;

[1602] means for transmitting the audio data to a user terminal and playing it back;

[1603] A system including:

[1604] (Claim 2)

[1605] 10. The system of claim 1, wherein the means for receiving voice input uses a microphone on the user's terminal.

[1606] (Claim 3)

[1607] 10. The system of claim 1, wherein the speech recognition engine uses a cloud-based speech recognition service.

[1608] "Example 1"

[1609] (Claim 1)

[1610] means for receiving voice input from a user;

[1611] means for storing said voice input as a digital voice file and transmitting it to a server;

[1612] means for converting the audio data into a text format;

[1613] means for analyzing a user's intent from the text data and generating an appropriate response;

[1614] means for translating the generated response into multiple languages;

[1615] means for generating voice data from the translated text by speech synthesis;

[1616] means for transmitting the audio data to a user's terminal and playing it back;

[1617] A system including:

[1618] (Claim 2)

[1619] 10. The system of claim 1, wherein the means for receiving speech input uses an input device on the user's terminal.

[1620] (Claim 3)

[1621] 10. The system of claim 1, wherein the speech recognition engine uses a cloud-based speech recognition service.

[1622] "Application Example 1"

[1623] (Claim 1)

[1624] means for receiving voice input from a user;

[1625] means for converting said speech input into a text format;

[1626] means for analyzing a user's intent from the text data and generating an appropriate response;

[1627] means for translating the generated response into another language;

[1628] means for generating voice data from the translated text by speech synthesis;

[1629] means for transmitting the audio data to a user's smartphone and playing it back;

[1630] A means of realizing dialogue for ordering food delivery in multiple languages;

[1631] A system including:

[1632] (Claim 2)

[1633] 10. The system of claim 1, wherein the means for receiving voice input uses a microphone on the user's smartphone.

[1634] (Claim 3)

[1635] 10. The system of claim 1, wherein the speech recognition engine uses a cloud-based speech recognition service.

[1636] "Example 2: Combining Emotion Engines"

[1637] (Claim 1)

[1638] means for receiving audio input from a user and storing it as a digital audio file;

[1639] means for transmitting the voice data to a server;

[1640] means for converting the voice input into a text format using a voice recognition service;

[1641] means for analyzing the voice data converted into text format using an emotion engine to recognize the emotional state of the user;

[1642] means including a natural language processing engine for analyzing a user's intent and generating an appropriate response;

[1643] means for adjusting the tone of the generated response based on emotion recognition results;

[1644] means for translating the tailored response into another language;

[1645] means for generating voice data from the translated text using a voice synthesis engine;

[1646] means for transmitting the generated voice data to a terminal and playing it back;

[1647] A system including:

[1648] (Claim 2)

[1649] 10. The system of claim 1, wherein the means for receiving voice input uses a microphone on the user's terminal.

[1650] (Claim 3)

[1651] 10. The system of claim 1, wherein the speech recognition engine and speech synthesis engine use cloud-based services.

[1652] "Application example 2 when combining emotion engines"

[1653] (Claim 1)

[1654] means for receiving voice input from a user;

[1655] means for converting said speech input into a text format;

[1656] means for analyzing a user's intent from the text data and generating an appropriate response;

[1657] means for translating the generated response into another language;

[1658] means for generating voice data from the translated text by speech synthesis;

[1659] means for transmitting the audio data to a user terminal and playing it back;

[1660] a means for recognizing a user's emotions and adjusting responses appropriately;

[1661] means for transmitting the emotion recognition results and the voice data for security notification;

[1662] A system including:

[1663] (Claim 2)

[1664] 10. The system of claim 1, wherein the means for receiving voice input uses a microphone on the user's terminal.

[1665] (Claim 3)

[1666] 10. The system of claim 1, wherein the speech recognition engine uses a cloud-based speech recognition service. [Explanation of symbols]

[1667] 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 voice input from a user; means for converting said speech input into a text format; means for analyzing a user's intent from the text data and generating an appropriate response; means for translating the generated response into another language; means for generating voice data from the translated text by speech synthesis; means for transmitting the audio data to a user terminal and playing it back; A system including:

2. 2. The system of claim 1, wherein the means for receiving voice input uses a microphone on the user's terminal.

3. The system of claim 1 , wherein the speech recognition engine uses a cloud-based speech recognition service.

Images