system

Abstract

We provide the system. [Solution] A device for acquiring voice information, A device that identifies the speaker from acquired voice information, A device that restricts response information when it is determined that the identified speaker is not a specific user, A device that presents restricted response information to the user, A device that applies a response generation algorithm to provide a safe response when different speakers are present, A system that includes this.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In a conversation using an AI agent, there is a risk that the user's privacy information may be inadvertently leaked to others. In particular, in a scenario where multiple people participate, there is a problem that the user's personal information may leak out when the AI agent inadvertently provides information related to privacy. An object of the present invention is to solve this problem and provide an environment in which users can use the AI agent with confidence.

Means for Solving the Problems

[0005] This invention comprises means for acquiring audio data and means for identifying the speaker from the acquired audio data. It also includes means for restricting the response content when it is determined that the identified speaker is not a specific user, and means for providing the response content to the user. This enables speaker recognition, response restriction, and protection of privacy information, thereby preventing the leakage of the user's personal information.

[0006] "Audio data" refers to information that represents sound in digital format and is acquired through devices such as microphones.

[0007] The term "speaker" refers to the person who produced the voice that formed the basis of the audio data.

[0008] A "user" is defined as an entity that operates or utilizes an AI agent, and is a subject of privacy in this invention.

[0009] "To identify" means to judge an individual or object using specific means and to distinguish it from others.

[0010] "To restrict" means to curb or prohibit the dissemination or use of information under specific conditions.

[0011] "Response" refers to the reply or reaction generated in response to the input information or question.

[0012] "Privacy information" refers to all information that is directly or indirectly related to the identification of an individual and that requires protection. [Brief explanation of the drawing]

[0013] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of the data processing device and smart device according to the first embodiment. [Figure 3]It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Embodiments for Carrying Out the Invention

[0014] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0015] First, the language used in the following description will be explained.

[0016] In the following embodiments, the labeled processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0017] In the following embodiments, the labeled RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0018] In the following embodiments, the labeled storage is one or more non-volatile storage devices that store various programs, various parameters, and the like. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0019] In the following embodiments, the labeled communication I / F (Interface) is an interface including a communication processor, an antenna, and the like. The communication I / F manages communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0020] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0021] [First Embodiment]

[0022] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0023] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0024] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0025] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0026] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0027] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0028] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0029] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0031] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0032] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0033] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0034] This invention is a privacy protection system using speech recognition technology, and is particularly aimed at protecting user privacy when an AI agent participates in a conversation involving multiple people. This system consists of multiple modules operating on a server and a terminal. Specific embodiments are described below.

[0035] System Configuration

[0036] The server is equipped with a speech recognition engine and is responsible for receiving and processing voice data transmitted from the terminal. The terminal has the means to capture the user's voice data and communicate with the server.

[0037] How to use

[0038] When a user initiates a voice interaction with an AI agent, the device uses its microphone to capture the user's speech as audio data. This audio data is transmitted to a server in real time. The server identifies the speaker based on the audio data, and if a person other than the user is detected, it returns that information to the device.

[0039] After receiving information from the server, the device notifies the user that another person is present in the conversation. At this point, the device enables restricted mode for privacy protection and continues the conversation after obtaining the user's confirmation.

[0040] When restricted mode is enabled, the server applies a specific response generation algorithm to queries that may contain private information to provide secure content. For example, based on user instructions such as "Do not share my personal information," the server generalizes the response.

[0041] Specific example

[0042] As an example, imagine a user asking an AI agent, "Tell me my schedule for this week," in an environment where a friend is nearby. In this system, as soon as the server detects the friend's voice, a notification is sent to the device. The device informs the user, "There is another person in the conversation. Privacy mode is enabled." As a result, the server's response will be privacy-conscious, such as, "You can check the details of the schedule individually later."

[0043] In this way, users can use the AI ​​agent with peace of mind and prevent the leakage of personal information. This system provides a means to protect user privacy at a high level, especially when used in public places or workplaces.

[0044] The following describes the processing flow.

[0045] Step 1:

[0046] The device captures the user's spoken voice using its microphone and sends it to the server as digital audio data.

[0047] Step 2:

[0048] The server analyzes the received audio data using a speech recognition engine and compares it with a voice profile that characterizes the speaker to determine whether it is the user's voice.

[0049] Step 3:

[0050] The server determines, based on the identification results, whether there are speakers other than the user in the conversation. It then sends this result to the terminal.

[0051] Step 4:

[0052] The device receives the determination result from the server, and if another person is included in the conversation, it displays or voices a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled."

[0053] Step 5:

[0054] When privacy mode is enabled, the server generates appropriate responses to incoming user queries in a way that does not include private information. For example, responses containing the user's personal information are replaced with generalized language or secure content.

[0055] Step 6:

[0056] The terminal provides the user with the response received from the server. In this process, the response is presented in a format that protects personal information through privacy mode.

[0057] Step 7:

[0058] Through these processes, users can continue their conversation with the AI ​​agent in a safe and private manner.

[0059] (Example 1)

[0060] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0061] In conversational systems using speech recognition technology, there is a need to effectively detect the presence of others in real time and generate appropriate responses while protecting user privacy at a high level. In particular, it is necessary to minimize the risk of unintentional leakage of personal information when used in public places or workplaces.

[0062] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0063] In this invention, the server includes means for transmitting voice information to an information processing device, means for authenticating the speaker using a voice recognition function, and means for setting a privacy level. This makes it possible to automatically activate privacy mode when another person is involved in the conversation and respond to the user while protecting personal information.

[0064] "Audio information" refers to information obtained by converting speech acquired through input devices such as microphones into digital data.

[0065] The term "speaker" refers to the person making a statement based on the acquired audio information, and is identified based on their voice pattern.

[0066] "Users" refer to individuals who directly operate or use this system, and are recipients of information that the system particularly needs to protect.

[0067] An "information processing device" is a device that performs various analyses and processes using acquired audio information, and includes servers and computer devices.

[0068] "Speech recognition functionality" is a technology that analyzes voice input, converts it into text information, and identifies the speaker, and is a core technology used in this system.

[0069] "Privacy level" refers to the degree of security set to control the content of responses for the purpose of protecting personal information.

[0070] A "response generation algorithm" refers to a method or computational means for generating information presented to the user based on voice information, and is used to produce highly secure responses.

[0071] This invention is a privacy protection system utilizing speech recognition technology, and is particularly aimed at ensuring privacy when an AI agent assists a user in conversation. The following details an embodiment of the system.

[0072] server

[0073] The server is equipped with a high-performance speech recognition engine and receives and processes voice information transmitted from the terminal in real time. The necessary environment for this is a general cloud computing platform and a powerful voice analysis engine. Specifically, a general-purpose voice processing API can be used. The server analyzes the voice information and identifies whether the speaker is someone other than the user. It also generates a secure response that takes privacy levels into consideration using an AI model. The generated response is generalized and may include something like, "You can check the details of the schedule individually later."

[0074] terminal

[0075] The device captures the user's speech using a highly sensitive microphone. This device is often a portable device or smartphone equipped with noise-canceling capabilities. The device sends the acquired audio information to a server, and a sophisticated security protocol is used for data transfer. When the server detects another person, the device uses that information to display a message to the user such as, "There is another person in the conversation. Privacy mode is enabled."

[0076] User

[0077] Users obtain necessary information by conversing with an AI agent via voice, with the server and terminal assisting in protecting privacy during this process. A voice confirmation feature ensures that personal information can be handled safely even in public places. This feature alerts the user if other people are involved in the conversation. An example of such a prompt might be the user saying, "Don't share my personal information."

[0078] Specific example

[0079] For example, if a user asks an AI assistant in a cafe, "Tell me about tomorrow's meeting," the server filters out ambient noise while detecting other people's voices. After detection, the device activates privacy mode and provides a generalized response to the user's query. In this way, an environment is created where users can use AI with peace of mind.

[0080] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0081] Step 1:

[0082] The device captures the user's speech using a high-sensitivity microphone. The input is the user's voice, and the output is digital audio data. During this process, noise cancellation technology is used to reduce external noise and obtain clear audio data.

[0083] Step 2:

[0084] The terminal encrypts the captured digital audio data and sends it to the server. The input is the audio data acquired by the terminal, and the output is the securely transmitted audio data. This process applies an encryption protocol (e.g., TLS) to ensure security.

[0085] Step 3:

[0086] The server analyzes the received audio data using a speech recognition engine to identify the speaker. The input is encrypted audio data, and the output is speaker identification information. The server analyzes the audio data and identifies the speaker by comparing it to known speech patterns.

[0087] Step 4:

[0088] If the server determines from the analysis results that the speaker is someone other than the user, it sends that information to the terminal. The input is the result of speaker identification, and the output is a notification of detection of another person. The server returns the determination result to the terminal in real time.

[0089] Step 5:

[0090] When the device receives information about the detection of another person, it displays a message to the user stating, "There is another person in the conversation. Privacy mode is enabled." The input is notification information from the server, and the output is a warning message to the user. This allows the user to recognize the presence of another person and take necessary action.

[0091] Step 6:

[0092] When a user selects privacy mode, the device sends that selection to the server. The input is the user's selection information, and the output is the selection notification sent to the server. This selection automatically triggers privacy protection measures.

[0093] Step 7:

[0094] When the server confirms privacy mode, it uses a generative AI model to generate a secure response. The input is the user's query information and privacy level, and the output is a generalized secure response. The generative AI model provides information that takes privacy into consideration.

[0095] Step 8:

[0096] The terminal displays the response sent from the server to the user. The input is the response data from the server, and the output is the presentation of information to the user. The response is displayed as audio or text, allowing the user to obtain information with confidence.

[0097] (Application Example 1)

[0098] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0099] In smart cities, using voice assistants in public places poses a risk of privacy breaches due to others overhearing conversations. Furthermore, in environments with multiple speakers, responses that protect the privacy of specific users are required. Conventional technologies have been unable to adequately address these issues.

[0100] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0101] In this invention, the server includes a device for acquiring voice information, a device for identifying the speaker from the acquired voice information, and a device for applying a response generation algorithm to provide a secure response when different speakers are present. This makes it possible to securely protect the user's privacy information even when using a voice assistant in a public place.

[0102] "Audio information" refers to a data format in which audio is digitized, and is used for speech recognition and processing.

[0103] "Device" refers to a combination of hardware and software designed to perform a specific function, and in this invention, it refers to a system for acquiring and processing sound.

[0104] "Speaker" refers to the person who produces auditory information, and in speech recognition, it is the target for identifying an individual who possesses a specific speech pattern.

[0105] A "user" is the person who uses this system, and usually refers to the individual who gives instructions to the voice assistant.

[0106] "Response information" refers to the information that a voice assistant provides to the user, including answers to questions and guidance.

[0107] A "response generation algorithm" refers to a processing method for generating appropriate response information based on user input and circumstances, and can apply special conditions to provide a secure response.

[0108] "Privacy Mode" is a function that adjusts the system's response to protect the user's personal information and is activated when another person's speech is detected.

[0109] To implement this invention, a device and a server are required to acquire and analyze voice information. A smartphone or smart glasses with a built-in microphone can be used as the voice acquisition device. This device is responsible for acquiring voice information from the user and transmitting it to the server.

[0110] The server receives audio information and identifies the speaker using speech recognition technology. Specifically, it uses a speech recognition engine such as Google® Cloud Speech-to-Text API to convert audio information into text in real time and analyze the speaker's characteristics. At this time, the server uses a generative AI model to detect whether the speaker is a user other than a specific user.

[0111] If different speakers are present, the server applies a response generation algorithm to generate secure response information. This process uses AI models and predefined contexts to create a privacy-conscious, generic response. The generated response information is sent to the user's device and presented to the user as audio or text.

[0112] As a concrete example, consider a scenario where a user on public transport asks a voice assistant, "What are the plans for tomorrow?" In this case, the server analyzes the surrounding audio and, if it detects other speakers, activates privacy mode and returns a response such as, "More details later."

[0113] An example of a prompt would be: "Please provide a privacy-sensitive response for the voice assistant in the following scenario: The user asks, 'What is my health information?' There are third parties present."

[0114] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0115] Step 1:

[0116] The device acquires the user's speech using a microphone. The acquired voice information is temporarily stored on the device as digital data. Because this data retains the characteristics of the voice, it can be identified in the next processing step.

[0117] Step 2:

[0118] The device transmits the acquired voice information to the server. Upon transmission of this data, the server begins analyzing the data using speech recognition technology. An internet connection is required for data transmission.

[0119] Step 3:

[0120] The server receives the audio information and uses a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the audio information into text data. The converted text data contains the speaker's voice characteristics and is used as data to identify the speaker.

[0121] Step 4:

[0122] The server uses a generative AI model based on the converted text data to identify the speaker. This identification process involves analyzing speech patterns to distinguish between a specific user and others. As a result, it is determined whether the speaker is a specific user or someone else.

[0123] Step 5:

[0124] If the server determines that the speaker is someone else, it applies a response generation algorithm. This algorithm is used to generate safe and generalized response information. A generative AI model is used for response generation, and context-appropriate prompt sentences are applied.

[0125] Step 6:

[0126] The server sends the generated response information to the terminal. This response information is designed with privacy in mind and may include messages such as, "Please check the details later."

[0127] Step 7:

[0128] The terminal transmits the received response information to the user. This is done using audio output or screen display to allow the user to confirm the response.

[0129] This processing flow makes it possible to use voice assistants in public places while ensuring user privacy.

[0130] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0131] This invention is a privacy protection system using speech recognition technology and an emotion engine, aiming to protect privacy with high accuracy by simultaneously recognizing the speaker and emotions based on the user's voice data. Specific embodiments are described below.

[0132] System Configuration

[0133] The server is equipped with a speech recognition engine and an emotion engine, and has the function of receiving and processing voice data transmitted from the terminal. The terminal provides a means to acquire the user's voice data and communicate with the server.

[0134] How to use

[0135] When a user begins interacting with the AI ​​agent, the device captures the audio using its microphone and sends it to the server as digital audio data. The server analyzes the received audio data with a speech recognition engine to simultaneously identify the emotions of both the speaker and the user. The emotion engine identifies the emotional state of the user based on the tone and speed of their voice.

[0136] Based on the speaker identification results, if someone other than the user is participating in the conversation, the server returns that information to the terminal. The terminal then notifies the user that "Someone else is participating in the conversation. Privacy mode is enabled." In addition, the tone of the response and the level of detail of the information are dynamically adjusted according to the recognized emotions of the user.

[0137] Specific example

[0138] For example, consider a scenario where a user asks "What are today's tasks?" in a tired voice. Assume a family member is nearby. In this system, the server detects the family member's voice and recognizes the user's emotional state as "fatigued." The device then informs the user, "Someone else is in the conversation. Privacy mode is enabled. More details will be provided later." At this point, the server adjusts its response, taking the user's state into consideration, to avoid overloading the user with excessive information.

[0139] In this way, users can use the AI ​​agent with peace of mind, and interaction that flexibly responds to the user's emotions can be achieved while protecting their privacy. This system provides a means to ensure both user privacy and comfort, especially when used in homes or public places.

[0140] The following describes the processing flow.

[0141] Step 1:

[0142] The device captures the user's speech using a microphone and sends it to the server as digital audio data.

[0143] Step 2:

[0144] The server processes the received audio data with a speech recognition engine and identifies the user's speech by matching it with a speech profile that characterizes the speaker.

[0145] Step 3:

[0146] The server inputs the voice data into the emotion engine, which analyzes the tone, speed, and volume of the user's voice to recognize their emotions.

[0147] Step 4:

[0148] The server determines whether there are participants other than the user in the conversation based on the speaker identification results and emotion recognition results. This determination result is then sent to the terminal.

[0149] Step 5:

[0150] The device receives the detection result from the server and, if another person is in the conversation, notifies the user with the message, "Another person is participating in the conversation. Privacy mode is enabled."

[0151] Step 6:

[0152] The server generates responses to user queries with privacy mode enabled and taking user sentiment into consideration. The responses are selected to exclude private information and are adjusted to a tone that respects the user's feelings.

[0153] Step 7:

[0154] The device provides the user with a tailored response, which is considerate of the user's emotional state.

[0155] Step 8:

[0156] Users can continue their conversations with AI agents in a safe and satisfying manner, while their privacy is protected and flexible, emotionally-driven dialogue is possible.

[0157] (Example 2)

[0158] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0159] Conventional speech recognition systems have shortcomings, including insufficient speaker identification and subsequent privacy protection, as well as the inability to respond flexibly to the user's emotional state. Furthermore, there is a risk of information leakage when individuals other than the user participate in conversations in certain environments. There is a need to address these challenges and simultaneously improve both privacy and user experience.

[0160] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0161] In this invention, the server includes means for acquiring voice data, means for identifying the speaker from the acquired voice data, and means for analyzing the emotional state from the acquired voice data. This enables the activation of a privacy mode based on speaker identification and the adjustment of the tone of response and level of detail of information according to the user's emotions.

[0162] "Audio data" refers to data obtained by capturing the user's voice and converting it into a digital format.

[0163] "Speaker" refers to the individual or person identified as the source of the speech based on the acquired audio data.

[0164] "Privacy Mode" is a protective feature that restricts responses to prevent the leakage of potentially unnecessary information when it is determined that another person is participating in the conversation.

[0165] "Emotional state" refers to a state that indicates psychological or emotional changes or tendencies, analyzed based on the user's voice data.

[0166] "Response tone" refers to the pitch and manner of speaking characteristics of the responses that the system provides to the user, and is usually adjusted based on the emotional state.

[0167] "Information detail" is an indicator that shows the specificity and depth of information included in the response to the user, and it is adjusted according to the user's emotional state.

[0168] This invention is a system that utilizes speech recognition technology and an emotion analysis engine to enable flexible, emotion-responsive dialogue while protecting user privacy. Specific embodiments are described below.

[0169] The server is equipped with a high-performance speech recognition engine and sentiment analysis engine. The speech recognition engine uses commercially available speech analysis software (e.g., common APIs) to convert speech data into text. This process also includes an authentication process to identify who the speaker is. The sentiment analysis engine analyzes parameters such as tone, speed, and volume obtained from the speech data to identify the user's emotional state.

[0170] The terminal is responsible for acquiring the user's voice using a microphone and transmitting it to the server as digital audio data. Since the terminal transmits this audio data to the server in real time, the communication infrastructure requires high reliability.

[0171] For example, if a user asks "What are today's tasks?" in a tired voice, the server analyzes the audio data to identify the user as the speaker and simultaneously analyzes their emotional state as "fatigue." Furthermore, if the presence of other people, such as family members, is detected in the vicinity, the server activates privacy mode based on the speaker identification information and notifies the device accordingly.

[0172] Based on information from the server, the device displays a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled." The server also takes the user's emotional state into account and adjusts the tone of its response and the level of detail in the information provided to avoid unnecessarily burdening the user.

[0173] A concrete example of a prompt in this system would be, "When a user is feeling tired and someone is nearby, how should they respond to ask an AI agent about a task while maintaining their privacy and comfort?"

[0174] In this way, this invention makes it possible to maximize the user experience while maintaining privacy in homes and public places.

[0175] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0176] Step 1:

[0177] The device captures audio using a microphone when the user begins interacting with the AI ​​agent. The input data is the user's voice signal, and the device's audio capture software converts this audio into a digital signal. Specifically, it samples the audio signal and formats it as a digital audio file.

[0178] Step 2:

[0179] The terminal sends captured digital audio data to the server. The input is the converted digital audio data, and the output is a transmission completion report to the server. The terminal streams the audio data using a communication protocol (e.g., HTTP or WebSocket). Specifically, it divides the audio data into packets for transmission and uses compression techniques to minimize latency.

[0180] Step 3:

[0181] The server passes the received audio data to the speech recognition engine. The input is digital audio data from the terminal, and the output is the content converted into text. The server uses the speech recognition engine to generate text from the audio waveform and identify the speaker. Specifically, it performs a speech feature extraction process and model comparison.

[0182] Step 4:

[0183] The server simultaneously uses an emotion analysis engine to identify emotional states from audio data. The input is the same audio data, and the output is the analyzed emotional state tag. The emotion analysis engine analyzes the tone, pitch, and speed of the audio to label emotions. Specifically, it statistically analyzes speech metrics related to emotions and maps them to emotion categories.

[0184] Step 5:

[0185] The server determines the privacy mode based on speaker recognition and emotional state. If the speaker is not the user or if the emotional state meets certain conditions, it returns output to the terminal that enhances privacy. Specifically, it performs a cross-check of speaker information and sets a privacy flag.

[0186] Step 6:

[0187] The terminal notifies the user based on information received from the server. Input is the privacy mode status and response content, while output is the notification message to the user. The terminal uses a display or speech synthesis system to notify the user, "Another person is participating in the conversation. Privacy mode is enabled." Specifically, it performs the notification process via display or voice.

[0188] Step 7:

[0189] The server adjusts the tone and level of detail of its responses to the user according to the recognized emotional state. The input is the result of the emotional analysis, and the output is the configured response parameters. The server considers the user's emotional state to adopt an appropriate tone in its response and controls the level of detail to prevent excessive information transmission. Specifically, the response generation module performs tone adjustment and information calculation.

[0190] (Application Example 2)

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

[0192] Modern voice assistant technology struggles to adequately protect privacy while providing appropriate responses tailored to the user's emotional state. This presents risks of eavesdropping and overload when the user is stressed. The challenge lies in simultaneously improving both privacy and user experience.

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

[0194] In this invention, the server includes a device for acquiring voice data, a device for identifying the speaker from the acquired voice data, a device for restricting the response content if the identified speaker is determined to be someone other than a specific user, a device for presenting the response content to the user, a device for analyzing the user's emotional state, and a device for adjusting the response according to the user's emotional state. This makes it possible to provide responses that match the user's emotional state while protecting the user's privacy.

[0195] "Voice data" refers to digital recordings of users' speech, which are used for purposes such as speech recognition and sentiment analysis.

[0196] A "device" is a component of hardware or software designed to perform a specific function, such as acquiring, analyzing, or responding to audio data.

[0197] The term "speaker" refers to the person who made the utterance in the audio data, and is the subject that needs to be identified and recognized.

[0198] "Discrimination" is the process of analyzing acquired audio data to determine whether or not it belongs to a specific user.

[0199] A "device for restricting response content" is a device that has a mechanism to adjust the scope of information provided in order to protect user privacy, based on the results of identifying the speaker.

[0200] "Users" are those who use the voice recognition system, and their privacy and the quality of their experience should be taken into consideration.

[0201] "Emotional state" refers to the psychological state that indicates emotions contained in the user's utterances, such as joy, sadness, or stress.

[0202] "Analysis" is the process of evaluating the user's emotional state based on voice data and reflecting the results in the response.

[0203] "Response adjustment" is the process of changing the content and tone of a response according to the user's emotional state.

[0204] To implement this invention, the terminal must first be equipped with a device for acquiring audio data. The terminal captures the user's speech with a microphone and transmits it to the server as digital audio data. The server uses speech recognition technology to identify the speaker from this audio data. Specifically, it uses a library such as speech_recognition to convert the audio data into text and then analyzes that text data to identify the speaker.

[0205] If the server determines that the speaker is someone other than the user, it will restrict the content of the response to protect privacy. For example, it will use technologies such as PrivacyGuard to automatically adjust the details of the response so that important information is not leaked even if it is overheard by others.

[0206] In addition, the server utilizes emotion engines such as EmotionEngine to analyze the user's emotional state and dynamically adjusts its response based on the results. For example, if the user is experiencing stress, the server will simplify data submission and respond in a gentle tone. This reduces the burden on the user.

[0207] To give a concrete example, when a user asks a home robot, "What are my plans for today?", the server detects the voices of family members from the acquired audio data and also recognizes the user's "calmness" from their voice. In this case, the server can generate a response such as, "There are other people nearby right now, but please listen without worry. I have a meeting in the morning and free time in the afternoon."

[0208] An example of a prompt for a generative AI model is, "What kind of system can achieve emotional and privacy protection from user voice data in a home robot?" Using this information, users can receive appropriate feedback in a privacy-protected manner.

[0209] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0210] Step 1:

[0211] The device captures the user's voice in real time using a microphone. The acquired voice data is converted into a digital format. At this stage, the input is analog voice data, and the output is digital voice data. The specific operation of the data conversion is to use analog-to-digital conversion technology.

[0212] Step 2:

[0213] The terminal sends digital audio data to the server. The input is the digital audio data converted in step 1, and the output is the audio data received by the server. The specific operation here is the packetization and transmission of data using a network protocol (e.g., TCP / IP).

[0214] Step 3:

[0215] The server converts received digital audio data into text using a speech recognition engine. The input is digital audio data, and the output is the text data of the audio. Specifically, it uses the speech_recognition library to generate text from audio.

[0216] Step 4:

[0217] The server analyzes the text data and identifies the speaker. The input is the text data from step 3, and the output is the speaker identification result. Specific operations include matching the data against a user profile.

[0218] Step 5:

[0219] The server analyzes the user's emotional state using speaker identification results and an emotion engine. Input is speech-to-text data and identified speaker information, while output is the emotion recognition result. Specific operations include a process that utilizes the EmotionEngine to infer emotions from indicators such as tone of voice and word choice.

[0220] Step 6:

[0221] The server is configured to automatically restrict the response content if the speaker is not a user. The input is the speaker identification result, and the output is the restricted response content. The specific operation is performed by filtering the response content using PrivacyGuard technology.

[0222] Step 7:

[0223] The server adjusts its response based on the user's emotional state to generate appropriate feedback. The input is the emotion recognition result and restricted response content, while the output is the adjusted final response. Specific operations include generation processes that modify the tone and level of detail of the response.

[0224] Step 8:

[0225] The server sends the final response to the terminal and presents it to the user. The input is the generated response data from step 7, and the output is what the terminal displays to the user or provides as audio feedback. The specific actions involve network transmission and output processing at the terminal.

[0226] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0227] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0228] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0229] [Second Embodiment]

[0230] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0231] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0232] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0233] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0234] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0235] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0236] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0237] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0238] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0239] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0240] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0241] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0242] This invention is a privacy protection system using speech recognition technology, and is particularly aimed at protecting user privacy when an AI agent participates in a conversation involving multiple people. This system consists of multiple modules operating on a server and a terminal. Specific embodiments are described below.

[0243] System Configuration

[0244] The server is equipped with a speech recognition engine and is responsible for receiving and processing voice data transmitted from the terminal. The terminal has the means to capture the user's voice data and communicate with the server.

[0245] How to use

[0246] When a user initiates a voice interaction with an AI agent, the device uses its microphone to capture the user's speech as audio data. This audio data is transmitted to a server in real time. The server identifies the speaker based on the audio data, and if a person other than the user is detected, it returns that information to the device.

[0247] After receiving information from the server, the device notifies the user that another person is present in the conversation. At this point, the device enables restricted mode for privacy protection and continues the conversation after obtaining the user's confirmation.

[0248] When restricted mode is enabled, the server applies a specific response generation algorithm to queries that may contain private information to provide secure content. For example, based on user instructions such as "Do not share my personal information," the server generalizes the response.

[0249] Specific example

[0250] As an example, imagine a user asking an AI agent, "Tell me my schedule for this week," in an environment where a friend is nearby. In this system, as soon as the server detects the friend's voice, a notification is sent to the device. The device informs the user, "There is another person in the conversation. Privacy mode is enabled." As a result, the server's response will be privacy-conscious, such as, "You can check the details of the schedule individually later."

[0251] In this way, users can use the AI ​​agent with peace of mind and prevent the leakage of personal information. This system provides a means to protect user privacy at a high level, especially when used in public places or workplaces.

[0252] The following describes the processing flow.

[0253] Step 1:

[0254] The device captures the user's spoken voice using its microphone and sends it to the server as digital audio data.

[0255] Step 2:

[0256] The server analyzes the received audio data using a speech recognition engine and compares it with a voice profile that characterizes the speaker to determine whether it is the user's voice.

[0257] Step 3:

[0258] The server determines, based on the identification results, whether there are speakers other than the user in the conversation. It then sends this result to the terminal.

[0259] Step 4:

[0260] The device receives the determination result from the server, and if another person is included in the conversation, it displays or voices a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled."

[0261] Step 5:

[0262] When privacy mode is enabled, the server generates appropriate responses to incoming user queries in a way that does not include private information. For example, responses containing the user's personal information are replaced with generalized language or secure content.

[0263] Step 6:

[0264] The terminal provides the user with the response received from the server. In this process, the response is presented in a format that protects personal information through privacy mode.

[0265] Step 7:

[0266] Through these processes, users can continue their conversation with the AI ​​agent in a safe and private manner.

[0267] (Example 1)

[0268] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0269] In conversational systems using speech recognition technology, there is a need to effectively detect the presence of others in real time and generate appropriate responses while protecting user privacy at a high level. In particular, it is necessary to minimize the risk of unintentional leakage of personal information when used in public places or workplaces.

[0270] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0271] In this invention, the server includes means for transmitting voice information to an information processing device, means for authenticating the speaker using a voice recognition function, and means for setting a privacy level. This makes it possible to automatically activate privacy mode when another person is involved in the conversation and respond to the user while protecting personal information.

[0272] "Audio information" refers to information obtained by converting speech acquired through input devices such as microphones into digital data.

[0273] The term "speaker" refers to the person making a statement based on the acquired audio information, and is identified based on their voice pattern.

[0274] "Users" refer to individuals who directly operate or use this system, and are recipients of information that the system particularly needs to protect.

[0275] An "information processing device" is a device that performs various analyses and processes using acquired audio information, and includes servers and computer devices.

[0276] "Speech recognition functionality" is a technology that analyzes voice input, converts it into text information, and identifies the speaker, and is a core technology used in this system.

[0277] "Privacy level" refers to the degree of security set to control the content of responses for the purpose of protecting personal information.

[0278] A "response generation algorithm" refers to a method or computational means for generating information presented to the user based on voice information, and is used to produce highly secure responses.

[0279] This invention is a privacy protection system utilizing speech recognition technology, and is particularly aimed at ensuring privacy when an AI agent assists a user in conversation. The following details an embodiment of the system.

[0280] server

[0281] The server is equipped with a high-performance speech recognition engine and receives and processes voice information transmitted from the terminal in real time. The necessary environment for this is a general cloud computing platform and a powerful voice analysis engine. Specifically, a general-purpose voice processing API can be used. The server analyzes the voice information and identifies whether the speaker is someone other than the user. It also generates a secure response that takes privacy levels into consideration using an AI model. The generated response is generalized and may include something like, "You can check the details of the schedule individually later."

[0282] terminal

[0283] The terminal captures the user's speech with a high - sensitivity microphone. This terminal is often a portable device or a smartphone equipped with a noise - canceling function. Since the terminal transmits the acquired voice information to the server, an advanced security protocol is used for data transfer. When the server returns detection information of others, the terminal presents a message such as "There are others in the conversation. The privacy mode is enabled" to the user based on that information.

[0284] User

[0285] The user obtains the necessary information by conversing with the AI agent in voice, and the server and the terminal support privacy protection in this process. Even in public places, it has a voice confirmation function so that personal information can be handled with confidence. With this function, the user is warned that other people are included in the conversation. As an example of the prompt at that time, it is conceivable that the user gives an instruction such as "Do not share my personal information".

[0286] Specific example

[0287] For example, when the user asks the AI assistant in a cafe "Tell me about tomorrow's meeting", the server filters the surrounding noise while detecting the voices of others. After detection, the terminal activates the privacy mode and returns a generalized response to the user's query. In this way, the user can construct an environment where they can use AI with confidence.

[0288] The flow of the specific process in Example 1 will be described using FIG. 11.

[0289] Step 1:

[0290] The terminal captures the user's speech with a high - sensitivity microphone. The input is the user's voice, and the output is digital voice data. In this process, noise - canceling technology is used to reduce external noise and obtain clear voice data.

[0291] Step 2:

[0292] The terminal encrypts the captured digital audio data and sends it to the server. The input is the audio data acquired by the terminal, and the output is the securely transmitted audio data. This process applies an encryption protocol (e.g., TLS) to ensure security.

[0293] Step 3:

[0294] The server analyzes the received audio data using a speech recognition engine to identify the speaker. The input is encrypted audio data, and the output is speaker identification information. The server analyzes the audio data and identifies the speaker by comparing it to known speech patterns.

[0295] Step 4:

[0296] If the server determines from the analysis results that the speaker is someone other than the user, it sends that information to the terminal. The input is the result of speaker identification, and the output is a notification of detection of another person. The server returns the determination result to the terminal in real time.

[0297] Step 5:

[0298] When the device receives information about the detection of another person, it displays a message to the user stating, "There is another person in the conversation. Privacy mode is enabled." The input is notification information from the server, and the output is a warning message to the user. This allows the user to recognize the presence of another person and take necessary action.

[0299] Step 6:

[0300] When a user selects privacy mode, the device sends that selection to the server. The input is the user's selection information, and the output is the selection notification sent to the server. This selection automatically triggers privacy protection measures.

[0301] Step 7:

[0302] When the server checks the privacy mode, it uses the generative AI model to generate a safe response. The input is the user's query information and privacy level, and the output is a generalized safe response. The generative AI model provides privacy - considered information.

[0303] Step 8:

[0304] The terminal presents the response sent from the server to the user. The input is the response data from the server, and the output is the information presentation to the user. The response is displayed in voice or text, and the user can obtain information with confidence.

[0305] (Application Example 1)

[0306] Next, Application Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0307] In a smart city, when using a voice assistant in a public place, there is a risk of privacy leakage due to others overhearing the conversation content. Also, in an environment where there are multiple speakers, a response that protects the privacy of a specific user is required. In the prior art, there were problems that could not be fully solved.

[0308] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0309] In this invention, the server includes a device for acquiring voice information, a device for identifying a speaker from the acquired voice information, and a device for applying a response generation algorithm to provide a safe response when different speakers exist. Thereby, even when using a voice assistant in a public place, it is possible to securely protect the user's privacy information.

[0310] "Audio information" refers to a data format in which audio is digitized, and is used for speech recognition and processing.

[0311] "Device" refers to a combination of hardware and software designed to perform a specific function, and in this invention, it refers to a system for acquiring and processing sound.

[0312] "Speaker" refers to the person who produces auditory information, and in speech recognition, it is the target for identifying an individual who possesses a specific speech pattern.

[0313] A "user" is the person who uses this system, and usually refers to the individual who gives instructions to the voice assistant.

[0314] "Response information" refers to the information that a voice assistant provides to the user, including answers to questions and guidance.

[0315] A "response generation algorithm" refers to a processing method for generating appropriate response information based on user input and circumstances, and can apply special conditions to provide a secure response.

[0316] "Privacy Mode" is a function that adjusts the system's response to protect the user's personal information and is activated when another person's speech is detected.

[0317] To implement this invention, a device and a server are required to acquire and analyze voice information. A smartphone or smart glasses with a built-in microphone can be used as the voice acquisition device. This device is responsible for acquiring voice information from the user and transmitting it to the server.

[0318] The server receives audio information and uses speech recognition technology to identify the speaker. Specifically, it uses a speech recognition engine such as the Google Cloud Speech-to-Text API to convert the audio information into text in real time and analyze the speaker's characteristics. At this time, the server uses a generative AI model to detect whether the speaker is a user other than a specific user.

[0319] If different speakers are present, the server applies a response generation algorithm to generate secure response information. This process uses AI models and predefined contexts to create a privacy-conscious, generic response. The generated response information is sent to the user's device and presented to the user as audio or text.

[0320] As a concrete example, consider a scenario where a user on public transport asks a voice assistant, "What are the plans for tomorrow?" In this case, the server analyzes the surrounding audio and, if it detects other speakers, activates privacy mode and returns a response such as, "More details later."

[0321] An example of a prompt would be: "Please provide a privacy-sensitive response for the voice assistant in the following scenario: The user asks, 'What is my health information?' There are third parties present."

[0322] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0323] Step 1:

[0324] The device acquires the user's speech using a microphone. The acquired voice information is temporarily stored on the device as digital data. Because this data retains the characteristics of the voice, it can be identified in the next processing step.

[0325] Step 2:

[0326] The device transmits the acquired voice information to the server. Upon transmission of this data, the server begins analyzing the data using speech recognition technology. An internet connection is required for data transmission.

[0327] Step 3:

[0328] The server receives the audio information and uses a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the audio information into text data. The converted text data contains the speaker's voice characteristics and is used as data to identify the speaker.

[0329] Step 4:

[0330] The server uses a generative AI model based on the converted text data to identify the speaker. This identification process involves analyzing speech patterns to distinguish between a specific user and others. As a result, it is determined whether the speaker is a specific user or someone else.

[0331] Step 5:

[0332] If the server determines that the speaker is someone else, it applies a response generation algorithm. This algorithm is used to generate safe and generalized response information. A generative AI model is used for response generation, and context-appropriate prompt sentences are applied.

[0333] Step 6:

[0334] The server sends the generated response information to the terminal. This response information is designed with privacy in mind and may include messages such as, "Please check the details later."

[0335] Step 7:

[0336] The terminal transmits the received response information to the user. This is done using audio output or screen display to allow the user to confirm the response.

[0337] This processing flow makes it possible to use voice assistants in public places while ensuring user privacy.

[0338] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0339] This invention is a privacy protection system using speech recognition technology and an emotion engine, aiming to protect privacy with high accuracy by simultaneously recognizing the speaker and emotions based on the user's voice data. Specific embodiments are described below.

[0340] System Configuration

[0341] The server is equipped with a speech recognition engine and an emotion engine, and has the function of receiving and processing voice data transmitted from the terminal. The terminal provides a means to acquire the user's voice data and communicate with the server.

[0342] How to use

[0343] When a user begins interacting with the AI ​​agent, the device captures the audio using its microphone and sends it to the server as digital audio data. The server analyzes the received audio data with a speech recognition engine to simultaneously identify the emotions of both the speaker and the user. The emotion engine identifies the emotional state of the user based on the tone and speed of their voice.

[0344] Based on the speaker identification results, if someone other than the user is participating in the conversation, the server returns that information to the terminal. The terminal then notifies the user that "Someone else is participating in the conversation. Privacy mode is enabled." In addition, the tone of the response and the level of detail of the information are dynamically adjusted according to the recognized emotions of the user.

[0345] Specific example

[0346] For example, consider a scenario where a user asks "What are today's tasks?" in a tired voice. Assume a family member is nearby. In this system, the server detects the family member's voice and recognizes the user's emotional state as "fatigued." The device then informs the user, "Someone else is in the conversation. Privacy mode is enabled. More details will be provided later." At this point, the server adjusts its response, taking the user's state into consideration, to avoid overloading the user with excessive information.

[0347] In this way, users can use the AI ​​agent with peace of mind, and interaction that flexibly responds to the user's emotions can be achieved while protecting their privacy. This system provides a means to ensure both user privacy and comfort, especially when used in homes or public places.

[0348] The following describes the processing flow.

[0349] Step 1:

[0350] The device captures the user's speech using a microphone and sends it to the server as digital audio data.

[0351] Step 2:

[0352] The server processes the received audio data with a speech recognition engine and identifies the user's speech by matching it with a speech profile that characterizes the speaker.

[0353] Step 3:

[0354] The server inputs the voice data into the emotion engine, which analyzes the tone, speed, and volume of the user's voice to recognize their emotions.

[0355] Step 4:

[0356] The server determines whether there are participants other than the user in the conversation based on the speaker identification results and emotion recognition results. This determination result is then sent to the terminal.

[0357] Step 5:

[0358] The device receives the detection result from the server and, if another person is in the conversation, notifies the user with the message, "Another person is participating in the conversation. Privacy mode is enabled."

[0359] Step 6:

[0360] The server generates responses to user queries with privacy mode enabled and taking user sentiment into consideration. The responses are selected to exclude private information and are adjusted to a tone that respects the user's feelings.

[0361] Step 7:

[0362] The device provides the user with a tailored response, which is considerate of the user's emotional state.

[0363] Step 8:

[0364] Users can continue their conversations with AI agents in a safe and satisfying manner, while their privacy is protected and flexible, emotionally-driven dialogue is possible.

[0365] (Example 2)

[0366] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0367] Conventional speech recognition systems have shortcomings, including insufficient speaker identification and subsequent privacy protection, as well as the inability to respond flexibly to the user's emotional state. Furthermore, there is a risk of information leakage when individuals other than the user participate in conversations in certain environments. There is a need to address these challenges and simultaneously improve both privacy and user experience.

[0368] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0369] In this invention, the server includes means for acquiring voice data, means for identifying the speaker from the acquired voice data, and means for analyzing the emotional state from the acquired voice data. This enables the activation of a privacy mode based on speaker identification and the adjustment of the tone of response and level of detail of information according to the user's emotions.

[0370] "Audio data" refers to data obtained by capturing the user's voice and converting it into a digital format.

[0371] "Speaker" refers to the individual or person identified as the source of the speech based on the acquired audio data.

[0372] "Privacy Mode" is a protective feature that restricts responses to prevent the leakage of potentially unnecessary information when it is determined that another person is participating in the conversation.

[0373] "Emotional state" refers to a state that indicates psychological or emotional changes or tendencies, analyzed based on the user's voice data.

[0374] "Response tone" refers to the pitch and manner of speaking characteristics of the responses that the system provides to the user, and is usually adjusted based on the emotional state.

[0375] "Information detail" is an indicator that shows the specificity and depth of information included in the response to the user, and it is adjusted according to the emotional state.

[0376] This invention is a system that utilizes speech recognition technology and an emotion analysis engine to enable flexible, emotion-responsive dialogue while protecting user privacy. Specific embodiments are described below.

[0377] The server is equipped with a high-performance speech recognition engine and sentiment analysis engine. The speech recognition engine uses commercially available speech analysis software (e.g., common APIs) to convert speech data into text. This process also includes an authentication process to identify who the speaker is. The sentiment analysis engine analyzes parameters such as tone, speed, and volume obtained from the speech data to identify the user's emotional state.

[0378] The terminal is responsible for acquiring the user's voice using a microphone and transmitting it to the server as digital audio data. Since the terminal transmits this audio data to the server in real time, the communication infrastructure requires high reliability.

[0379] For example, if a user asks "What are today's tasks?" in a tired voice, the server analyzes the audio data to identify the user as the speaker and simultaneously analyzes their emotional state as "fatigue." Furthermore, if the presence of other people, such as family members, is detected in the vicinity, the server activates privacy mode based on the speaker identification information and notifies the device accordingly.

[0380] Based on information from the server, the device displays a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled." The server also takes the user's emotional state into account and adjusts the tone of its response and the level of detail in the information provided to avoid unnecessarily burdening the user.

[0381] A concrete example of a prompt in this system would be, "When a user is feeling tired and someone is nearby, how should they respond to ask an AI agent about a task while maintaining their privacy and comfort?"

[0382] In this way, this invention makes it possible to maximize the user experience while maintaining privacy in homes and public places.

[0383] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0384] Step 1:

[0385] The device captures audio using a microphone when the user begins interacting with the AI ​​agent. The input data is the user's voice signal, and the device's audio capture software converts this audio into a digital signal. Specifically, it samples the audio signal and formats it as a digital audio file.

[0386] Step 2:

[0387] The terminal sends captured digital audio data to the server. The input is the converted digital audio data, and the output is a transmission completion report to the server. The terminal streams the audio data using a communication protocol (e.g., HTTP or WebSocket). Specifically, it divides the audio data into packets for transmission and uses compression techniques to minimize latency.

[0388] Step 3:

[0389] The server passes the received audio data to the speech recognition engine. The input is digital audio data from the terminal, and the output is the content converted into text. The server uses the speech recognition engine to generate text from the audio waveform and identify the speaker. Specifically, it performs a speech feature extraction process and model comparison.

[0390] Step 4:

[0391] The server simultaneously uses an emotion analysis engine to identify emotional states from audio data. The input is the same audio data, and the output is the analyzed emotional state tag. The emotion analysis engine analyzes the tone, pitch, and speed of the audio to label emotions. Specifically, it statistically analyzes speech metrics related to emotions and maps them to emotion categories.

[0392] Step 5:

[0393] The server determines the privacy mode based on speaker recognition and emotional state. If the speaker is not the user or if the emotional state meets certain conditions, it returns output to the terminal that enhances privacy. Specifically, it performs a cross-check of speaker information and sets a privacy flag.

[0394] Step 6:

[0395] The terminal notifies the user based on information received from the server. Input is the privacy mode status and response content, while output is the notification message to the user. The terminal uses a display or speech synthesis system to notify the user, "Another person is participating in the conversation. Privacy mode is enabled." Specifically, it performs the notification process via display or voice.

[0396] Step 7:

[0397] The server adjusts the tone and level of detail of its responses to the user according to the recognized emotional state. The input is the result of the emotional analysis, and the output is the configured response parameters. The server considers the user's emotional state to adopt an appropriate tone in its response and controls the level of detail to prevent excessive information transmission. Specifically, the response generation module performs tone adjustment and information calculation.

[0398] (Application Example 2)

[0399] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0400] Modern voice assistant technology struggles to adequately protect privacy while providing appropriate responses tailored to the user's emotional state. This presents risks of eavesdropping and overload when the user is stressed. The challenge lies in simultaneously improving both privacy and user experience.

[0401] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0402] In this invention, the server includes a device for acquiring voice data, a device for identifying the speaker from the acquired voice data, a device for restricting the response content if the identified speaker is determined to be someone other than a specific user, a device for presenting the response content to the user, a device for analyzing the user's emotional state, and a device for adjusting the response according to the user's emotional state. This makes it possible to provide responses that match the user's emotional state while protecting the user's privacy.

[0403] "Voice data" refers to digital recordings of users' speech, which are used for purposes such as speech recognition and sentiment analysis.

[0404] A "device" is a component of hardware or software designed to perform a specific function, such as acquiring, analyzing, or responding to audio data.

[0405] The term "speaker" refers to the person who made the utterance in the audio data, and is the subject that needs to be identified and recognized.

[0406] "Discrimination" is the process of analyzing acquired audio data to determine whether or not it belongs to a specific user.

[0407] A "device for restricting response content" is a device that has a mechanism to adjust the scope of information provided in order to protect user privacy, based on the results of identifying the speaker.

[0408] "Users" are those who use the voice recognition system, and their privacy and the quality of their experience should be taken into consideration.

[0409] "Emotional state" refers to the psychological state that indicates emotions contained in the user's utterances, such as joy, sadness, or stress.

[0410] "Analysis" is the process of evaluating the user's emotional state based on voice data and reflecting the results in the response.

[0411] "Response adjustment" is the process of changing the content and tone of a response according to the user's emotional state.

[0412] To implement this invention, the terminal must first be equipped with a device for acquiring audio data. The terminal captures the user's speech with a microphone and transmits it to the server as digital audio data. The server uses speech recognition technology to identify the speaker from this audio data. Specifically, it uses a library such as speech_recognition to convert the audio data into text and then analyzes that text data to identify the speaker.

[0413] If the server determines that the speaker is someone other than the user, it will restrict the content of the response to protect privacy. For example, it will use technologies such as PrivacyGuard to automatically adjust the details of the response so that important information is not leaked even if it is overheard by others.

[0414] In addition, the server utilizes emotion engines such as EmotionEngine to analyze the user's emotional state and dynamically adjusts its response based on the results. For example, if the user is experiencing stress, the server will simplify data submission and respond in a gentle tone. This reduces the burden on the user.

[0415] To give a concrete example, when a user asks a home robot, "What are my plans for today?", the server detects the voices of family members from the acquired audio data and also recognizes the user's "calmness" from their voice. In this case, the server can generate a response such as, "There are other people nearby right now, but please listen without worry. I have a meeting in the morning and free time in the afternoon."

[0416] An example of a prompt for a generative AI model is, "What kind of system can achieve emotional and privacy protection from user voice data in a home robot?" Using this information, users can receive appropriate feedback in a privacy-protected manner.

[0417] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0418] Step 1:

[0419] The device captures the user's voice in real time using a microphone. The acquired voice data is converted into a digital format. At this stage, the input is analog voice data, and the output is digital voice data. The specific operation of the data conversion is to use analog-to-digital conversion technology.

[0420] Step 2:

[0421] The terminal sends digital audio data to the server. The input is the digital audio data converted in step 1, and the output is the audio data received by the server. The specific operation here is the packetization and transmission of data using a network protocol (e.g., TCP / IP).

[0422] Step 3:

[0423] The server converts received digital audio data into text using a speech recognition engine. The input is digital audio data, and the output is the text data of the audio. Specifically, it uses the speech_recognition library to generate text from audio.

[0424] Step 4:

[0425] The server analyzes the text data and identifies the speaker. The input is the text data from step 3, and the output is the speaker identification result. Specific operations include matching the data against a user profile.

[0426] Step 5:

[0427] The server analyzes the user's emotional state using speaker identification results and an emotion engine. Input is speech-to-text data and identified speaker information, while output is the emotion recognition result. Specific operations include a process that utilizes the EmotionEngine to infer emotions from indicators such as tone of voice and word choice.

[0428] Step 6:

[0429] The server is configured to automatically restrict the response content if the speaker is not a user. The input is the speaker identification result, and the output is the restricted response content. The specific operation is performed by filtering the response content using PrivacyGuard technology.

[0430] Step 7:

[0431] The server adjusts its response based on the user's emotional state to generate appropriate feedback. The input is the emotion recognition result and restricted response content, while the output is the adjusted final response. Specific operations include generation processes that modify the tone and level of detail of the response.

[0432] Step 8:

[0433] The server sends the final response to the terminal and presents it to the user. The input is the generated response data from step 7, and the output is what the terminal displays to the user or provides as audio feedback. The specific actions involve network transmission and output processing at the terminal.

[0434] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0435] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0436] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0437] [Third Embodiment]

[0438] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0439] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0440] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0441] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0442] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0443] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0444] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0445] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0446] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0447] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0448] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0449] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0450] This invention is a privacy protection system using speech recognition technology, and is particularly aimed at protecting user privacy when an AI agent participates in a conversation involving multiple people. This system consists of multiple modules operating on a server and a terminal. Specific embodiments are described below.

[0451] System Configuration

[0452] The server is equipped with a speech recognition engine and is responsible for receiving and processing voice data transmitted from the terminal. The terminal has the means to capture the user's voice data and communicate with the server.

[0453] How to use

[0454] When a user initiates a voice interaction with an AI agent, the device uses its microphone to capture the user's speech as audio data. This audio data is transmitted to a server in real time. The server identifies the speaker based on the audio data, and if a person other than the user is detected, it returns that information to the device.

[0455] After receiving information from the server, the device notifies the user that another person is present in the conversation. At this point, the device enables restricted mode for privacy protection and continues the conversation after obtaining the user's confirmation.

[0456] When restricted mode is enabled, the server applies a specific response generation algorithm to queries that may contain private information to provide secure content. For example, based on user instructions such as "Do not share my personal information," the server generalizes the response.

[0457] Specific example

[0458] As an example, imagine a user asking an AI agent, "Tell me my schedule for this week," in an environment where a friend is nearby. In this system, as soon as the server detects the friend's voice, a notification is sent to the device. The device informs the user, "There is another person in the conversation. Privacy mode is enabled." As a result, the server's response will be privacy-conscious, such as, "You can check the details of the schedule individually later."

[0459] In this way, users can use the AI ​​agent with peace of mind and prevent the leakage of personal information. This system provides a means to protect user privacy at a high level, especially when used in public places or workplaces.

[0460] The following describes the processing flow.

[0461] Step 1:

[0462] The device captures the user's spoken voice using its microphone and sends it to the server as digital audio data.

[0463] Step 2:

[0464] The server analyzes the received audio data using a speech recognition engine and compares it with a voice profile that characterizes the speaker to determine whether it is the user's voice.

[0465] Step 3:

[0466] The server determines, based on the identification results, whether there are speakers other than the user in the conversation. It then sends this result to the terminal.

[0467] Step 4:

[0468] The device receives the determination result from the server, and if another person is included in the conversation, it displays or voices a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled."

[0469] Step 5:

[0470] When privacy mode is enabled, the server generates appropriate responses to incoming user queries in a way that does not include private information. For example, responses containing the user's personal information are replaced with generalized language or secure content.

[0471] Step 6:

[0472] The terminal provides the user with the response received from the server. In this process, the response is presented in a format that protects personal information through privacy mode.

[0473] Step 7:

[0474] Through these processes, users can continue their conversation with the AI ​​agent in a safe and private manner.

[0475] (Example 1)

[0476] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0477] In conversational systems using speech recognition technology, there is a need to effectively detect the presence of others in real time and generate appropriate responses while protecting user privacy at a high level. In particular, it is necessary to minimize the risk of unintentional leakage of personal information when used in public places or workplaces.

[0478] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0479] In this invention, the server includes means for transmitting voice information to an information processing device, means for authenticating the speaker using a voice recognition function, and means for setting a privacy level. This makes it possible to automatically activate privacy mode when another person is involved in the conversation and respond to the user while protecting personal information.

[0480] "Audio information" refers to information obtained by converting speech acquired through input devices such as microphones into digital data.

[0481] The term "speaker" refers to the person making a statement based on the acquired audio information, and is identified based on their voice pattern.

[0482] "Users" refer to individuals who directly operate or use this system, and are recipients of information that the system particularly needs to protect.

[0483] An "information processing device" is a device that performs various analyses and processes using acquired audio information, and includes servers and computer devices.

[0484] "Speech recognition functionality" is a technology that analyzes voice input, converts it into text information, and identifies the speaker, and is a core technology used in this system.

[0485] "Privacy level" refers to the degree of security set to control the content of responses for the purpose of protecting personal information.

[0486] A "response generation algorithm" refers to a method or computational means for generating information presented to the user based on voice information, and is used to produce highly secure responses.

[0487] This invention is a privacy protection system utilizing speech recognition technology, and is particularly aimed at ensuring privacy when an AI agent assists a user in conversation. The following details an embodiment of the system.

[0488] server

[0489] The server is equipped with a high-performance speech recognition engine and receives and processes voice information transmitted from the terminal in real time. The necessary environment for this is a general cloud computing platform and a powerful voice analysis engine. Specifically, a general-purpose voice processing API can be used. The server analyzes the voice information and identifies whether the speaker is someone other than the user. It also generates a secure response that takes privacy levels into consideration using an AI model. The generated response is generalized and may include something like, "You can check the details of the schedule individually later."

[0490] terminal

[0491] The device captures the user's speech using a highly sensitive microphone. This device is often a portable device or smartphone equipped with noise-canceling capabilities. The device sends the acquired audio information to a server, and a sophisticated security protocol is used for data transfer. When the server detects another person, the device uses that information to display a message to the user such as, "There is another person in the conversation. Privacy mode is enabled."

[0492] User

[0493] Users obtain necessary information by conversing with an AI agent via voice, with the server and terminal assisting in protecting privacy during this process. A voice confirmation feature ensures that personal information can be handled safely even in public places. This feature alerts the user if other people are involved in the conversation. An example of such a prompt might be the user saying, "Don't share my personal information."

[0494] Specific example

[0495] For example, if a user asks an AI assistant in a cafe, "Tell me about tomorrow's meeting," the server filters out ambient noise while detecting other people's voices. After detection, the device activates privacy mode and provides a generalized response to the user's query. In this way, an environment is created where users can use AI with peace of mind.

[0496] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0497] Step 1:

[0498] The device captures the user's speech using a high-sensitivity microphone. The input is the user's voice, and the output is digital audio data. During this process, noise cancellation technology is used to reduce external noise and obtain clear audio data.

[0499] Step 2:

[0500] The terminal encrypts the captured digital audio data and sends it to the server. The input is the audio data acquired by the terminal, and the output is the securely transmitted audio data. This process applies an encryption protocol (e.g., TLS) to ensure security.

[0501] Step 3:

[0502] The server analyzes the received audio data using a speech recognition engine to identify the speaker. The input is encrypted audio data, and the output is speaker identification information. The server analyzes the audio data and identifies the speaker by comparing it to known speech patterns.

[0503] Step 4:

[0504] If the server determines from the analysis results that the speaker is someone other than the user, it sends that information to the terminal. The input is the result of speaker identification, and the output is a notification of detection of another person. The server returns the determination result to the terminal in real time.

[0505] Step 5:

[0506] When the device receives information about the detection of another person, it displays a message to the user stating, "There is another person in the conversation. Privacy mode is enabled." The input is notification information from the server, and the output is a warning message to the user. This allows the user to recognize the presence of another person and take necessary action.

[0507] Step 6:

[0508] When a user selects privacy mode, the device sends that selection to the server. The input is the user's selection information, and the output is the selection notification sent to the server. This selection automatically triggers privacy protection measures.

[0509] Step 7:

[0510] When the server confirms privacy mode, it uses a generative AI model to generate a secure response. The input is the user's query information and privacy level, and the output is a generalized secure response. The generative AI model provides information that takes privacy into consideration.

[0511] Step 8:

[0512] The terminal displays the response sent from the server to the user. The input is the response data from the server, and the output is the presentation of information to the user. The response is displayed as audio or text, allowing the user to obtain information with confidence.

[0513] (Application Example 1)

[0514] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0515] In smart cities, using voice assistants in public places poses a risk of privacy breaches due to others overhearing conversations. Furthermore, in environments with multiple speakers, responses that protect the privacy of specific users are required. Conventional technologies have been unable to adequately address these issues.

[0516] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0517] In this invention, the server includes a device for acquiring voice information, a device for identifying the speaker from the acquired voice information, and a device for applying a response generation algorithm to provide a secure response when different speakers are present. This makes it possible to securely protect the user's privacy information even when using a voice assistant in a public place.

[0518] "Audio information" refers to a data format in which audio is digitized, and is used for speech recognition and processing.

[0519] "Device" refers to a combination of hardware and software designed to perform a specific function, and in this invention, it refers to a system for acquiring and processing sound.

[0520] "Speaker" refers to the person who produces auditory information, and in speech recognition, it is the target for identifying an individual who possesses a specific speech pattern.

[0521] A "user" is the person who uses this system, and usually refers to the individual who gives instructions to the voice assistant.

[0522] "Response information" refers to the information that a voice assistant provides to the user, including answers to questions and guidance.

[0523] A "response generation algorithm" refers to a processing method for generating appropriate response information based on user input and circumstances, and can apply special conditions to provide a secure response.

[0524] "Privacy Mode" is a function that adjusts the system's response to protect the user's personal information and is activated when another person's speech is detected.

[0525] To implement this invention, a device and a server are required to acquire and analyze voice information. A smartphone or smart glasses with a built-in microphone can be used as the voice acquisition device. This device is responsible for acquiring voice information from the user and transmitting it to the server.

[0526] The server receives audio information and uses speech recognition technology to identify the speaker. Specifically, it uses a speech recognition engine such as the Google Cloud Speech-to-Text API to convert the audio information into text in real time and analyze the speaker's characteristics. At this time, the server uses a generative AI model to detect whether the speaker is a user other than a specific user.

[0527] If different speakers are present, the server applies a response generation algorithm to generate secure response information. This process uses AI models and predefined contexts to create a privacy-conscious, generic response. The generated response information is sent to the user's device and presented to the user as audio or text.

[0528] As a concrete example, consider a scenario where a user on public transport asks a voice assistant, "What are the plans for tomorrow?" In this case, the server analyzes the surrounding audio and, if it detects other speakers, activates privacy mode and returns a response such as, "More details later."

[0529] An example of a prompt would be: "Please provide a privacy-sensitive response for the voice assistant in the following scenario: The user asks, 'What is my health information?' There are third parties present."

[0530] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0531] Step 1:

[0532] The device acquires the user's speech using a microphone. The acquired voice information is temporarily stored on the device as digital data. Because this data retains the characteristics of the voice, it can be identified in the next processing step.

[0533] Step 2:

[0534] The device transmits the acquired voice information to the server. Upon transmission of this data, the server begins analyzing the data using speech recognition technology. An internet connection is required for data transmission.

[0535] Step 3:

[0536] The server receives the audio information and uses a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the audio information into text data. The converted text data contains the speaker's voice characteristics and is used as data to identify the speaker.

[0537] Step 4:

[0538] The server uses a generative AI model based on the converted text data to identify the speaker. This identification process involves analyzing speech patterns to distinguish between a specific user and others. As a result, it is determined whether the speaker is a specific user or someone else.

[0539] Step 5:

[0540] If the server determines that the speaker is someone else, it applies a response generation algorithm. This algorithm is used to generate safe and generalized response information. A generative AI model is used for response generation, and context-appropriate prompt sentences are applied.

[0541] Step 6:

[0542] The server sends the generated response information to the terminal. This response information is designed with privacy in mind and may include messages such as, "Please check the details later."

[0543] Step 7:

[0544] The terminal transmits the received response information to the user. This is done using audio output or screen display to allow the user to confirm the response.

[0545] This processing flow makes it possible to use voice assistants in public places while ensuring user privacy.

[0546] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0547] This invention is a privacy protection system using speech recognition technology and an emotion engine, aiming to protect privacy with high accuracy by simultaneously recognizing the speaker and emotions based on the user's voice data. Specific embodiments are described below.

[0548] System Configuration

[0549] The server is equipped with a speech recognition engine and an emotion engine, and has the function of receiving and processing voice data transmitted from the terminal. The terminal provides a means to acquire the user's voice data and communicate with the server.

[0550] How to use

[0551] When a user begins interacting with the AI ​​agent, the device captures the audio using its microphone and sends it to the server as digital audio data. The server analyzes the received audio data with a speech recognition engine to simultaneously identify the emotions of both the speaker and the user. The emotion engine identifies the emotional state of the user based on the tone and speed of their voice.

[0552] Based on the speaker identification results, if someone other than the user is participating in the conversation, the server returns that information to the terminal. The terminal then notifies the user that "Someone else is participating in the conversation. Privacy mode is enabled." In addition, the tone of the response and the level of detail of the information are dynamically adjusted according to the recognized emotions of the user.

[0553] Specific example

[0554] For example, consider a scenario where a user asks "What are today's tasks?" in a tired voice. Assume a family member is nearby. In this system, the server detects the family member's voice and recognizes the user's emotional state as "fatigued." The device then informs the user, "Someone else is in the conversation. Privacy mode is enabled. More details will be provided later." At this point, the server adjusts its response, taking the user's state into consideration, to avoid overloading the user with excessive information.

[0555] In this way, users can use the AI ​​agent with peace of mind, and interaction that flexibly responds to the user's emotions can be achieved while protecting their privacy. This system provides a means to ensure both user privacy and comfort, especially when used in homes or public places.

[0556] The following describes the processing flow.

[0557] Step 1:

[0558] The device captures the user's speech using a microphone and sends it to the server as digital audio data.

[0559] Step 2:

[0560] The server processes the received audio data with a speech recognition engine and identifies the user's speech by matching it with a speech profile that characterizes the speaker.

[0561] Step 3:

[0562] The server inputs the voice data into the emotion engine, which analyzes the tone, speed, and volume of the user's voice to recognize their emotions.

[0563] Step 4:

[0564] The server determines whether there are participants other than the user in the conversation based on the speaker identification results and emotion recognition results. This determination result is then sent to the terminal.

[0565] Step 5:

[0566] The device receives the detection result from the server and, if another person is in the conversation, notifies the user with the message, "Another person is participating in the conversation. Privacy mode is enabled."

[0567] Step 6:

[0568] The server generates responses to user queries with privacy mode enabled and taking user sentiment into consideration. The responses are selected to exclude private information and are adjusted to a tone that respects the user's feelings.

[0569] Step 7:

[0570] The device provides the user with a tailored response, which is considerate of the user's emotional state.

[0571] Step 8:

[0572] Users can continue their conversations with AI agents in a safe and satisfying manner, while their privacy is protected and flexible, emotionally-driven dialogue is possible.

[0573] (Example 2)

[0574] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0575] Conventional speech recognition systems have shortcomings, including insufficient speaker identification and subsequent privacy protection, as well as the inability to respond flexibly to the user's emotional state. Furthermore, there is a risk of information leakage when individuals other than the user participate in conversations in certain environments. There is a need to address these challenges and simultaneously improve both privacy and user experience.

[0576] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0577] In this invention, the server includes means for acquiring voice data, means for identifying the speaker from the acquired voice data, and means for analyzing the emotional state from the acquired voice data. This enables the activation of a privacy mode based on speaker identification and the adjustment of the tone of response and level of detail of information according to the user's emotions.

[0578] "Audio data" refers to data obtained by capturing the user's voice and converting it into a digital format.

[0579] "Speaker" refers to the individual or person identified as the source of the speech based on the acquired audio data.

[0580] "Privacy Mode" is a protective feature that restricts responses to prevent the leakage of potentially unnecessary information when it is determined that another person is participating in the conversation.

[0581] "Emotional state" refers to a state that indicates psychological or emotional changes or tendencies, analyzed based on the user's voice data.

[0582] "Response tone" refers to the pitch and manner of speaking characteristics of the responses that the system provides to the user, and is usually adjusted based on the emotional state.

[0583] "Information detail" is an indicator that shows the specificity and depth of information included in the response to the user, and it is adjusted according to the user's emotional state.

[0584] This invention is a system that utilizes speech recognition technology and an emotion analysis engine to enable flexible, emotion-responsive dialogue while protecting user privacy. Specific embodiments are described below.

[0585] The server is equipped with a high-performance speech recognition engine and sentiment analysis engine. The speech recognition engine uses commercially available speech analysis software (e.g., common APIs) to convert speech data into text. This process also includes an authentication process to identify who the speaker is. The sentiment analysis engine analyzes parameters such as tone, speed, and volume obtained from the speech data to identify the user's emotional state.

[0586] The terminal is responsible for acquiring the user's voice using a microphone and transmitting it to the server as digital audio data. Since the terminal transmits this audio data to the server in real time, the communication infrastructure requires high reliability.

[0587] For example, if a user asks "What are today's tasks?" in a tired voice, the server analyzes the audio data to identify the user as the speaker and simultaneously analyzes their emotional state as "fatigue." Furthermore, if the presence of other people, such as family members, is detected in the vicinity, the server activates privacy mode based on the speaker identification information and notifies the device accordingly.

[0588] Based on information from the server, the device displays a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled." The server also takes the user's emotional state into account and adjusts the tone of its response and the level of detail in the information provided to avoid unnecessarily burdening the user.

[0589] A concrete example of a prompt in this system would be, "When a user is feeling tired and someone is nearby, how should they respond to ask an AI agent about a task while maintaining their privacy and comfort?"

[0590] In this way, this invention makes it possible to maximize the user experience while maintaining privacy in homes and public places.

[0591] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0592] Step 1:

[0593] The device captures audio using a microphone when the user begins interacting with the AI ​​agent. The input data is the user's voice signal, and the device's audio capture software converts this audio into a digital signal. Specifically, it samples the audio signal and formats it as a digital audio file.

[0594] Step 2:

[0595] The terminal sends captured digital audio data to the server. The input is the converted digital audio data, and the output is a transmission completion report to the server. The terminal streams the audio data using a communication protocol (e.g., HTTP or WebSocket). Specifically, it divides the audio data into packets for transmission and uses compression techniques to minimize latency.

[0596] Step 3:

[0597] The server passes the received audio data to the speech recognition engine. The input is digital audio data from the terminal, and the output is the content converted into text. The server uses the speech recognition engine to generate text from the audio waveform and identify the speaker. Specifically, it performs a speech feature extraction process and model comparison.

[0598] Step 4:

[0599] The server simultaneously uses an emotion analysis engine to identify emotional states from audio data. The input is the same audio data, and the output is the analyzed emotional state tag. The emotion analysis engine analyzes the tone, pitch, and speed of the audio to label emotions. Specifically, it statistically analyzes speech metrics related to emotions and maps them to emotion categories.

[0600] Step 5:

[0601] The server determines the privacy mode based on speaker recognition and emotional state. If the speaker is not the user or if the emotional state meets certain conditions, it returns output to the terminal that enhances privacy. Specifically, it performs a cross-check of speaker information and sets a privacy flag.

[0602] Step 6:

[0603] The terminal notifies the user based on information received from the server. Input is the privacy mode status and response content, while output is the notification message to the user. The terminal uses a display or speech synthesis system to notify the user, "Another person is participating in the conversation. Privacy mode is enabled." Specifically, it performs the notification process via display or voice.

[0604] Step 7:

[0605] The server adjusts the tone and level of detail of its responses to the user according to the recognized emotional state. The input is the result of the emotional analysis, and the output is the configured response parameters. The server considers the user's emotional state to adopt an appropriate tone in its response and controls the level of detail to prevent excessive information transmission. Specifically, the response generation module performs tone adjustment and information calculation.

[0606] (Application Example 2)

[0607] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0608] Modern voice assistant technology struggles to adequately protect privacy while providing appropriate responses tailored to the user's emotional state. This presents risks of eavesdropping and overload when the user is stressed. The challenge lies in simultaneously improving both privacy and user experience.

[0609] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0610] In this invention, the server includes a device for acquiring voice data, a device for identifying the speaker from the acquired voice data, a device for restricting the response content if the identified speaker is determined to be someone other than a specific user, a device for presenting the response content to the user, a device for analyzing the user's emotional state, and a device for adjusting the response according to the user's emotional state. This makes it possible to provide responses that match the user's emotional state while protecting the user's privacy.

[0611] "Voice data" refers to digital recordings of users' speech, which are used for purposes such as speech recognition and sentiment analysis.

[0612] A "device" is a component of hardware or software designed to perform a specific function, such as acquiring, analyzing, or responding to audio data.

[0613] The term "speaker" refers to the person who made the utterance in the audio data, and is the subject that needs to be identified and recognized.

[0614] "Discrimination" is the process of analyzing acquired audio data to determine whether or not it belongs to a specific user.

[0615] A "device for restricting response content" is a device that has a mechanism to adjust the scope of information provided in order to protect user privacy, based on the results of identifying the speaker.

[0616] "Users" are those who use the voice recognition system, and their privacy and the quality of their experience should be taken into consideration.

[0617] "Emotional state" refers to the psychological state that indicates emotions contained in the user's utterances, such as joy, sadness, or stress.

[0618] "Analysis" is the process of evaluating the user's emotional state based on voice data and reflecting the results in the response.

[0619] "Response adjustment" is the process of changing the content and tone of a response according to the user's emotional state.

[0620] To implement this invention, the terminal must first be equipped with a device for acquiring audio data. The terminal captures the user's speech with a microphone and transmits it to the server as digital audio data. The server uses speech recognition technology to identify the speaker from this audio data. Specifically, it uses a library such as speech_recognition to convert the audio data into text and then analyzes that text data to identify the speaker.

[0621] If the server determines that the speaker is someone other than the user, it will restrict the content of the response to protect privacy. For example, it will use technologies such as PrivacyGuard to automatically adjust the details of the response so that important information is not leaked even if it is overheard by others.

[0622] In addition, the server utilizes emotion engines such as EmotionEngine to analyze the user's emotional state and dynamically adjusts its response based on the results. For example, if the user is experiencing stress, the server will simplify data submission and respond in a gentle tone. This reduces the burden on the user.

[0623] To give a concrete example, when a user asks a home robot, "What are my plans for today?", the server detects the voices of family members from the acquired audio data and also recognizes the user's "calmness" from their voice. In this case, the server can generate a response such as, "There are other people nearby right now, but please listen without worry. I have a meeting in the morning and free time in the afternoon."

[0624] An example of a prompt for a generative AI model is, "What kind of system can achieve emotional and privacy protection from user voice data in a home robot?" Using this information, users can receive appropriate feedback in a privacy-protected manner.

[0625] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0626] Step 1:

[0627] The device captures the user's voice in real time using a microphone. The acquired voice data is converted into a digital format. At this stage, the input is analog voice data, and the output is digital voice data. The specific operation of the data conversion is to use analog-to-digital conversion technology.

[0628] Step 2:

[0629] The terminal sends digital audio data to the server. The input is the digital audio data converted in step 1, and the output is the audio data received by the server. The specific operation here is the packetization and transmission of data using a network protocol (e.g., TCP / IP).

[0630] Step 3:

[0631] The server converts received digital audio data into text using a speech recognition engine. The input is digital audio data, and the output is the text data of the audio. Specifically, it uses the speech_recognition library to generate text from audio.

[0632] Step 4:

[0633] The server analyzes the text data and identifies the speaker. The input is the text data from step 3, and the output is the speaker identification result. Specific operations include matching the data against a user profile.

[0634] Step 5:

[0635] The server analyzes the user's emotional state using speaker identification results and an emotion engine. Input is speech-to-text data and identified speaker information, while output is the emotion recognition result. Specific operations include a process that utilizes the EmotionEngine to infer emotions from indicators such as tone of voice and word choice.

[0636] Step 6:

[0637] The server is configured to automatically restrict the response content if the speaker is not a user. The input is the speaker identification result, and the output is the restricted response content. The specific operation is performed by filtering the response content using PrivacyGuard technology.

[0638] Step 7:

[0639] The server adjusts its response based on the user's emotional state to generate appropriate feedback. The input is the emotion recognition result and restricted response content, while the output is the adjusted final response. Specific operations include generation processes that modify the tone and level of detail of the response.

[0640] Step 8:

[0641] The server sends the final response to the terminal and presents it to the user. The input is the generated response data from step 7, and the output is what the terminal displays to the user or provides as audio feedback. The specific actions involve network transmission and output processing at the terminal.

[0642] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0643] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0644] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0645] [Fourth Embodiment]

[0646] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0647] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0648] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0649] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0650] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0651] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0652] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0653] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0654] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0655] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0656] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0657] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0658] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0659] This invention is a privacy protection system using speech recognition technology, and is particularly aimed at protecting user privacy when an AI agent participates in a conversation involving multiple people. This system consists of multiple modules operating on a server and a terminal. Specific embodiments are described below.

[0660] System Configuration

[0661] The server is equipped with a speech recognition engine and is responsible for receiving and processing voice data transmitted from the terminal. The terminal has the means to capture the user's voice data and communicate with the server.

[0662] How to use

[0663] When a user initiates a voice interaction with an AI agent, the device uses its microphone to capture the user's speech as audio data. This audio data is transmitted to a server in real time. The server identifies the speaker based on the audio data, and if a person other than the user is detected, it returns that information to the device.

[0664] After receiving information from the server, the device notifies the user that another person is present in the conversation. At this point, the device enables restricted mode for privacy protection and continues the conversation after obtaining the user's confirmation.

[0665] When restricted mode is enabled, the server applies a specific response generation algorithm to queries that may contain private information to provide secure content. For example, based on user instructions such as "Do not share my personal information," the server generalizes the response.

[0666] Specific example

[0667] As an example, imagine a user asking an AI agent, "Tell me my schedule for this week," in an environment where a friend is nearby. In this system, as soon as the server detects the friend's voice, a notification is sent to the device. The device informs the user, "There is another person in the conversation. Privacy mode is enabled." As a result, the server's response will be privacy-conscious, such as, "You can check the details of the schedule individually later."

[0668] In this way, users can use the AI ​​agent with peace of mind and prevent the leakage of personal information. This system provides a means to protect user privacy at a high level, especially when used in public places or workplaces.

[0669] The following describes the processing flow.

[0670] Step 1:

[0671] The device captures the user's spoken voice using its microphone and sends it to the server as digital audio data.

[0672] Step 2:

[0673] The server analyzes the received audio data using a speech recognition engine and compares it with a voice profile that characterizes the speaker to determine whether it is the user's voice.

[0674] Step 3:

[0675] The server determines, based on the identification results, whether there are speakers other than the user in the conversation. It then sends this result to the terminal.

[0676] Step 4:

[0677] The device receives the determination result from the server, and if another person is included in the conversation, it displays or voices a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled."

[0678] Step 5:

[0679] When privacy mode is enabled, the server generates appropriate responses to incoming user queries in a way that does not include private information. For example, responses containing the user's personal information are replaced with generalized language or secure content.

[0680] Step 6:

[0681] The terminal provides the user with the response received from the server. In this process, the response is presented in a format that protects personal information through privacy mode.

[0682] Step 7:

[0683] Through these processes, users can continue their conversation with the AI ​​agent in a safe and private manner.

[0684] (Example 1)

[0685] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0686] In conversational systems using speech recognition technology, there is a need to effectively detect the presence of others in real time and generate appropriate responses while protecting user privacy at a high level. In particular, it is necessary to minimize the risk of unintentional leakage of personal information when used in public places or workplaces.

[0687] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0688] In this invention, the server includes means for transmitting voice information to an information processing device, means for authenticating the speaker using a voice recognition function, and means for setting a privacy level. This makes it possible to automatically activate privacy mode when another person is involved in the conversation and respond to the user while protecting personal information.

[0689] "Audio information" refers to information obtained by converting speech acquired through input devices such as microphones into digital data.

[0690] The term "speaker" refers to the person making a statement based on the acquired audio information, and is identified based on their voice pattern.

[0691] "Users" refer to individuals who directly operate or use this system, and are recipients of information that the system particularly needs to protect.

[0692] An "information processing device" is a device that performs various analyses and processes using acquired audio information, and includes servers and computer devices.

[0693] "Speech recognition functionality" is a technology that analyzes voice input, converts it into text information, and identifies the speaker, and is a core technology used in this system.

[0694] "Privacy level" refers to the degree of security set to control the content of responses for the purpose of protecting personal information.

[0695] A "response generation algorithm" refers to a method or computational means for generating information presented to the user based on voice information, and is used to produce highly secure responses.

[0696] This invention is a privacy protection system utilizing speech recognition technology, and is particularly aimed at ensuring privacy when an AI agent assists a user in conversation. The following details an embodiment of the system.

[0697] server

[0698] The server is equipped with a high-performance speech recognition engine and receives and processes voice information transmitted from the terminal in real time. The necessary environment for this is a general cloud computing platform and a powerful voice analysis engine. Specifically, a general-purpose voice processing API can be used. The server analyzes the voice information and identifies whether the speaker is someone other than the user. It also generates a secure response that takes privacy levels into consideration using an AI model. The generated response is generalized and may include something like, "You can check the details of the schedule individually later."

[0699] terminal

[0700] The device captures the user's speech using a highly sensitive microphone. This device is often a portable device or smartphone equipped with noise-canceling capabilities. The device sends the acquired audio information to a server, and a sophisticated security protocol is used for data transfer. When the server detects another person, the device uses that information to display a message to the user such as, "There is another person in the conversation. Privacy mode is enabled."

[0701] User

[0702] Users obtain necessary information by conversing with an AI agent via voice, with the server and terminal assisting in protecting privacy during this process. A voice confirmation feature ensures that personal information can be handled safely even in public places. This feature alerts the user if other people are involved in the conversation. An example of such a prompt might be the user saying, "Don't share my personal information."

[0703] Specific example

[0704] For example, if a user asks an AI assistant in a cafe, "Tell me about tomorrow's meeting," the server filters out ambient noise while detecting other people's voices. After detection, the device activates privacy mode and provides a generalized response to the user's query. In this way, an environment is created where users can use AI with peace of mind.

[0705] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0706] Step 1:

[0707] The device captures the user's speech using a high-sensitivity microphone. The input is the user's voice, and the output is digital audio data. During this process, noise cancellation technology is used to reduce external noise and obtain clear audio data.

[0708] Step 2:

[0709] The terminal encrypts the captured digital audio data and sends it to the server. The input is the audio data acquired by the terminal, and the output is the securely transmitted audio data. This process applies an encryption protocol (e.g., TLS) to ensure security.

[0710] Step 3:

[0711] The server analyzes the received audio data using a speech recognition engine to identify the speaker. The input is encrypted audio data, and the output is speaker identification information. The server analyzes the audio data and identifies the speaker by comparing it to known speech patterns.

[0712] Step 4:

[0713] If the server determines from the analysis results that the speaker is someone other than the user, it sends that information to the terminal. The input is the result of speaker identification, and the output is a notification of detection of another person. The server returns the determination result to the terminal in real time.

[0714] Step 5:

[0715] When the device receives information about the detection of another person, it displays a message to the user stating, "There is another person in the conversation. Privacy mode is enabled." The input is notification information from the server, and the output is a warning message to the user. This allows the user to recognize the presence of another person and take necessary action.

[0716] Step 6:

[0717] When a user selects privacy mode, the device sends that selection to the server. The input is the user's selection information, and the output is the selection notification sent to the server. This selection automatically triggers privacy protection measures.

[0718] Step 7:

[0719] When the server confirms privacy mode, it uses a generative AI model to generate a secure response. The input is the user's query information and privacy level, and the output is a generalized secure response. The generative AI model provides information that takes privacy into consideration.

[0720] Step 8:

[0721] The terminal displays the response sent from the server to the user. The input is the response data from the server, and the output is the presentation of information to the user. The response is displayed as audio or text, allowing the user to obtain information with confidence.

[0722] (Application Example 1)

[0723] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0724] In smart cities, using voice assistants in public places poses a risk of privacy breaches due to others overhearing conversations. Furthermore, in environments with multiple speakers, responses that protect the privacy of specific users are required. Conventional technologies have been unable to adequately address these issues.

[0725] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0726] In this invention, the server includes a device for acquiring voice information, a device for identifying the speaker from the acquired voice information, and a device for applying a response generation algorithm to provide a secure response when different speakers are present. This makes it possible to securely protect the user's privacy information even when using a voice assistant in a public place.

[0727] "Audio information" refers to a data format in which audio is digitized, and is used for speech recognition and processing.

[0728] "Device" refers to a combination of hardware and software designed to perform a specific function, and in this invention, it refers to a system for acquiring and processing sound.

[0729] "Speaker" refers to the person who produces auditory information, and in speech recognition, it is the target for identifying an individual who possesses a specific speech pattern.

[0730] A "user" is the person who uses this system, and usually refers to the individual who gives instructions to the voice assistant.

[0731] "Response information" refers to the information that a voice assistant provides to the user, including answers to questions and guidance.

[0732] A "response generation algorithm" refers to a processing method for generating appropriate response information based on user input and circumstances, and can apply special conditions to provide a secure response.

[0733] "Privacy Mode" is a function that adjusts the system's response to protect the user's personal information and is activated when another person's speech is detected.

[0734] To implement this invention, a device and a server are required to acquire and analyze voice information. A smartphone or smart glasses with a built-in microphone can be used as the voice acquisition device. This device is responsible for acquiring voice information from the user and transmitting it to the server.

[0735] The server receives audio information and uses speech recognition technology to identify the speaker. Specifically, it uses a speech recognition engine such as the Google Cloud Speech-to-Text API to convert the audio information into text in real time and analyze the speaker's characteristics. At this time, the server uses a generative AI model to detect whether the speaker is a user other than a specific user.

[0736] If different speakers are present, the server applies a response generation algorithm to generate secure response information. This process uses AI models and predefined contexts to create a privacy-conscious, generic response. The generated response information is sent to the user's device and presented to the user as audio or text.

[0737] As a concrete example, consider a scenario where a user on public transport asks a voice assistant, "What are the plans for tomorrow?" In this case, the server analyzes the surrounding audio and, if it detects other speakers, activates privacy mode and returns a response such as, "More details later."

[0738] An example of a prompt would be: "Please provide a privacy-sensitive response for the voice assistant in the following scenario: The user asks, 'What is my health information?' There are third parties present."

[0739] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0740] Step 1:

[0741] The device acquires the user's speech using a microphone. The acquired voice information is temporarily stored on the device as digital data. Because this data retains the characteristics of the voice, it can be identified in the next processing step.

[0742] Step 2:

[0743] The device transmits the acquired voice information to the server. Upon transmission of this data, the server begins analyzing the data using speech recognition technology. An internet connection is required for data transmission.

[0744] Step 3:

[0745] The server receives the audio information and uses a speech recognition engine (e.g., Google Cloud Speech-to-Text API) to convert the audio information into text data. The converted text data contains the speaker's voice characteristics and is used as data to identify the speaker.

[0746] Step 4:

[0747] The server uses a generative AI model based on the converted text data to identify the speaker. This identification process involves analyzing speech patterns to distinguish between a specific user and others. As a result, it is determined whether the speaker is a specific user or someone else.

[0748] Step 5:

[0749] If the server determines that the speaker is someone else, it applies a response generation algorithm. This algorithm is used to generate safe and generalized response information. A generative AI model is used for response generation, and context-appropriate prompt sentences are applied.

[0750] Step 6:

[0751] The server sends the generated response information to the terminal. This response information is designed with privacy in mind and may include messages such as, "Please check the details later."

[0752] Step 7:

[0753] The terminal transmits the received response information to the user. This is done using audio output or screen display to allow the user to confirm the response.

[0754] This processing flow makes it possible to use voice assistants in public places while ensuring user privacy.

[0755] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0756] This invention is a privacy protection system using speech recognition technology and an emotion engine, aiming to protect privacy with high accuracy by simultaneously recognizing the speaker and emotions based on the user's voice data. Specific embodiments are described below.

[0757] System Configuration

[0758] The server is equipped with a speech recognition engine and an emotion engine, and has the function of receiving and processing voice data transmitted from the terminal. The terminal provides a means to acquire the user's voice data and communicate with the server.

[0759] How to use

[0760] When a user begins interacting with the AI ​​agent, the device captures the audio using its microphone and sends it to the server as digital audio data. The server analyzes the received audio data with a speech recognition engine to simultaneously identify the emotions of both the speaker and the user. The emotion engine identifies the emotional state of the user based on the tone and speed of their voice.

[0761] Based on the speaker identification results, if someone other than the user is participating in the conversation, the server returns that information to the terminal. The terminal then notifies the user that "Someone else is participating in the conversation. Privacy mode is enabled." In addition, the tone of the response and the level of detail of the information are dynamically adjusted according to the recognized emotions of the user.

[0762] Specific example

[0763] For example, consider a scenario where a user asks "What are today's tasks?" in a tired voice. Assume a family member is nearby. In this system, the server detects the family member's voice and recognizes the user's emotional state as "fatigued." The device then informs the user, "Someone else is in the conversation. Privacy mode is enabled. More details will be provided later." At this point, the server adjusts its response, taking the user's state into consideration, to avoid overloading the user with excessive information.

[0764] In this way, users can use the AI ​​agent with peace of mind, and interaction that flexibly responds to the user's emotions can be achieved while protecting their privacy. This system provides a means to ensure both user privacy and comfort, especially when used in homes or public places.

[0765] The following describes the processing flow.

[0766] Step 1:

[0767] The device captures the user's speech using a microphone and sends it to the server as digital audio data.

[0768] Step 2:

[0769] The server processes the received audio data with a speech recognition engine and identifies the user's speech by matching it with a speech profile that characterizes the speaker.

[0770] Step 3:

[0771] The server inputs the voice data into the emotion engine, which analyzes the tone, speed, and volume of the user's voice to recognize their emotions.

[0772] Step 4:

[0773] The server determines whether there are participants other than the user in the conversation based on the speaker identification results and emotion recognition results. This determination result is then sent to the terminal.

[0774] Step 5:

[0775] The device receives the detection result from the server and, if another person is in the conversation, notifies the user with the message, "Another person is participating in the conversation. Privacy mode is enabled."

[0776] Step 6:

[0777] The server generates responses to user queries with privacy mode enabled and taking user sentiment into consideration. The responses are selected to exclude private information and are adjusted to a tone that respects the user's feelings.

[0778] Step 7:

[0779] The device provides the user with a tailored response, which is considerate of the user's emotional state.

[0780] Step 8:

[0781] Users can continue their conversations with AI agents in a safe and satisfying manner, while their privacy is protected and flexible, emotionally-driven dialogue is possible.

[0782] (Example 2)

[0783] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0784] Conventional speech recognition systems have shortcomings, including insufficient speaker identification and subsequent privacy protection, as well as the inability to respond flexibly to the user's emotional state. Furthermore, there is a risk of information leakage when individuals other than the user participate in conversations in certain environments. There is a need to address these challenges and simultaneously improve both privacy and user experience.

[0785] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0786] In this invention, the server includes means for acquiring voice data, means for identifying the speaker from the acquired voice data, and means for analyzing the emotional state from the acquired voice data. This enables the activation of a privacy mode based on speaker identification and the adjustment of the tone of response and level of detail of information according to the user's emotions.

[0787] "Audio data" refers to data obtained by capturing the user's voice and converting it into a digital format.

[0788] "Speaker" refers to the individual or person identified as the source of the speech based on the acquired audio data.

[0789] "Privacy Mode" is a protective feature that restricts responses to prevent the leakage of potentially unnecessary information when it is determined that another person is participating in the conversation.

[0790] "Emotional state" refers to a state that indicates psychological or emotional changes or tendencies, analyzed based on the user's voice data.

[0791] "Response tone" refers to the pitch and manner of speaking characteristics of the responses that the system provides to the user, and is usually adjusted based on the emotional state.

[0792] "Information detail" is an indicator that shows the specificity and depth of information included in the response to the user, and it is adjusted according to the user's emotional state.

[0793] This invention is a system that utilizes speech recognition technology and an emotion analysis engine to enable flexible, emotion-responsive dialogue while protecting user privacy. Specific embodiments are described below.

[0794] The server is equipped with a high-performance speech recognition engine and sentiment analysis engine. The speech recognition engine uses commercially available speech analysis software (e.g., common APIs) to convert speech data into text. This process also includes an authentication process to identify who the speaker is. The sentiment analysis engine analyzes parameters such as tone, speed, and volume obtained from the speech data to identify the user's emotional state.

[0795] The terminal is responsible for acquiring the user's voice using a microphone and transmitting it to the server as digital audio data. Since the terminal transmits this audio data to the server in real time, the communication infrastructure requires high reliability.

[0796] For example, if a user asks "What are today's tasks?" in a tired voice, the server analyzes the audio data to identify the user as the speaker and simultaneously analyzes their emotional state as "fatigue." Furthermore, if the presence of other people, such as family members, is detected in the vicinity, the server activates privacy mode based on the speaker identification information and notifies the device accordingly.

[0797] Based on information from the server, the device displays a notification to the user stating, "Another person is participating in the conversation. Privacy mode is enabled." The server also takes the user's emotional state into account and adjusts the tone of its response and the level of detail in the information provided to avoid unnecessarily burdening the user.

[0798] A concrete example of a prompt in this system would be, "When a user is feeling tired and someone is nearby, how should they respond to ask an AI agent about a task while maintaining their privacy and comfort?"

[0799] In this way, this invention makes it possible to maximize the user experience while maintaining privacy in homes and public places.

[0800] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0801] Step 1:

[0802] The device captures audio using a microphone when the user begins interacting with the AI ​​agent. The input data is the user's voice signal, and the device's audio capture software converts this audio into a digital signal. Specifically, it samples the audio signal and formats it as a digital audio file.

[0803] Step 2:

[0804] The terminal sends captured digital audio data to the server. The input is the converted digital audio data, and the output is a transmission completion report to the server. The terminal streams the audio data using a communication protocol (e.g., HTTP or WebSocket). Specifically, it divides the audio data into packets for transmission and uses compression techniques to minimize latency.

[0805] Step 3:

[0806] The server passes the received audio data to the speech recognition engine. The input is digital audio data from the terminal, and the output is the content converted into text. The server uses the speech recognition engine to generate text from the audio waveform and identify the speaker. Specifically, it performs a speech feature extraction process and model comparison.

[0807] Step 4:

[0808] The server simultaneously uses an emotion analysis engine to identify emotional states from audio data. The input is the same audio data, and the output is the analyzed emotional state tag. The emotion analysis engine analyzes the tone, pitch, and speed of the audio to label emotions. Specifically, it statistically analyzes speech metrics related to emotions and maps them to emotion categories.

[0809] Step 5:

[0810] The server determines the privacy mode based on speaker recognition and emotional state. If the speaker is not the user or if the emotional state meets certain conditions, it returns output to the terminal that enhances privacy. Specifically, it performs a cross-check of speaker information and sets a privacy flag.

[0811] Step 6:

[0812] The terminal notifies the user based on information received from the server. Input is the privacy mode status and response content, while output is the notification message to the user. The terminal uses a display or speech synthesis system to notify the user, "Another person is participating in the conversation. Privacy mode is enabled." Specifically, it performs the notification process via display or voice.

[0813] Step 7:

[0814] The server adjusts the tone and level of detail of its responses to the user according to the recognized emotional state. The input is the result of the emotional analysis, and the output is the configured response parameters. The server considers the user's emotional state to adopt an appropriate tone in its response and controls the level of detail to prevent excessive information transmission. Specifically, the response generation module performs tone adjustment and information calculation.

[0815] (Application Example 2)

[0816] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0817] Modern voice assistant technology struggles to adequately protect privacy while providing appropriate responses tailored to the user's emotional state. This presents risks of eavesdropping and overload when the user is stressed. The challenge lies in simultaneously improving both privacy and user experience.

[0818] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0819] In this invention, the server includes a device for acquiring voice data, a device for identifying the speaker from the acquired voice data, a device for restricting the response content if the identified speaker is determined to be someone other than a specific user, a device for presenting the response content to the user, a device for analyzing the user's emotional state, and a device for adjusting the response according to the user's emotional state. This makes it possible to provide responses that match the user's emotional state while protecting the user's privacy.

[0820] "Voice data" refers to digital recordings of users' speech, which are used for purposes such as speech recognition and sentiment analysis.

[0821] A "device" is a component of hardware or software designed to perform a specific function, such as acquiring, analyzing, or responding to audio data.

[0822] The term "speaker" refers to the person who made the utterance in the audio data, and is the subject that needs to be identified and recognized.

[0823] "Discrimination" is the process of analyzing acquired audio data to determine whether or not it belongs to a specific user.

[0824] A "device for restricting response content" is a device that has a mechanism to adjust the scope of information provided in order to protect user privacy, based on the results of identifying the speaker.

[0825] "Users" are those who use the voice recognition system, and their privacy and the quality of their experience should be taken into consideration.

[0826] "Emotional state" refers to the psychological state that indicates emotions contained in the user's utterances, such as joy, sadness, or stress.

[0827] "Analysis" is the process of evaluating the user's emotional state based on voice data and reflecting the results in the response.

[0828] "Response adjustment" is the process of changing the content and tone of a response according to the user's emotional state.

[0829] To implement this invention, the terminal must first be equipped with a device for acquiring audio data. The terminal captures the user's speech with a microphone and transmits it to the server as digital audio data. The server uses speech recognition technology to identify the speaker from this audio data. Specifically, it uses a library such as speech_recognition to convert the audio data into text and then analyzes that text data to identify the speaker.

[0830] If the server determines that the speaker is someone other than the user, it will restrict the content of the response to protect privacy. For example, it will use technologies such as PrivacyGuard to automatically adjust the details of the response so that important information is not leaked even if it is overheard by others.

[0831] In addition, the server utilizes emotion engines such as EmotionEngine to analyze the user's emotional state and dynamically adjusts its response based on the results. For example, if the user is experiencing stress, the server will simplify data submission and respond in a gentle tone. This reduces the burden on the user.

[0832] To give a concrete example, when a user asks a home robot, "What are my plans for today?", the server detects the voices of family members from the acquired audio data and also recognizes the user's "calmness" from their voice. In this case, the server can generate a response such as, "There are other people nearby right now, but please listen without worry. I have a meeting in the morning and free time in the afternoon."

[0833] An example of a prompt for a generative AI model is, "What kind of system can achieve emotional and privacy protection from user voice data in a home robot?" Using this information, users can receive appropriate feedback in a privacy-protected manner.

[0834] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0835] Step 1:

[0836] The device captures the user's voice in real time using a microphone. The acquired voice data is converted into a digital format. At this stage, the input is analog voice data, and the output is digital voice data. The specific operation of the data conversion is to use analog-to-digital conversion technology.

[0837] Step 2:

[0838] The terminal sends digital audio data to the server. The input is the digital audio data converted in step 1, and the output is the audio data received by the server. The specific operation here is the packetization and transmission of data using a network protocol (e.g., TCP / IP).

[0839] Step 3:

[0840] The server converts received digital audio data into text using a speech recognition engine. The input is digital audio data, and the output is the text data of the audio. Specifically, it uses the speech_recognition library to generate text from audio.

[0841] Step 4:

[0842] The server analyzes the text data and identifies the speaker. The input is the text data from step 3, and the output is the speaker identification result. Specific operations include matching the data against a user profile.

[0843] Step 5:

[0844] The server analyzes the user's emotional state using speaker identification results and an emotion engine. Input is speech-to-text data and identified speaker information, while output is the emotion recognition result. Specific operations include a process that utilizes the EmotionEngine to infer emotions from indicators such as tone of voice and word choice.

[0845] Step 6:

[0846] The server is configured to automatically restrict the response content if the speaker is not a user. The input is the speaker identification result, and the output is the restricted response content. The specific operation is performed by filtering the response content using PrivacyGuard technology.

[0847] Step 7:

[0848] The server adjusts its response based on the user's emotional state to generate appropriate feedback. The input is the emotion recognition result and restricted response content, while the output is the adjusted final response. Specific operations include generation processes that modify the tone and level of detail of the response.

[0849] Step 8:

[0850] The server sends the final response to the terminal and presents it to the user. The input is the generated response data from step 7, and the output is what the terminal displays to the user or provides as audio feedback. The specific actions involve network transmission and output processing at the terminal.

[0851] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0852] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0853] 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 this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0854] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0855] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0856] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0857] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0858] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0859] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0860] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0861] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0862] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0863] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0864] 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.

[0865] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0866] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0867] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0868] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0869] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0870] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0871] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0872] The following is further disclosed regarding the embodiments described above.

[0873] (Claim 1)

[0874] Means for acquiring audio data,

[0875] A means for identifying the speaker from acquired audio data,

[0876] A means to restrict the response content when it is determined that the identified speaker is not a specific user,

[0877] Means of providing the response content to the user,

[0878] A system that includes this.

[0879] (Claim 2)

[0880] The system according to claim 1, comprising means for detecting participants in a conversation and notifying the user based on acquired audio data.

[0881] (Claim 3)

[0882] The system according to claim 1, comprising means for authenticating the speaker using speech recognition technology and controlling the handling of private information.

[0883] "Example 1"

[0884] (Claim 1)

[0885] Means for acquiring audio information,

[0886] A means for identifying the speaker of a voice from acquired voice information,

[0887] A means for restricting the response content when the identified voice speaker is determined to be someone other than a specific user,

[0888] A means of presenting the response content to the user,

[0889] A means for transmitting audio information to an information processing device,

[0890] A means of authenticating the speaker using speech recognition technology,

[0891] Means for setting privacy levels,

[0892] A means of generalizing the generated response,

[0893] A system that includes this.

[0894] (Claim 2)

[0895] The system according to claim 1, which detects participants in a conversation based on acquired audio information and reports them to the user.

[0896] (Claim 3)

[0897] The system according to claim 1, which provides secure information using a response generation algorithm.

[0898] "Application Example 1"

[0899] (Claim 1)

[0900] A device for acquiring voice information,

[0901] A device that identifies the speaker from acquired voice information,

[0902] A device that restricts response information when it is determined that the identified speaker is not a specific user,

[0903] A device that presents restricted response information to the user,

[0904] A device that applies a response generation algorithm to provide a safe response when different speakers are present,

[0905] A system that includes this.

[0906] (Claim 2)

[0907] The system according to claim 1, comprising a device that detects participants in a conversation based on acquired voice information and notifies the user.

[0908] (Claim 3)

[0909] The system according to claim 1, comprising a device that uses speech recognition technology to authenticate the speaker, control the handling of personal information, and activate a privacy mode when a different speaker is detected.

[0910] "Example 2 of combining an emotion engine"

[0911] (Claim 1)

[0912] Means for acquiring audio data,

[0913] A means for identifying the speaker from acquired audio data,

[0914] If the identified speaker is determined to be someone other than a specific user, a means to enable privacy mode and restrict the content of the response,

[0915] A method for analyzing emotional states from acquired audio data,

[0916] A means of adjusting the tone of response and the level of detail of information according to the analyzed emotional state,

[0917] Means of providing the response content to the user,

[0918] A system that includes this.

[0919] (Claim 2)

[0920] The system according to claim 1, which includes means for detecting participants in a conversation based on acquired audio data and notifying the user of the presence of other people based on the detected participant information.

[0921] (Claim 3)

[0922] The system according to claim 1, comprising means for authenticating the speaker using speech recognition technology, identifying the emotional state using an emotion engine, and dynamically controlling the handling of privacy information and the response to the user.

[0923] "Application example 2 when combining with an emotional engine"

[0924] (Claim 1)

[0925] A device for acquiring audio data,

[0926] A device for identifying the speaker from acquired audio data,

[0927] A device for restricting the content of responses when it is determined that the identified speaker is not a specific user,

[0928] A device for presenting the response content to the user,

[0929] A device for analyzing the emotional state of users,

[0930] A device for adjusting responses according to the user's emotional state,

[0931] A system that includes this.

[0932] (Claim 2)

[0933] The system according to claim 1, comprising a device for detecting participants in a conversation and notifying the user based on acquired audio data.

[0934] (Claim 3)

[0935] The system according to claim 1, comprising a device for performing speaker authentication and controlling the management of personal information using speech recognition technology. [Explanation of Symbols]

[0936] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A device for acquiring voice information, A device that identifies the speaker from acquired voice information, A device that restricts response information when it is determined that the identified speaker is not a specific user, A device that presents restricted response information to the user, A device that applies a response generation algorithm to provide a safe response when different speakers are present, A system that includes this.

2. The system according to claim 1, comprising a device that detects participants in a conversation based on acquired voice information and notifies the user.

3. The system according to claim 1, comprising a device that uses speech recognition technology to authenticate the speaker, control the handling of personal information, and activate a privacy mode when a different speaker is detected.

Images