System

Abstract

A system is provided.SOLUTION: The system includes a sound recognition means for collecting a sound datum, a communication means for transmitting the sound datum to a server, a sound recognition engine means for converting the sound datum into a text in the server, a generation AI means for summarizing a sound recognition result, a speech synthesis means for converting the summarized text datum into the sound datum, and a bone-conduction means for transmitting the sound datum to a wearer by bone-conduction.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] In everyday life, it is important to be able to quickly and accurately understand what the other person is saying. However, when communicating with someone who speaks quickly, verbosely, or whose explanations are difficult to understand, it can be difficult to grasp the main points of the conversation, which can be stressful. Furthermore, in an international environment where multiple languages ​​are spoken, language barriers can be an obstacle to communication. In such situations, there is a demand for technology that can efficiently understand the content of conversations and enable smooth communication. [Means for solving the problem]

[0005] This invention relates to a system that efficiently summarizes the content of a conversation and conveys it to the wearer. Specifically, the system includes a speech recognition unit that collects speech data, a communication unit that transmits the speech data to a server, a speech recognition engine that converts the speech data to text on the server, a generation AI unit that summarizes the speech recognition results, a speech synthesis unit that converts the summarized text data into speech data, and a bone conduction unit that transmits the speech data to the wearer via bone conduction. Furthermore, a multilingual translation function can be added to create a system that can be used in a multinational environment. This system enables quick understanding of the main points of a conversation and smooth communication. Furthermore, by using a wearable device, users can use the system in a natural way.

[0006] "Voice recognition means" is a general term for devices and software that have the function of collecting surrounding sounds and converting them into digital voice data.

[0007] "Communication means" is a general term for wireless communication technology and wired communication technology used to transmit collected voice data to a server.

[0008] "Speech recognition engine means" is a general term for software and devices that convert voice data into text data in the server.

[0009] "Generative AI means" is a general term for systems and software that use artificial intelligence technology to analyze text from speech recognition results and generate summaries.

[0010] "Speech synthesis means" is a general term for software or devices that convert summarized text data into synthetic speech.

[0011] "Bone conduction means" is a general term for devices and modules that transmit audio data to the wearer using bone conduction technology.

[0012] "Wearable device" is a general term for devices that can be worn by the wearer, and specifically includes eyeglasses and headsets. [Brief explanation of the drawings]

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

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

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

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

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

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

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

[0020] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0021] [First embodiment]

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

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

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

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

[0026] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

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

[0028] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

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

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

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

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

[0034] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device, such as glasses, worn by a user. The program processing of this system is specifically described below.

[0035] Collection and transmission of voice data

[0036] 1. The device activates the built-in microphone and collects surrounding audio in real time.

[0037] 2. The device stores the collected voice data as digital data.

[0038] 3. The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi.

[0039] Speech-to-text conversion and summary generation

[0040] 4. The server records the received audio data.

[0041] 5. The server starts the speech recognition engine and converts the voice data into text data.

[0042] Example: The server analyzes user A's utterance "I will report on the progress of the next project" and converts it into the text "I will report on the progress of the next project."

[0043] 6. The server uses generative AI means to summarize the converted text data.

[0044] Example: Summarize the text "I will report on the progress of the next project" as "Project progress report."

[0045] Sending summary data and synthesizing speech

[0046] 7. The server compresses the generated summary text and sends it to the terminal.

[0047] 8. The terminal runs the summary text received through a speech synthesis engine and converts it into voice data.

[0048] Example: The device outputs the summary text "Project progress report" as synthesized speech.

[0049] Sound transmission through bone conduction

[0050] 9. The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[0051] Example: A user receives audio summaries such as "Project Status Report" via bone conduction without using their ears.

[0052] User Feedback

[0053] 10. Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[0054] Example: User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next conversation.

[0055] Addition of multilingual translation function

[0056] Furthermore, this system can be equipped with a multilingual translation function, specifically, a function to translate conversations in different languages ​​in real time and to convey the translation results to the user in a summarized form.

[0057] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[0058] The processing flow will be explained below.

[0059] Step 1:

[0060] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[0061] Step 2:

[0062] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[0063] Step 3:

[0064] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[0065] Step 4:

[0066] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[0067] Step 5:

[0068] The server invokes a speech recognition engine to convert the voice data into text data, where a speech recognition algorithm analyzes the voice data and generates corresponding text.

[0069] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[0070] Step 6:

[0071] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[0072] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[0073] Step 7:

[0074] The server compresses the generated summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0075] Step 8:

[0076] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0077] Example: Generate the summary text "Project progress report" as machine speech "Project progress report."

[0078] Step 9:

[0079] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0080] Step 10:

[0081] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[0082] Example: User B receives a summary audio such as "Project Progress Report" and quickly grasps the main points of User A's speech.

[0083] Example 1

[0084] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0085] Conventional voice data collection and analysis systems are unable to collect, transmit, and analyze voice data in real time quickly enough, making it difficult for users to easily grasp information. Furthermore, when it comes to multilingual support, translation accuracy and real-time performance have been issues. The present invention aims to solve these problems.

[0086] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0087] In this invention, the server includes means for activating and collecting the built-in microphone in real time, means for saving the voice data as digital data, means for compressing the voice data, means for recording the received voice data, and means for compressing and transmitting the generated summary text. This enables the collection, analysis, summarization, and transmission of voice data in real time, and allows for rapid response to voice data in multiple languages.

[0088] "Speech recognition means" refers to a device or technology that collects voice data and captures the voice as digital data.

[0089] "Communication means" refers to the technology or protocol used to transmit collected audio data to the server, and generally includes wireless communication such as Bluetooth or Wi-Fi.

[0090] "Speech recognition engine means" refers to software or algorithms that analyze received voice data and convert it into text data.

[0091] "Generative AI methods" refer to artificial intelligence techniques that analyze and summarize text data, in particular using generative models.

[0092] "Speech synthesis means" refers to software or technology that converts text data into speech data.

[0093] "Bone conduction means" is a technology that transmits audio data as vibrations through the bones to deliver audio information to the user.

[0094] "Means for activating and collecting audio from a built-in microphone in real time" refers to a technology that activates a microphone built into a device and collects audio on the spot in real time.

[0095] "Means for storing audio data as digital data" refers to technology or devices that store collected audio data in digital form.

[0096] "Means for compressing audio data" refers to a compression algorithm or technique for reducing the file size of audio data.

[0097] "Means for recording received voice data" refers to the technology that allows the server to store the received voice data in a database so that it can be accessed later.

[0098] "Means for compressing and transmitting generated summary text" means a technology or algorithm for compressing the summary text generated by the generating AI means and transmitting it to the terminal.

[0099] A "wearable device" is a portable electronic device that can be worn by a user.

[0100] The present invention relates to a system including a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device such as glasses worn by a user.

[0101] Collection and transmission of voice data

[0102] First, the device activates its built-in microphone to collect surrounding sounds in real time. For example, if the user is in a meeting, the device's microphone will record the speech of the meeting participants in real time. This collected audio data is then stored internally as digital data.

[0103] The device then compresses the collected audio data and sends it to a server via Bluetooth or Wi-Fi, for example in AAC format.

[0104] Speech-to-text conversion and summary generation

[0105] The server records the received voice data and invokes a speech recognition engine such as the Google Cloud Speech-to-Text API to convert the voice data into text data. For example, User A's utterance "I will report on the progress of the next project" is converted into text. The server then summarizes this text data using OpenAI's generative AI method. Specifically, the text "I will report on the progress of the next project" is summarized as "Project progress report."

[0106] Sending summary data and synthesizing speech

[0107] The server compresses the summarized text data using a compression algorithm such as GZIP and then transmits it back to the device via Bluetooth or Wi-Fi. The device then converts the received summary text into voice data using a speech synthesis engine such as Amazon Polly.

[0108] Sound transmission through bone conduction

[0109] The device then inputs the synthesized voice data into the bone conduction module and transmits it to the user. For example, the user can receive a summary of a "project progress report" without using their ears through bone conduction technology, which transmits sound information via the user's jawbone and skull.

[0110] User Feedback

[0111] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation. Specifically, User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next part of the conversation.

[0112] Addition of multilingual translation function

[0113] The system can also be equipped with a multilingual translation function, for example, to translate conversations in different languages ​​in real time and provide a summary of the translation results to the user.

[0114] Examples of prompt statements

[0115] User says: "I'll report on the progress of my next project."

[0116] Prompt for generative AI model: "Summarize the following text: Report on the progress of the following project."

[0117] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[0118] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0119] Step 1:

[0120] The device activates the built-in microphone and collects surrounding audio in real time.

[0121] Input: Ambient audio.

[0122] Output: Collected audio data.

[0123] Specific operation: The glasses-type wearable device detects the user's voice input and activates the microphone. For example, the device records what is being said during a meeting.

[0124] Step 2:

[0125] The voice data collected by the device is stored as digital data.

[0126] Input: Collected audio data.

[0127] Output: Audio data in digital format.

[0128] Specific operation: The audio data is saved in the device's internal memory in MP3 format. For example, the collected audio data is temporarily saved in a buffer.

[0129] Step 3:

[0130] The device compresses the audio data it collects and sends it to the server via Bluetooth or Wi-Fi.

[0131] Input: Audio data in digital format.

[0132] Output: Compressed audio data, notification of completion of transmission to the server.

[0133] Specific operation: Using a compression algorithm such as AAC, the compressed audio data is sent to a server via Bluetooth or Wi-Fi. For example, the data is sent via a home Wi-Fi network.

[0134] Step 4:

[0135] The server records the received audio data.

[0136] Input: Compressed audio data.

[0137] Output: Recorded audio data.

[0138] Specific operation: The received voice data is stored in a database and managed with a timestamp and identification information. For example, the data is recorded in a relational database on the cloud.

[0139] Step 5:

[0140] The server activates a voice recognition engine and converts the voice data into text data.

[0141] Input: Recorded audio data.

[0142] Output: The converted text data.

[0143] Specific operation: The voice data is analyzed using the Google Cloud Speech-to-Text API and converted into text data. For example, the speech of User A is converted into the text "I will report on the progress of the next project."

[0144] Step 6:

[0145] The server uses generative AI means to summarize the converted text data.

[0146] Input: The converted text data.

[0147] Output: Summarized text data.

[0148] Specific operation: Send a prompt to a generative AI model (e.g., GPT-4) and obtain a summarized text. For example, input the prompt "Summarize the following text: Report on the progress of the following project" and obtain the summary "Project progress report."

[0149] Step 7:

[0150] The server compresses the generated summary text and transmits it to the terminal.

[0151] Input: Abstracted text data.

[0152] Output: Compressed summary text data.

[0153] Specific operation: Compress the summary text using the GZIP algorithm and send it to the terminal via the Wi-Fi module. For example, send the summary text through your home Wi-Fi network.

[0154] Step 8:

[0155] The summary text received by the terminal is passed through a speech synthesis engine and converted into voice data.

[0156] Input: Received summary text data.

[0157] Output: Audio data.

[0158] Specific operation: Calls a speech synthesis engine such as Amazon Polly and generates synthetic speech data from text. For example, it outputs the received text "Project progress report" as synthetic speech.

[0159] Step 9:

[0160] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[0161] Input: Synthetic speech data.

[0162] Output: Audio information via bone conduction.

[0163] Specific operation: The synthesized voice data is sent to the bone conduction module, and the voice is transmitted to the user via the jawbone or skull. For example, the synthesized voice is transmitted as vibrations through the jawbone.

[0164] Step 10:

[0165] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[0166] Input: Audio information via bone conduction.

[0167] Output: Understanding, next steps in the conversation.

[0168] Specific actions: The user understands the audio from the bone conduction speaker and considers the next statement or action based on the content. For example, the user understands a "project progress report" and performs the action of providing the necessary feedback.

[0169] (Application example 1)

[0170] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0171] Modern manufacturing demands improved work efficiency and safety. Collaboration between workers and robots within factories is particularly important, but language barriers and communication delays pose challenges. Current systems make it difficult for workers who speak different languages ​​to communicate with each other, reducing the efficiency of instruction transmission. Furthermore, as communication methods utilizing bone conduction technology have not yet been fully established, there is a need for intuitive and rapid instruction transmission.

[0172] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0173] In this invention, the server includes a speech recognition engine means for converting voice data into text, a generation AI means, and a speech synthesis means, which enable a voice instruction means for giving instructions to factory automation equipment using a visual support device worn by a worker.

[0174] A "voice recognition means" is a device or component that collects voice data and stores it as digital data.

[0175] "Communication means" refers to the technology and protocols used to transmit collected audio data to the server, including Bluetooth and Wi-Fi.

[0176] "Speech recognition engine means" refers to software or algorithms that run on the server and convert received voice data into text.

[0177] "Generative AI means" refers to artificial intelligence models or algorithms that analyze the converted text data and generate summaries or instructions.

[0178] "Speech synthesis means" refers to software or technology for converting the generated summary text data into speech data.

[0179] "Bone conduction means" refers to a device or module for transmitting audio data to a wearer using bone conduction technology.

[0180] A "visual support device" is a wearable device worn by a worker that provides visual and audio information. Specific examples include smart glasses.

[0181] "Factory automation equipment" refers to automated equipment and robots used in factories.

[0182] "Voice instruction means" refers to a system or technology that allows a worker to give instructions to automated factory equipment through voice using a visual assistance device.

[0183] The present invention relates to a voice guidance system for automated factory equipment using a visual assistance device worn by an operator. The system includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, a bone conduction unit, and a voice guidance unit.

[0184] Voice recognition means

[0185] First, the built-in microphone of the visual assistance device (e.g., smart glasses) is activated and collects the worker's voice in real time. The collected voice data is saved as digital data, allowing the worker's voice to be accurately captured.

[0186] communication means

[0187] The communication device then compresses the collected voice data and transmits it to a server via Bluetooth or Wi-Fi, efficiently transmitting large amounts of voice data.

[0188] Speech Recognition Engine Means

[0189] The server records the received voice data and activates a speech recognition engine. Specifically, it uses Google's speech recognition API or similar technology to convert the voice data into text data. For example, instructions such as "Please prepare to assemble the next part" are converted into text.

[0190] Generation AI means

[0191] The server inputs the converted text data into a generative AI means and generates a summary using a generative AI model (e.g., GPT-3). For example, the text "Please prepare to assemble the following parts" is summarized as "Prepare to assemble parts."

[0192] Voice synthesis means

[0193] The server sends the summarized text data to the visual support device, which then converts the summarized text into speech data using a speech synthesis engine (e.g., pyttsx3), generating a summary instruction as speech.

[0194] Bone conduction means

[0195] The visual assistance device inputs the generated voice data into the bone conduction module and transmits it to the worker, who can receive summary voice such as "Prepare for parts assembly" via bone conduction without using their ears.

[0196] Audio instruction tools

[0197] This allows workers wearing visual assistance devices to receive voice instructions and provide accurate and prompt voice guidance to automated factory equipment, which can, for example, facilitate smooth communication between workers who speak different languages ​​and improve the efficiency of instruction transmission.

[0198] Specific examples

[0199] Say: "Get ready to assemble the next part."

[0200] Generative AI summary: "Parts ready for assembly"

[0201] Bone conduction feedback: "Preparing to assemble parts"

[0202] Example prompts to input to a generative AI model:

[0203] Get ready to assemble the next part

[0204] In this invention, the voice recognition means, generation AI means, bone conduction means, etc. cooperate with each other at each step to provide optimal voice guidance to workers, which is expected to significantly improve work efficiency and safety in factories.

[0205] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0206] Step 1:

[0207] The terminal activates the built-in microphone and collects the worker's voice in real time. The collected voice data is saved as digital data. Here, the input is the worker's voice and the output is digital voice data. In this step, data conversion is performed to save it as digital voice data.

[0208] Step 2:

[0209] The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi. Here, the input is digital audio data and the output is compressed audio data. In this step, a compression algorithm is used to reduce the data size and the data is sent to the server using a communication protocol.

[0210] Step 3:

[0211] The server records the received voice data and starts the voice recognition engine. Here, the input is compressed voice data and the output is text data. In this step, the voice recognition process is performed to convert the voice data into text. Specifically, Google's voice recognition API is used.

[0212] Step 4:

[0213] The server uses a generative AI method to summarize the converted text data, where the input is text data and the output is a summary text. In this step, a generative AI model (e.g., GPT-3) analyzes the input text and generates a summary.

[0214] Step 5:

[0215] The server compresses the generated summary text and sends it to the terminal. Here, the input is the summary text and the output is the compressed summary text. In this step, a compression algorithm is used to reduce the data size and the data is sent to the terminal using a communication protocol.

[0216] Step 6:

[0217] The terminal converts the received summary text into speech data through a speech synthesis engine. The input is the summary text and the output is speech data. Specifically, the text is converted into speech using the pyttsx3 engine.

[0218] Step 7:

[0219] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the worker. Here, the input is voice data, and the output is the voice that the worker receives through bone conduction. In this step, the voice data is transmitted without using the ears using bone conduction technology.

[0220] Step 8:

[0221] The worker understands the summarized audio received through bone conduction technology and gives instructions to the automated factory equipment. Here, the input is the audio via bone conduction, and the output is audio instructions to the automated factory equipment. In this step, the worker performs the actual work based on the audio instructions.

[0222] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0223] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, a bone conduction unit, and an emotion engine. This system can be implemented using a wearable device such as glasses worn by a user. Specific program processing of this system is described in detail below.

[0224] Collection and transmission of voice data

[0225] 1. The device activates the built-in microphone and collects surrounding audio in real time. The collected audio data is temporarily stored in a buffer.

[0226] 2. The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing if necessary.

[0227] 3. The device sends the compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are pre-set.

[0228] Speech-to-text conversion and summary generation

[0229] 4. The server saves the received voice data in a specific directory and prepares to pass it to the voice recognition engine.

[0230] 5. The server invokes the speech recognition engine to convert the audio data to text data. Here, the speech recognition algorithm analyzes the audio data and generates corresponding text.

[0231] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[0232] 6. The server starts the generation AI means and receives the text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[0233] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[0234] Recognizing user emotions with an emotion engine

[0235] 7. The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it to the emotion engine.

[0236] 8. The emotion engine analyzes the collected data and recognizes the user's current emotional state (e.g., excited, relaxed, anxious, etc.).

[0237] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[0238] Emotion regulation and transmission of summary data

[0239] 9. The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[0240] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[0241] 10. The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0242] Speech synthesis of summary text and transmission via bone conduction

[0243] 11. The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0244] Example: Generate the summary text "progress report" as machine speech "progress report."

[0245] 12. The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0246] User Feedback

[0247] 13. Users can understand the summary audio received through bone conduction and efficiently grasp the content of the conversation.

[0248] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[0249] This system allows users to not only efficiently receive speech recognition and summarized information, but also receive optimal feedback according to their emotional state, enabling them to quickly and accurately understand the content of conversations and achieve smooth communication.

[0250] The processing flow will be explained below.

[0251] Step 1:

[0252] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[0253] Step 2:

[0254] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[0255] Step 3:

[0256] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are preset.

[0257] Step 4:

[0258] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[0259] Step 5:

[0260] The server activates a speech recognition engine to convert the speech data into text data, which then analyzes the speech and generates corresponding text.

[0261] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[0262] Step 6:

[0263] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[0264] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[0265] Step 7:

[0266] The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it on to the emotion engine.

[0267] Step 8:

[0268] The emotion engine analyzes the collected biometric data to recognize the user's current emotional state, which can be classified as excited, relaxed, anxious, etc.

[0269] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[0270] Step 9:

[0271] The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[0272] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[0273] Step 10:

[0274] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0275] Step 11:

[0276] The device passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0277] Example: Generate the summary text "progress report" as machine speech "progress report."

[0278] Step 12:

[0279] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0280] Step 13:

[0281] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[0282] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[0283] Example 2

[0284] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0285] Conventional speech recognition systems only convert speech data into text, and have the problem of being unable to provide information that takes into account the user's emotional state. Furthermore, the collection of speech data, summary generation, emotion recognition, and speech synthesis are all performed separately, which reduces the overall efficiency of the system. Therefore, there is a need for an integrated system that can reduce the burden on users and provide information quickly and accurately.

[0286] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a processor that converts voice data into text, a generative model that summarizes the text data, an emotion recognition means that analyzes the user's emotional state, and a voice synthesis means that converts the summarized text data into voice data. This makes it possible to perform everything from voice collection to emotion analysis, summary generation, and voice synthesis in an integrated manner.

[0287] A "device for collecting audio data" is a device that has the function of collecting surrounding audio in real time and storing it in a buffer.

[0288] A "communication means" is a device that has the function of compressing collected voice data and sending it to a server using Bluetooth, Wi-Fi, etc.

[0289] "Processor" means a central processing unit that executes speech recognition algorithms within the server to convert speech data into text data.

[0290] A "generative model" is an artificial intelligence model that analyzes text data, extracts important content, and generates a summary.

[0291] The "emotion recognition means" is a device or algorithm for analyzing biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[0292] The "speech synthesis means" is a synthesis device that has the function of converting summarized text data into natural speech.

[0293] An "acoustic device" is a device that has the function of transmitting audio data to a user through bone conduction.

[0294] A "wearable device" is a portable device that can be worn by a user and that collects and communicates audio data.

[0295] This invention is a system that integrates voice recognition, generative AI, emotion recognition, voice synthesis, and voice transmission via bone conduction, and provides efficient information provision using a wearable device worn by the user.

[0296] The system includes the following main components:

[0297] 1. Device for collecting voice data: A microphone built into the wearable device collects the user's voice and surrounding voices in real time and stores the data in a buffer.

[0298] 2. Communication method: The collected voice data is converted into digital form, compressed if necessary, and then sent to the server via Bluetooth, Wi-Fi, etc.

[0299] 3. Processor: The central processing unit installed in the server converts the received voice data into text data using a voice recognition algorithm (e.g., Google Cloud Speech-to-Text API).

[0300] 4. Generative models: These include artificial intelligence models (e.g., GPT-4) that analyze text data, extract key content, and generate summaries. Generative models use prompts as input and generate corresponding summary text.

[0301] Example prompt: "I'll report on the progress of my next project."

[0302] 5. Emotion recognition means: Sensors built into smart glasses and other wearable devices collect biometric data such as the user's voice, facial expressions, and pulse rate, and then use emotion recognition algorithms to analyze the user's emotional state.

[0303] Example: Judging whether a user is "excited" based on their tone of voice and facial expression.

[0304] 6. Speech synthesis means: Includes a speech synthesis engine (e.g., Amazon Polly) for converting the summarized text data into natural-sounding speech, which then converts the summarized text into speech data and transmits it to the user.

[0305] Example: Generate the summary text "progress report" as machine speech "progress report."

[0306] 7. Acoustic Device: Includes devices that utilize bone conduction technology to transmit synthesized audio data directly to the user's inner ear through the skull, allowing the user to receive audio information without blocking their ears.

[0307] This system allows users to benefit from:

[0308] Not only can you receive voice recognition and summarized information efficiently, but you can also receive optimal feedback based on your emotional state.

[0309] The entire process, from collecting voice data to analyzing, summarizing, synthesizing voice, and finally transmitting it, is managed centrally, ensuring that information is provided quickly and accurately.

[0310] This system is particularly useful in situations where large amounts of information need to be efficiently processed and transmitted in real time, such as business meetings and educational settings. The present invention allows users to achieve smooth communication and information comprehension.

[0311] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0312] Step 1:

[0313] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[0314] Input: Ambient audio

[0315] Output: Buffered audio data

[0316] Specific operation: The microphone of the wearable device (such as smart glasses) is activated and collects the user's speech and surrounding sounds. The collected data is temporarily stored in a buffer in memory.

[0317] Step 2:

[0318] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[0319] Input: Buffered audio data

[0320] Output: Digital audio data (compressed)

[0321] Specific operation: A processor operates to convert analog audio data into digital, and then a codec is applied to compress the audio data.

[0322] Step 3:

[0323] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[0324] Input: Digital audio data (compressed)

[0325] Output: Audio data sent to the server

[0326] What happens: Compressed audio data is sent via Wi-Fi or Bluetooth modules and reaches the specified API endpoint on the server.

[0327] Step 4:

[0328] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[0329] Input: Audio data sent to the server

[0330] Output: Input file for speech recognition engine

[0331] Specific operation: The voice data is stored in a specific directory on the server, and the voice recognition engine is configured to refer to this file.

[0332] Step 5:

[0333] The server activates a speech recognition engine to convert the speech data into text data. The speech recognition algorithm analyzes the speech data and generates corresponding text.

[0334] Input: Input file for the speech recognition engine

[0335] Output: Text data

[0336] Specific operation: The process of converting speech to text is carried out by calling the Google Cloud Speech-to-Text API, etc. For example, speech saying "I will report on the progress of the next project" is converted to text saying "I will report on the progress of the next project."

[0337] Step 6:

[0338] The server launches the generative model and receives text data from the speech recognition engine. The generative AI analyzes the text data, extracts important content, and generates a summary.

[0339] Input: Text data

[0340] Output: Summary text

[0341] Specific operation: A GPT model (e.g., GPT-4) analyzes the text data "I will report on the progress of the next project" and summarizes it as "Project progress report."

[0342] Step 7:

[0343] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate and passes it to an emotion recognition means.

[0344] Input: User biometric data (voice, facial expression, pulse, etc.)

[0345] Output: Emotional state data

[0346] How it works: Sensors built into smart glasses or wearable devices collect biometric data from users and send it to emotion recognition algorithms.

[0347] Step 8:

[0348] An emotion recognition means analyzes the collected data and recognizes the user's current emotional state (eg, excited, relaxed, anxious, etc.).

[0349] Input: Biometric data

[0350] Output: Emotional state data

[0351] Specific operation: The emotion recognition algorithm analyzes the tone of voice and facial expressions to determine the user's emotional state, such as "excited" or "relaxed."

[0352] Step 9:

[0353] The server adjusts the content of the generated summary text based on the analysis results of the emotion recognition means, making the summary more concise if the user is emotionally excited.

[0354] Input: Summary text, emotional state data

[0355] Output: Adjusted summary text

[0356] Specific operation: The server adjusts the generated summary "Project Progress Report" to be more concise as "Progress Report" based on the emotional state data.

[0357] Step 10:

[0358] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0359] Input: Adjusted summary text

[0360] Output: Summary text sent to terminal

[0361] What it does: The adjusted summary text is sent to your device using Bluetooth or Wi-Fi.

[0362] Step 11:

[0363] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0364] Input: Adjusted summary text

[0365] Output: Synthesized speech data

[0366] What happens: A speech synthesis engine such as Amazon Polly converts the text "progress report" into machine-generated speech.

[0367] Step 12:

[0368] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0369] Input: Synthetic speech data

[0370] Output: Transmission of audio information via bone conduction

[0371] Specific operation: The synthesized voice data is transmitted directly to the user's inner ear through the bone conduction module, and the user hears the voice.

[0372] Step 13:

[0373] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[0374] Input: Audio information via bone conduction

[0375] Output: Understood conversation

[0376] Specific operation: The user receives a summary audio such as "progress report" via bone conduction and quickly grasps the main points of the talk.

[0377] The above are the specific steps of the program processing of this system, which allows users to efficiently enjoy speech recognition, summary generation, emotion recognition, speech synthesis, and bone conduction communication.

[0378] (Application example 2)

[0379] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0380] Security operations require accurate and rapid understanding of on-site situations in real time and the most appropriate course of action. However, conventional methods are prone to information delays and miscommunication, making it difficult to respond immediately based on the urgency of the situation and the emotional state of the guards. To solve this problem, a system is needed that summarizes the on-site situation in real time and provides information that is appropriately adjusted according to the emotional state of the guards.

[0381] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a voice recognition engine means for converting voice data into text, a generation AI means for summarizing the voice recognition results, and an emotion engine means for collecting user emotion data and adjusting the summarized text data based on the emotion data. This makes it possible to immediately grasp the situation on-site and provide optimal information according to the user's emotional state.

[0382] A "voice recognition means" is a device or method that collects voice data in real time and converts it into digital voice data.

[0383] "Communication means" refers to a means for transmitting collected voice data to a server, and includes wireless communication technologies such as Bluetooth and Wi-Fi.

[0384] "Speech recognition engine means" refers to software or algorithms that analyze voice data and convert it into corresponding text data.

[0385] "Generative AI means" refers to means that use artificial intelligence technology to analyze converted text data, extract important content, and generate a summary.

[0386] The "speech synthesis means" refers to a means for converting summarized text data into speech data, and includes an algorithm for generating natural-sounding speech.

[0387] "Bone conduction means" refers to devices or technologies that transmit sound directly through the wearer's skull to the inner ear.

[0388] "Emotion engine means" refers to technology or devices that analyze biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[0389] The present invention is a security system that includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generative AI unit, a voice synthesis unit, a bone conduction unit, and an emotion engine unit. Specifically, the system uses a wearable device worn by a security guard. With this system, the security guard reports the situation on-site by voice, and the voice data is sent to a server for summarization and emotion analysis. Appropriately adjusted information is then provided to the security guard in real time via bone conduction.

[0390] The device activates its built-in microphone and collects audio from the site in real time. This audio data is temporarily stored in a buffer and converted into digital audio data. The compressed audio data is then sent to a server via communication methods such as Bluetooth or Wi-Fi.

[0391] The server converts the received voice data into text data using a voice recognition engine means. For example, voice data reported by a security guard that "a suspicious person has broken in" is converted into text data. Next, a generation AI means analyzes the text data, extracts important content, and generates a summary. For example, "a suspicious person has broken in" is summarized as "suspicious person has broken in." After that, an emotion engine means recognizes the user's emotional state from their voice and facial expression, and adjusts the summary text based on the emotion data. For example, if the user is in a tense state, the summary text is further simplified to "Emergency! Suspicious person has broken in."

[0392] The resulting adjusted summary text data is converted into natural-sounding speech using a speech synthesis device. This speech data is transmitted to the security guard via bone conduction. The security guard receives the speech information via bone conduction, allowing them to quickly grasp important information on the scene.

[0393] For example, when a security guard reports that "a suspicious person has entered the entrance," the system summarizes the voice data as "Suspicious person has entered," and after sensing the guard's state of tension, further simplifies it to "Emergency! Suspicious person has entered," and transmits it via bone conduction. In this way, the security guard can instantly take the most appropriate action depending on the situation.

[0394] Example prompt sentence:

[0395] "A suspicious person entered the entrance" Summarize the situation in one sentence: "A suspicious person entered the entrance"

[0396] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0397] Step 1:

[0398] The terminal activates the built-in microphone and collects audio from the scene in real time. The input is real-time audio data, and the output is digital audio data temporarily stored in a buffer.

[0399] Step 2:

[0400] The terminal converts the buffered audio data into a digital format and compresses it if necessary. The input is the audio data in the buffer, and the output is compressed digital audio data.

[0401] Step 3:

[0402] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The input is compressed digital audio data, and the output is audio data sent to the server.

[0403] Step 4:

[0404] The server stores the received voice data in a specific directory and prepares to pass it to the voice recognition engine means. The input is the voice data sent to the server, and the output is the voice data for the recognition engine to process.

[0405] Step 5:

[0406] The server converts the voice data into text data using a voice recognition engine means, where the input is the received voice data and the output is the converted text data.

[0407] Step 6:

[0408] The server uses generative AI methods to analyze the text data, extract important content, and generate a summary. The input is the converted text data, and the output is the summarized text data.

[0409] Step 7:

[0410] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate, and passes it to the emotion engine means. The input is the user's biometric data, and the output is data representing the user's emotional state.

[0411] Step 8:

[0412] The server uses an emotion engine means to analyze the collected data and recognize the user's current emotional state, where the input is the user's biometric data and the output is the emotional state data.

[0413] Step 9:

[0414] The server adjusts the content of the generated summary text based on the emotional state data. The input is the emotional state data and the summary text data, and the output is the adjusted summary text data.

[0415] Step 10:

[0416] The server compresses the adjusted summary text data and sends it to the terminal, where the input is the adjusted summary text data and the output is the compressed text data sent to the terminal.

[0417] Step 11:

[0418] The terminal passes the received summary text to a speech synthesis engine means for converting it into synthesized speech data, with the adjusted summary text data as input and the synthesized speech data as output.

[0419] Step 12:

[0420] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user. The input is the synthesized voice data, and the output is the transmission of voice through the bone conduction module.

[0421] Step 13:

[0422] The user understands the summary speech received through bone conduction and efficiently grasps the content of the conversation and the situation on the scene. The input is the speech data received through bone conduction, and the output is the understood information.

[0423] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0424] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0425] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0426] [Second embodiment]

[0427] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0428] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0429] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0430] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0431] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0432] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0433] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0434] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0435] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0436] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0437] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0438] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0439] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device, such as glasses, worn by a user. The program processing of this system is specifically described below.

[0440] Collection and transmission of voice data

[0441] 1. The device activates the built-in microphone and collects surrounding audio in real time.

[0442] 2. The device stores the collected voice data as digital data.

[0443] 3. The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi.

[0444] Speech-to-text conversion and summary generation

[0445] 4. The server records the received audio data.

[0446] 5. The server starts the speech recognition engine and converts the voice data into text data.

[0447] Example: The server analyzes user A's utterance "I will report on the progress of the next project" and converts it into the text "I will report on the progress of the next project."

[0448] 6. The server uses generative AI means to summarize the converted text data.

[0449] Example: Summarize the text "I will report on the progress of the next project" as "Project progress report."

[0450] Sending summary data and synthesizing speech

[0451] 7. The server compresses the generated summary text and sends it to the terminal.

[0452] 8. The terminal runs the summary text received through a speech synthesis engine and converts it into voice data.

[0453] Example: The device outputs the summary text "Project progress report" as synthesized speech.

[0454] Sound transmission through bone conduction

[0455] 9. The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[0456] Example: A user receives audio summaries such as "Project Status Report" via bone conduction without using their ears.

[0457] User Feedback

[0458] 10. Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[0459] Example: User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next conversation.

[0460] Addition of multilingual translation function

[0461] Furthermore, this system can be equipped with a multilingual translation function, specifically, a function to translate conversations in different languages ​​in real time and to convey the translation results to the user in a summarized form.

[0462] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[0463] The processing flow will be explained below.

[0464] Step 1:

[0465] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[0466] Step 2:

[0467] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[0468] Step 3:

[0469] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[0470] Step 4:

[0471] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[0472] Step 5:

[0473] The server invokes a speech recognition engine to convert the voice data into text data, where a speech recognition algorithm analyzes the voice data and generates corresponding text.

[0474] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[0475] Step 6:

[0476] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[0477] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[0478] Step 7:

[0479] The server compresses the generated summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0480] Step 8:

[0481] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0482] Example: Generate the summary text "Project progress report" as machine speech "Project progress report."

[0483] Step 9:

[0484] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0485] Step 10:

[0486] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[0487] Example: User B receives a summary audio such as "Project Progress Report" and quickly grasps the main points of User A's speech.

[0488] Example 1

[0489] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0490] Conventional voice data collection and analysis systems are unable to collect, transmit, and analyze voice data in real time quickly enough, making it difficult for users to easily grasp information. Furthermore, when it comes to multilingual support, translation accuracy and real-time performance have been issues. The present invention aims to solve these problems.

[0491] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0492] In this invention, the server includes means for activating and collecting the built-in microphone in real time, means for saving the voice data as digital data, means for compressing the voice data, means for recording the received voice data, and means for compressing and transmitting the generated summary text. This enables the collection, analysis, summarization, and transmission of voice data in real time, and allows for rapid response to voice data in multiple languages.

[0493] "Speech recognition means" refers to a device or technology that collects voice data and captures the voice as digital data.

[0494] "Communication means" refers to the technology or protocol used to transmit collected audio data to the server, and generally includes wireless communication such as Bluetooth or Wi-Fi.

[0495] "Speech recognition engine means" refers to software or algorithms that analyze received voice data and convert it into text data.

[0496] "Generative AI methods" refer to artificial intelligence techniques that analyze and summarize text data, in particular using generative models.

[0497] "Speech synthesis means" refers to software or technology that converts text data into speech data.

[0498] "Bone conduction means" is a technology that transmits audio data as vibrations through the bones to deliver audio information to the user.

[0499] "Means for activating and collecting audio from a built-in microphone in real time" refers to a technology that activates a microphone built into a device and collects audio on the spot in real time.

[0500] "Means for storing audio data as digital data" refers to technology or devices that store collected audio data in digital form.

[0501] "Means for compressing audio data" refers to a compression algorithm or technique for reducing the file size of audio data.

[0502] "Means for recording received voice data" refers to the technology that allows the server to store the received voice data in a database so that it can be accessed later.

[0503] "Means for compressing and transmitting generated summary text" means a technology or algorithm for compressing the summary text generated by the generating AI means and transmitting it to the terminal.

[0504] A "wearable device" is a portable electronic device that can be worn by a user.

[0505] The present invention relates to a system including a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device such as glasses worn by a user.

[0506] Collection and transmission of voice data

[0507] First, the device activates its built-in microphone to collect surrounding sounds in real time. For example, if the user is in a meeting, the device's microphone will record the speech of the meeting participants in real time. This collected audio data is then stored internally as digital data.

[0508] The device then compresses the collected audio data and sends it to a server via Bluetooth or Wi-Fi, for example in AAC format.

[0509] Speech-to-text conversion and summary generation

[0510] The server records the received voice data and invokes a speech recognition engine such as the Google Cloud Speech-to-Text API to convert the voice data into text data. For example, User A's utterance "I will report on the progress of the next project" is converted into text. The server then summarizes this text data using OpenAI's generative AI method. Specifically, the text "I will report on the progress of the next project" is summarized as "Project progress report."

[0511] Sending summary data and synthesizing speech

[0512] The server compresses the summarized text data using a compression algorithm such as GZIP and then transmits it back to the device via Bluetooth or Wi-Fi. The device then converts the received summary text into voice data using a speech synthesis engine such as Amazon Polly.

[0513] Sound transmission through bone conduction

[0514] The device then inputs the synthesized voice data into the bone conduction module and transmits it to the user. For example, the user can receive a summary of a "project progress report" without using their ears through bone conduction technology, which transmits sound information via the user's jawbone and skull.

[0515] User Feedback

[0516] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation. Specifically, User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next part of the conversation.

[0517] Addition of multilingual translation function

[0518] The system can also be equipped with a multilingual translation function, for example, to translate conversations in different languages ​​in real time and provide a summary of the translation results to the user.

[0519] Examples of prompt statements

[0520] User says: "I'll report on the progress of my next project."

[0521] Prompt for generative AI model: "Summarize the following text: Report on the progress of the following project."

[0522] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[0523] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0524] Step 1:

[0525] The device activates the built-in microphone and collects surrounding audio in real time.

[0526] Input: Ambient audio.

[0527] Output: Collected audio data.

[0528] Specific operation: The glasses-type wearable device detects the user's voice input and activates the microphone. For example, the device records what is being said during a meeting.

[0529] Step 2:

[0530] The voice data collected by the device is stored as digital data.

[0531] Input: Collected audio data.

[0532] Output: Audio data in digital format.

[0533] Specific operation: The audio data is saved in the device's internal memory in MP3 format. For example, the collected audio data is temporarily saved in a buffer.

[0534] Step 3:

[0535] The device compresses the audio data it collects and sends it to the server via Bluetooth or Wi-Fi.

[0536] Input: Audio data in digital format.

[0537] Output: Compressed audio data, notification of completion of transmission to the server.

[0538] Specific operation: Using a compression algorithm such as AAC, the compressed audio data is sent to a server via Bluetooth or Wi-Fi. For example, the data is sent via a home Wi-Fi network.

[0539] Step 4:

[0540] The server records the received audio data.

[0541] Input: Compressed audio data.

[0542] Output: Recorded audio data.

[0543] Specific operation: The received voice data is stored in a database and managed with a timestamp and identification information. For example, the data is recorded in a relational database on the cloud.

[0544] Step 5:

[0545] The server activates a voice recognition engine and converts the voice data into text data.

[0546] Input: Recorded audio data.

[0547] Output: The converted text data.

[0548] Specific operation: The voice data is analyzed using the Google Cloud Speech-to-Text API and converted into text data. For example, the speech of User A is converted into the text "I will report on the progress of the next project."

[0549] Step 6:

[0550] The server uses generative AI means to summarize the converted text data.

[0551] Input: The converted text data.

[0552] Output: Summarized text data.

[0553] Specific operation: Send a prompt to a generative AI model (e.g., GPT-4) and obtain a summarized text. For example, input the prompt "Summarize the following text: Report on the progress of the following project" and obtain the summary "Project progress report."

[0554] Step 7:

[0555] The server compresses the generated summary text and transmits it to the terminal.

[0556] Input: Abstracted text data.

[0557] Output: Compressed summary text data.

[0558] Specific operation: Compress the summary text using the GZIP algorithm and send it to the terminal via the Wi-Fi module. For example, send the summary text through your home Wi-Fi network.

[0559] Step 8:

[0560] The summary text received by the terminal is passed through a speech synthesis engine and converted into voice data.

[0561] Input: Received summary text data.

[0562] Output: Audio data.

[0563] Specific operation: Calls a speech synthesis engine such as Amazon Polly and generates synthetic speech data from text. For example, it outputs the received text "Project progress report" as synthetic speech.

[0564] Step 9:

[0565] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[0566] Input: Synthetic speech data.

[0567] Output: Audio information via bone conduction.

[0568] Specific operation: The synthesized voice data is sent to the bone conduction module, and the voice is transmitted to the user via the jawbone or skull. For example, the synthesized voice is transmitted as vibrations through the jawbone.

[0569] Step 10:

[0570] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[0571] Input: Audio information via bone conduction.

[0572] Output: Understanding, next steps in the conversation.

[0573] Specific actions: The user understands the audio from the bone conduction speaker and considers the next statement or action based on the content. For example, the user understands a "project progress report" and performs the action of providing the necessary feedback.

[0574] (Application example 1)

[0575] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0576] Modern manufacturing demands improved work efficiency and safety. Collaboration between workers and robots within factories is particularly important, but language barriers and communication delays pose challenges. Current systems make it difficult for workers who speak different languages ​​to communicate with each other, reducing the efficiency of instruction transmission. Furthermore, as communication methods utilizing bone conduction technology have not yet been fully established, there is a need for intuitive and rapid instruction transmission.

[0577] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0578] In this invention, the server includes a speech recognition engine means for converting voice data into text, a generation AI means, and a speech synthesis means, which enable a voice instruction means for giving instructions to factory automation equipment using a visual support device worn by a worker.

[0579] A "voice recognition means" is a device or component that collects voice data and stores it as digital data.

[0580] "Communication means" refers to the technology and protocols used to transmit collected audio data to the server, including Bluetooth and Wi-Fi.

[0581] "Speech recognition engine means" refers to software or algorithms that run on the server and convert received voice data into text.

[0582] "Generative AI means" refers to artificial intelligence models or algorithms that analyze the converted text data and generate summaries or instructions.

[0583] "Speech synthesis means" refers to software or technology for converting the generated summary text data into speech data.

[0584] "Bone conduction means" refers to a device or module for transmitting audio data to a wearer using bone conduction technology.

[0585] A "visual support device" is a wearable device worn by a worker that provides visual and audio information. Specific examples include smart glasses.

[0586] "Factory automation equipment" refers to automated equipment and robots used in factories.

[0587] "Voice instruction means" refers to a system or technology that allows a worker to give instructions to automated factory equipment through voice using a visual assistance device.

[0588] The present invention relates to a voice guidance system for automated factory equipment using a visual assistance device worn by an operator. The system includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, a bone conduction unit, and a voice guidance unit.

[0589] Voice recognition means

[0590] First, the built-in microphone of the visual assistance device (e.g., smart glasses) is activated and collects the worker's voice in real time. The collected voice data is saved as digital data, allowing the worker's voice to be accurately captured.

[0591] communication means

[0592] The communication device then compresses the collected voice data and transmits it to a server via Bluetooth or Wi-Fi, efficiently transmitting large amounts of voice data.

[0593] Speech Recognition Engine Means

[0594] The server records the received voice data and activates a speech recognition engine. Specifically, it uses Google's speech recognition API or similar technology to convert the voice data into text data. For example, instructions such as "Please prepare to assemble the next part" are converted into text.

[0595] Generation AI means

[0596] The server inputs the converted text data into a generative AI means and generates a summary using a generative AI model (e.g., GPT-3). For example, the text "Please prepare to assemble the following parts" is summarized as "Prepare to assemble parts."

[0597] Voice synthesis means

[0598] The server sends the summarized text data to the visual support device, which then converts the summarized text into speech data using a speech synthesis engine (e.g., pyttsx3), generating a summary instruction as speech.

[0599] Bone conduction means

[0600] The visual assistance device inputs the generated voice data into the bone conduction module and transmits it to the worker, who can receive summary voice such as "Prepare for parts assembly" via bone conduction without using their ears.

[0601] Audio instruction tools

[0602] This allows workers wearing visual assistance devices to receive voice instructions and provide accurate and prompt voice guidance to automated factory equipment, which can, for example, facilitate smooth communication between workers who speak different languages ​​and improve the efficiency of instruction transmission.

[0603] Specific examples

[0604] Say: "Get ready to assemble the next part."

[0605] Generative AI summary: "Parts ready for assembly"

[0606] Bone conduction feedback: "Preparing to assemble parts"

[0607] Example prompts to input to a generative AI model:

[0608] Get ready to assemble the next part

[0609] In this invention, the voice recognition means, generation AI means, bone conduction means, etc. cooperate with each other at each step to provide optimal voice guidance to workers, which is expected to significantly improve work efficiency and safety in factories.

[0610] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0611] Step 1:

[0612] The terminal activates the built-in microphone and collects the worker's voice in real time. The collected voice data is saved as digital data. Here, the input is the worker's voice and the output is digital voice data. In this step, data conversion is performed to save it as digital voice data.

[0613] Step 2:

[0614] The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi. Here, the input is digital audio data and the output is compressed audio data. In this step, a compression algorithm is used to reduce the data size and the data is sent to the server using a communication protocol.

[0615] Step 3:

[0616] The server records the received voice data and starts the voice recognition engine. Here, the input is compressed voice data and the output is text data. In this step, the voice recognition process is performed to convert the voice data into text. Specifically, Google's voice recognition API is used.

[0617] Step 4:

[0618] The server uses a generative AI method to summarize the converted text data, where the input is text data and the output is a summary text. In this step, a generative AI model (e.g., GPT-3) analyzes the input text and generates a summary.

[0619] Step 5:

[0620] The server compresses the generated summary text and sends it to the terminal. Here, the input is the summary text and the output is the compressed summary text. In this step, a compression algorithm is used to reduce the data size and the data is sent to the terminal using a communication protocol.

[0621] Step 6:

[0622] The terminal converts the received summary text into speech data through a speech synthesis engine. The input is the summary text and the output is speech data. Specifically, the text is converted into speech using the pyttsx3 engine.

[0623] Step 7:

[0624] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the worker. Here, the input is voice data, and the output is the voice that the worker receives through bone conduction. In this step, the voice data is transmitted without using the ears using bone conduction technology.

[0625] Step 8:

[0626] The worker understands the summarized audio received through bone conduction technology and gives instructions to the automated factory equipment. Here, the input is the audio via bone conduction, and the output is audio instructions to the automated factory equipment. In this step, the worker performs the actual work based on the audio instructions.

[0627] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0628] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, a bone conduction unit, and an emotion engine. This system can be implemented using a wearable device such as glasses worn by a user. Specific program processing of this system is described in detail below.

[0629] Collection and transmission of voice data

[0630] 1. The device activates the built-in microphone and collects surrounding audio in real time. The collected audio data is temporarily stored in a buffer.

[0631] 2. The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing if necessary.

[0632] 3. The device sends the compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are pre-set.

[0633] Speech-to-text conversion and summary generation

[0634] 4. The server saves the received voice data in a specific directory and prepares to pass it to the voice recognition engine.

[0635] 5. The server invokes the speech recognition engine to convert the audio data to text data. Here, the speech recognition algorithm analyzes the audio data and generates corresponding text.

[0636] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[0637] 6. The server starts the generation AI means and receives the text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[0638] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[0639] Recognizing user emotions with an emotion engine

[0640] 7. The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it to the emotion engine.

[0641] 8. The emotion engine analyzes the collected data and recognizes the user's current emotional state (e.g., excited, relaxed, anxious, etc.).

[0642] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[0643] Emotion regulation and transmission of summary data

[0644] 9. The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[0645] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[0646] 10. The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0647] Speech synthesis of summary text and transmission via bone conduction

[0648] 11. The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0649] Example: Generate the summary text "progress report" as machine speech "progress report."

[0650] 12. The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0651] User Feedback

[0652] 13. Users can understand the summary audio received through bone conduction and efficiently grasp the content of the conversation.

[0653] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[0654] This system allows users to not only efficiently receive speech recognition and summarized information, but also receive optimal feedback according to their emotional state, enabling them to quickly and accurately understand the content of conversations and achieve smooth communication.

[0655] The processing flow will be explained below.

[0656] Step 1:

[0657] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[0658] Step 2:

[0659] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[0660] Step 3:

[0661] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are preset.

[0662] Step 4:

[0663] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[0664] Step 5:

[0665] The server activates a speech recognition engine to convert the speech data into text data, which then analyzes the speech and generates corresponding text.

[0666] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[0667] Step 6:

[0668] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[0669] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[0670] Step 7:

[0671] The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it on to the emotion engine.

[0672] Step 8:

[0673] The emotion engine analyzes the collected biometric data to recognize the user's current emotional state, which can be classified as excited, relaxed, anxious, etc.

[0674] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[0675] Step 9:

[0676] The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[0677] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[0678] Step 10:

[0679] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0680] Step 11:

[0681] The device passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0682] Example: Generate the summary text "progress report" as machine speech "progress report."

[0683] Step 12:

[0684] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0685] Step 13:

[0686] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[0687] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[0688] Example 2

[0689] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0690] Conventional speech recognition systems only convert speech data into text, and have the problem of being unable to provide information that takes into account the user's emotional state. Furthermore, the collection of speech data, summary generation, emotion recognition, and speech synthesis are all performed separately, which reduces the overall efficiency of the system. Therefore, there is a need for an integrated system that can reduce the burden on users and provide information quickly and accurately.

[0691] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a processor that converts voice data into text, a generative model that summarizes the text data, an emotion recognition means that analyzes the user's emotional state, and a voice synthesis means that converts the summarized text data into voice data. This makes it possible to perform everything from voice collection to emotion analysis, summary generation, and voice synthesis in an integrated manner.

[0692] A "device for collecting audio data" is a device that has the function of collecting surrounding audio in real time and storing it in a buffer.

[0693] A "communication means" is a device that has the function of compressing collected voice data and sending it to a server using Bluetooth, Wi-Fi, etc.

[0694] "Processor" means a central processing unit that executes speech recognition algorithms within the server to convert speech data into text data.

[0695] A "generative model" is an artificial intelligence model that analyzes text data, extracts important content, and generates a summary.

[0696] The "emotion recognition means" is a device or algorithm for analyzing biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[0697] The "speech synthesis means" is a synthesis device that has the function of converting summarized text data into natural speech.

[0698] An "acoustic device" is a device that has the function of transmitting audio data to a user through bone conduction.

[0699] A "wearable device" is a portable device that can be worn by a user and that collects and communicates audio data.

[0700] This invention is a system that integrates voice recognition, generative AI, emotion recognition, voice synthesis, and voice transmission via bone conduction, and provides efficient information provision using a wearable device worn by the user.

[0701] The system includes the following main components:

[0702] 1. Device for collecting voice data: A microphone built into the wearable device collects the user's voice and surrounding voices in real time and stores the data in a buffer.

[0703] 2. Communication method: The collected voice data is converted into digital form, compressed if necessary, and then sent to the server via Bluetooth, Wi-Fi, etc.

[0704] 3. Processor: The central processing unit installed in the server converts the received voice data into text data using a voice recognition algorithm (e.g., Google Cloud Speech-to-Text API).

[0705] 4. Generative models: These include artificial intelligence models (e.g., GPT-4) that analyze text data, extract key content, and generate summaries. Generative models use prompts as input and generate corresponding summary text.

[0706] Example prompt: "I'll report on the progress of my next project."

[0707] 5. Emotion recognition means: Sensors built into smart glasses and other wearable devices collect biometric data such as the user's voice, facial expressions, and pulse rate, and then use emotion recognition algorithms to analyze the user's emotional state.

[0708] Example: Judging whether a user is "excited" based on their tone of voice and facial expression.

[0709] 6. Speech synthesis means: Includes a speech synthesis engine (e.g., Amazon Polly) for converting the summarized text data into natural-sounding speech, which then converts the summarized text into speech data and transmits it to the user.

[0710] Example: Generate the summary text "progress report" as machine speech "progress report."

[0711] 7. Acoustic Device: Includes devices that utilize bone conduction technology to transmit synthesized audio data directly to the user's inner ear through the skull, allowing the user to receive audio information without blocking their ears.

[0712] This system allows users to benefit from:

[0713] Not only can you receive voice recognition and summarized information efficiently, but you can also receive optimal feedback based on your emotional state.

[0714] The entire process, from collecting voice data to analyzing, summarizing, synthesizing voice, and finally transmitting it, is managed centrally, ensuring that information is provided quickly and accurately.

[0715] This system is particularly useful in situations where large amounts of information need to be efficiently processed and transmitted in real time, such as business meetings and educational settings. The present invention allows users to achieve smooth communication and information comprehension.

[0716] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0717] Step 1:

[0718] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[0719] Input: Ambient audio

[0720] Output: Buffered audio data

[0721] Specific operation: The microphone of the wearable device (such as smart glasses) is activated and collects the user's speech and surrounding sounds. The collected data is temporarily stored in a buffer in memory.

[0722] Step 2:

[0723] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[0724] Input: Buffered audio data

[0725] Output: Digital audio data (compressed)

[0726] Specific operation: A processor operates to convert analog audio data into digital, and then a codec is applied to compress the audio data.

[0727] Step 3:

[0728] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[0729] Input: Digital audio data (compressed)

[0730] Output: Audio data sent to the server

[0731] What happens: Compressed audio data is sent via Wi-Fi or Bluetooth modules and reaches the specified API endpoint on the server.

[0732] Step 4:

[0733] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[0734] Input: Audio data sent to the server

[0735] Output: Input file for speech recognition engine

[0736] Specific operation: The voice data is stored in a specific directory on the server, and the voice recognition engine is configured to refer to this file.

[0737] Step 5:

[0738] The server activates a speech recognition engine to convert the speech data into text data. The speech recognition algorithm analyzes the speech data and generates corresponding text.

[0739] Input: Input file for the speech recognition engine

[0740] Output: Text data

[0741] Specific operation: The process of converting speech to text is carried out by calling the Google Cloud Speech-to-Text API, etc. For example, speech saying "I will report on the progress of the next project" is converted to text saying "I will report on the progress of the next project."

[0742] Step 6:

[0743] The server launches the generative model and receives text data from the speech recognition engine. The generative AI analyzes the text data, extracts important content, and generates a summary.

[0744] Input: Text data

[0745] Output: Summary text

[0746] Specific operation: A GPT model (e.g., GPT-4) analyzes the text data "I will report on the progress of the next project" and summarizes it as "Project progress report."

[0747] Step 7:

[0748] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate and passes it to an emotion recognition means.

[0749] Input: User biometric data (voice, facial expression, pulse, etc.)

[0750] Output: Emotional state data

[0751] How it works: Sensors built into smart glasses or wearable devices collect biometric data from users and send it to emotion recognition algorithms.

[0752] Step 8:

[0753] An emotion recognition means analyzes the collected data and recognizes the user's current emotional state (eg, excited, relaxed, anxious, etc.).

[0754] Input: Biometric data

[0755] Output: Emotional state data

[0756] Specific operation: The emotion recognition algorithm analyzes the tone of voice and facial expressions to determine the user's emotional state, such as "excited" or "relaxed."

[0757] Step 9:

[0758] The server adjusts the content of the generated summary text based on the analysis results of the emotion recognition means, making the summary more concise if the user is emotionally excited.

[0759] Input: Summary text, emotional state data

[0760] Output: Adjusted summary text

[0761] Specific operation: The server adjusts the generated summary "Project Progress Report" to be more concise as "Progress Report" based on the emotional state data.

[0762] Step 10:

[0763] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0764] Input: Adjusted summary text

[0765] Output: Summary text sent to terminal

[0766] What it does: The adjusted summary text is sent to your device using Bluetooth or Wi-Fi.

[0767] Step 11:

[0768] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0769] Input: Adjusted summary text

[0770] Output: Synthesized speech data

[0771] What happens: A speech synthesis engine such as Amazon Polly converts the text "progress report" into machine-generated speech.

[0772] Step 12:

[0773] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0774] Input: Synthetic speech data

[0775] Output: Transmission of audio information via bone conduction

[0776] Specific operation: The synthesized voice data is transmitted directly to the user's inner ear through the bone conduction module, and the user hears the voice.

[0777] Step 13:

[0778] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[0779] Input: Audio information via bone conduction

[0780] Output: Understood conversation

[0781] Specific operation: The user receives a summary audio such as "progress report" via bone conduction and quickly grasps the main points of the talk.

[0782] The above are the specific steps of the program processing of this system, which allows users to efficiently enjoy speech recognition, summary generation, emotion recognition, speech synthesis, and bone conduction communication.

[0783] (Application example 2)

[0784] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0785] Security operations require accurate and rapid understanding of on-site situations in real time and the most appropriate course of action. However, conventional methods are prone to information delays and miscommunication, making it difficult to respond immediately based on the urgency of the situation and the emotional state of the guards. To solve this problem, a system is needed that summarizes the on-site situation in real time and provides information that is appropriately adjusted according to the emotional state of the guards.

[0786] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a voice recognition engine means for converting voice data into text, a generation AI means for summarizing the voice recognition results, and an emotion engine means for collecting user emotion data and adjusting the summarized text data based on the emotion data. This makes it possible to immediately grasp the situation on-site and provide optimal information according to the user's emotional state.

[0787] A "voice recognition means" is a device or method that collects voice data in real time and converts it into digital voice data.

[0788] "Communication means" refers to a means for transmitting collected voice data to a server, and includes wireless communication technologies such as Bluetooth and Wi-Fi.

[0789] "Speech recognition engine means" refers to software or algorithms that analyze voice data and convert it into corresponding text data.

[0790] "Generative AI means" refers to means that use artificial intelligence technology to analyze converted text data, extract important content, and generate a summary.

[0791] The "speech synthesis means" refers to a means for converting summarized text data into speech data, and includes an algorithm for generating natural-sounding speech.

[0792] "Bone conduction means" refers to devices or technologies that transmit sound directly through the wearer's skull to the inner ear.

[0793] "Emotion engine means" refers to technology or devices that analyze biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[0794] The present invention is a security system that includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generative AI unit, a voice synthesis unit, a bone conduction unit, and an emotion engine unit. Specifically, the system uses a wearable device worn by a security guard. With this system, the security guard reports the situation on-site by voice, and the voice data is sent to a server for summarization and emotion analysis. Appropriately adjusted information is then provided to the security guard in real time via bone conduction.

[0795] The device activates its built-in microphone and collects audio from the site in real time. This audio data is temporarily stored in a buffer and converted into digital audio data. The compressed audio data is then sent to a server via communication methods such as Bluetooth or Wi-Fi.

[0796] The server converts the received voice data into text data using a voice recognition engine means. For example, voice data reported by a security guard that "a suspicious person has broken in" is converted into text data. Next, a generation AI means analyzes the text data, extracts important content, and generates a summary. For example, "a suspicious person has broken in" is summarized as "suspicious person has broken in." After that, an emotion engine means recognizes the user's emotional state from their voice and facial expression, and adjusts the summary text based on the emotion data. For example, if the user is in a tense state, the summary text is further simplified to "Emergency! Suspicious person has broken in."

[0797] The resulting adjusted summary text data is converted into natural-sounding speech using a speech synthesis device. This speech data is transmitted to the security guard via bone conduction. The security guard receives the speech information via bone conduction, allowing them to quickly grasp important information on the scene.

[0798] For example, when a security guard reports that "a suspicious person has entered the entrance," the system summarizes the voice data as "Suspicious person has entered," and after sensing the guard's state of tension, further simplifies it to "Emergency! Suspicious person has entered," and transmits it via bone conduction. In this way, the security guard can instantly take the most appropriate action depending on the situation.

[0799] Example prompt sentence:

[0800] "A suspicious person entered the entrance" Summarize the situation in one sentence: "A suspicious person entered the entrance"

[0801] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0802] Step 1:

[0803] The terminal activates the built-in microphone and collects audio from the scene in real time. The input is real-time audio data, and the output is digital audio data temporarily stored in a buffer.

[0804] Step 2:

[0805] The terminal converts the buffered audio data into a digital format and compresses it if necessary. The input is the audio data in the buffer, and the output is compressed digital audio data.

[0806] Step 3:

[0807] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The input is compressed digital audio data, and the output is audio data sent to the server.

[0808] Step 4:

[0809] The server stores the received voice data in a specific directory and prepares to pass it to the voice recognition engine means. The input is the voice data sent to the server, and the output is the voice data for the recognition engine to process.

[0810] Step 5:

[0811] The server converts the voice data into text data using a voice recognition engine means, where the input is the received voice data and the output is the converted text data.

[0812] Step 6:

[0813] The server uses generative AI methods to analyze the text data, extract important content, and generate a summary. The input is the converted text data, and the output is the summarized text data.

[0814] Step 7:

[0815] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate, and passes it to the emotion engine means. The input is the user's biometric data, and the output is data representing the user's emotional state.

[0816] Step 8:

[0817] The server uses an emotion engine means to analyze the collected data and recognize the user's current emotional state, where the input is the user's biometric data and the output is the emotional state data.

[0818] Step 9:

[0819] The server adjusts the content of the generated summary text based on the emotional state data. The input is the emotional state data and the summary text data, and the output is the adjusted summary text data.

[0820] Step 10:

[0821] The server compresses the adjusted summary text data and sends it to the terminal, where the input is the adjusted summary text data and the output is the compressed text data sent to the terminal.

[0822] Step 11:

[0823] The terminal passes the received summary text to a speech synthesis engine means for converting it into synthesized speech data, with the adjusted summary text data as input and the synthesized speech data as output.

[0824] Step 12:

[0825] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user. The input is the synthesized voice data, and the output is the transmission of voice through the bone conduction module.

[0826] Step 13:

[0827] The user understands the summary speech received through bone conduction and efficiently grasps the content of the conversation and the situation on the scene. The input is the speech data received through bone conduction, and the output is the understood information.

[0828] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0829] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0830] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0831] [Third embodiment]

[0832] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0833] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0834] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0835] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0836] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0837] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0838] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0839] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0840] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0841] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0842] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0843] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0844] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device, such as glasses, worn by a user. The program processing of this system is specifically described below.

[0845] Collection and transmission of voice data

[0846] 1. The device activates the built-in microphone and collects surrounding audio in real time.

[0847] 2. The device stores the collected voice data as digital data.

[0848] 3. The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi.

[0849] Speech-to-text conversion and summary generation

[0850] 4. The server records the received audio data.

[0851] 5. The server starts the speech recognition engine and converts the voice data into text data.

[0852] Example: The server analyzes user A's utterance "I will report on the progress of the next project" and converts it into the text "I will report on the progress of the next project."

[0853] 6. The server uses generative AI means to summarize the converted text data.

[0854] Example: Summarize the text "I will report on the progress of the next project" as "Project progress report."

[0855] Sending summary data and synthesizing speech

[0856] 7. The server compresses the generated summary text and sends it to the terminal.

[0857] 8. The terminal runs the summary text received through a speech synthesis engine and converts it into voice data.

[0858] Example: The device outputs the summary text "Project progress report" as synthesized speech.

[0859] Sound transmission through bone conduction

[0860] 9. The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[0861] Example: A user receives audio summaries such as "Project Status Report" via bone conduction without using their ears.

[0862] User Feedback

[0863] 10. Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[0864] Example: User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next conversation.

[0865] Addition of multilingual translation function

[0866] Furthermore, this system can be equipped with a multilingual translation function, specifically, a function to translate conversations in different languages ​​in real time and to convey the translation results to the user in a summarized form.

[0867] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[0868] The processing flow will be explained below.

[0869] Step 1:

[0870] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[0871] Step 2:

[0872] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[0873] Step 3:

[0874] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[0875] Step 4:

[0876] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[0877] Step 5:

[0878] The server invokes a speech recognition engine to convert the voice data into text data, where a speech recognition algorithm analyzes the voice data and generates corresponding text.

[0879] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[0880] Step 6:

[0881] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[0882] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[0883] Step 7:

[0884] The server compresses the generated summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[0885] Step 8:

[0886] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[0887] Example: Generate the summary text "Project progress report" as machine speech "Project progress report."

[0888] Step 9:

[0889] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[0890] Step 10:

[0891] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[0892] Example: User B receives a summary audio such as "Project Progress Report" and quickly grasps the main points of User A's speech.

[0893] Example 1

[0894] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0895] Conventional voice data collection and analysis systems are unable to collect, transmit, and analyze voice data in real time quickly enough, making it difficult for users to easily grasp information. Furthermore, when it comes to multilingual support, translation accuracy and real-time performance have been issues. The present invention aims to solve these problems.

[0896] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0897] In this invention, the server includes means for activating and collecting the built-in microphone in real time, means for saving the voice data as digital data, means for compressing the voice data, means for recording the received voice data, and means for compressing and transmitting the generated summary text. This enables the collection, analysis, summarization, and transmission of voice data in real time, and allows for rapid response to voice data in multiple languages.

[0898] "Speech recognition means" refers to a device or technology that collects voice data and captures the voice as digital data.

[0899] "Communication means" refers to the technology or protocol used to transmit collected audio data to the server, and generally includes wireless communication such as Bluetooth or Wi-Fi.

[0900] "Speech recognition engine means" refers to software or algorithms that analyze received voice data and convert it into text data.

[0901] "Generative AI methods" refer to artificial intelligence techniques that analyze and summarize text data, in particular using generative models.

[0902] "Speech synthesis means" refers to software or technology that converts text data into speech data.

[0903] "Bone conduction means" is a technology that transmits audio data as vibrations through the bones to deliver audio information to the user.

[0904] "Means for activating and collecting audio from a built-in microphone in real time" refers to a technology that activates a microphone built into a device and collects audio on the spot in real time.

[0905] "Means for storing audio data as digital data" refers to technology or devices that store collected audio data in digital form.

[0906] "Means for compressing audio data" refers to a compression algorithm or technique for reducing the file size of audio data.

[0907] "Means for recording received voice data" refers to the technology that allows the server to store the received voice data in a database so that it can be accessed later.

[0908] "Means for compressing and transmitting generated summary text" means a technology or algorithm for compressing the summary text generated by the generating AI means and transmitting it to the terminal.

[0909] A "wearable device" is a portable electronic device that can be worn by a user.

[0910] The present invention relates to a system including a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device such as glasses worn by a user.

[0911] Collection and transmission of voice data

[0912] First, the device activates its built-in microphone to collect surrounding sounds in real time. For example, if the user is in a meeting, the device's microphone will record the speech of the meeting participants in real time. This collected audio data is then stored internally as digital data.

[0913] The device then compresses the collected audio data and sends it to a server via Bluetooth or Wi-Fi, for example in AAC format.

[0914] Speech-to-text conversion and summary generation

[0915] The server records the received voice data and invokes a speech recognition engine such as the Google Cloud Speech-to-Text API to convert the voice data into text data. For example, User A's utterance "I will report on the progress of the next project" is converted into text. The server then summarizes this text data using OpenAI's generative AI method. Specifically, the text "I will report on the progress of the next project" is summarized as "Project progress report."

[0916] Sending summary data and synthesizing speech

[0917] The server compresses the summarized text data using a compression algorithm such as GZIP and then transmits it back to the device via Bluetooth or Wi-Fi. The device then converts the received summary text into voice data using a speech synthesis engine such as Amazon Polly.

[0918] Sound transmission through bone conduction

[0919] The device then inputs the synthesized voice data into the bone conduction module and transmits it to the user. For example, the user can receive a summary of a "project progress report" without using their ears through bone conduction technology, which transmits sound information via the user's jawbone and skull.

[0920] User Feedback

[0921] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation. Specifically, User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next part of the conversation.

[0922] Addition of multilingual translation function

[0923] The system can also be equipped with a multilingual translation function, for example, to translate conversations in different languages ​​in real time and provide a summary of the translation results to the user.

[0924] Examples of prompt statements

[0925] User says: "I'll report on the progress of my next project."

[0926] Prompt for generative AI model: "Summarize the following text: Report on the progress of the following project."

[0927] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[0928] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0929] Step 1:

[0930] The device activates the built-in microphone and collects surrounding audio in real time.

[0931] Input: Ambient audio.

[0932] Output: Collected audio data.

[0933] Specific operation: The glasses-type wearable device detects the user's voice input and activates the microphone. For example, the device records what is being said during a meeting.

[0934] Step 2:

[0935] The voice data collected by the device is stored as digital data.

[0936] Input: Collected audio data.

[0937] Output: Audio data in digital format.

[0938] Specific operation: The audio data is saved in the device's internal memory in MP3 format. For example, the collected audio data is temporarily saved in a buffer.

[0939] Step 3:

[0940] The device compresses the audio data it collects and sends it to the server via Bluetooth or Wi-Fi.

[0941] Input: Audio data in digital format.

[0942] Output: Compressed audio data, notification of completion of transmission to the server.

[0943] Specific operation: Using a compression algorithm such as AAC, the compressed audio data is sent to a server via Bluetooth or Wi-Fi. For example, the data is sent via a home Wi-Fi network.

[0944] Step 4:

[0945] The server records the received audio data.

[0946] Input: Compressed audio data.

[0947] Output: Recorded audio data.

[0948] Specific operation: The received voice data is stored in a database and managed with a timestamp and identification information. For example, the data is recorded in a relational database on the cloud.

[0949] Step 5:

[0950] The server activates a voice recognition engine and converts the voice data into text data.

[0951] Input: Recorded audio data.

[0952] Output: The converted text data.

[0953] Specific operation: The voice data is analyzed using the Google Cloud Speech-to-Text API and converted into text data. For example, the speech of User A is converted into the text "I will report on the progress of the next project."

[0954] Step 6:

[0955] The server uses generative AI means to summarize the converted text data.

[0956] Input: The converted text data.

[0957] Output: Summarized text data.

[0958] Specific operation: Send a prompt to a generative AI model (e.g., GPT-4) and obtain a summarized text. For example, input the prompt "Summarize the following text: Report on the progress of the following project" and obtain the summary "Project progress report."

[0959] Step 7:

[0960] The server compresses the generated summary text and transmits it to the terminal.

[0961] Input: Abstracted text data.

[0962] Output: Compressed summary text data.

[0963] Specific operation: Compress the summary text using the GZIP algorithm and send it to the terminal via the Wi-Fi module. For example, send the summary text through your home Wi-Fi network.

[0964] Step 8:

[0965] The summary text received by the terminal is passed through a speech synthesis engine and converted into voice data.

[0966] Input: Received summary text data.

[0967] Output: Audio data.

[0968] Specific operation: Calls a speech synthesis engine such as Amazon Polly and generates synthetic speech data from text. For example, it outputs the received text "Project progress report" as synthetic speech.

[0969] Step 9:

[0970] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[0971] Input: Synthetic speech data.

[0972] Output: Audio information via bone conduction.

[0973] Specific operation: The synthesized voice data is sent to the bone conduction module, and the voice is transmitted to the user via the jawbone or skull. For example, the synthesized voice is transmitted as vibrations through the jawbone.

[0974] Step 10:

[0975] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[0976] Input: Audio information via bone conduction.

[0977] Output: Understanding, next steps in the conversation.

[0978] Specific actions: The user understands the audio from the bone conduction speaker and considers the next statement or action based on the content. For example, the user understands a "project progress report" and performs the action of providing the necessary feedback.

[0979] (Application example 1)

[0980] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0981] Modern manufacturing demands improved work efficiency and safety. Collaboration between workers and robots within factories is particularly important, but language barriers and communication delays pose challenges. Current systems make it difficult for workers who speak different languages ​​to communicate with each other, reducing the efficiency of instruction transmission. Furthermore, as communication methods utilizing bone conduction technology have not yet been fully established, there is a need for intuitive and rapid instruction transmission.

[0982] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0983] In this invention, the server includes a speech recognition engine means for converting voice data into text, a generation AI means, and a speech synthesis means, which enable a voice instruction means for giving instructions to factory automation equipment using a visual support device worn by a worker.

[0984] A "voice recognition means" is a device or component that collects voice data and stores it as digital data.

[0985] "Communication means" refers to the technology and protocols used to transmit collected audio data to the server, including Bluetooth and Wi-Fi.

[0986] "Speech recognition engine means" refers to software or algorithms that run on the server and convert received voice data into text.

[0987] "Generative AI means" refers to artificial intelligence models or algorithms that analyze the converted text data and generate summaries or instructions.

[0988] "Speech synthesis means" refers to software or technology for converting the generated summary text data into speech data.

[0989] "Bone conduction means" refers to a device or module for transmitting audio data to a wearer using bone conduction technology.

[0990] A "visual support device" is a wearable device worn by a worker that provides visual and audio information. Specific examples include smart glasses.

[0991] "Factory automation equipment" refers to automated equipment and robots used in factories.

[0992] "Voice instruction means" refers to a system or technology that allows a worker to give instructions to automated factory equipment through voice using a visual assistance device.

[0993] The present invention relates to a voice guidance system for automated factory equipment using a visual assistance device worn by an operator. The system includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, a bone conduction unit, and a voice guidance unit.

[0994] Voice recognition means

[0995] First, the built-in microphone of the visual assistance device (e.g., smart glasses) is activated and collects the worker's voice in real time. The collected voice data is saved as digital data, allowing the worker's voice to be accurately captured.

[0996] communication means

[0997] The communication device then compresses the collected voice data and transmits it to a server via Bluetooth or Wi-Fi, efficiently transmitting large amounts of voice data.

[0998] Speech Recognition Engine Means

[0999] The server records the received voice data and activates a speech recognition engine. Specifically, it uses Google's speech recognition API or similar technology to convert the voice data into text data. For example, instructions such as "Please prepare to assemble the next part" are converted into text.

[1000] Generation AI means

[1001] The server inputs the converted text data into a generative AI means and generates a summary using a generative AI model (e.g., GPT-3). For example, the text "Please prepare to assemble the following parts" is summarized as "Prepare to assemble parts."

[1002] Voice synthesis means

[1003] The server sends the summarized text data to the visual support device, which then converts the summarized text into speech data using a speech synthesis engine (e.g., pyttsx3), generating a summary instruction as speech.

[1004] Bone conduction means

[1005] The visual assistance device inputs the generated voice data into the bone conduction module and transmits it to the worker, who can receive summary voice such as "Prepare for parts assembly" via bone conduction without using their ears.

[1006] Audio instruction tools

[1007] This allows workers wearing visual assistance devices to receive voice instructions and provide accurate and prompt voice guidance to automated factory equipment, which can, for example, facilitate smooth communication between workers who speak different languages ​​and improve the efficiency of instruction transmission.

[1008] Specific examples

[1009] Say: "Get ready to assemble the next part."

[1010] Generative AI summary: "Parts ready for assembly"

[1011] Bone conduction feedback: "Preparing to assemble parts"

[1012] Example prompts to input to a generative AI model:

[1013] Get ready to assemble the next part

[1014] In this invention, the voice recognition means, generation AI means, bone conduction means, etc. cooperate with each other at each step to provide optimal voice guidance to workers, which is expected to significantly improve work efficiency and safety in factories.

[1015] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1016] Step 1:

[1017] The terminal activates the built-in microphone and collects the worker's voice in real time. The collected voice data is saved as digital data. Here, the input is the worker's voice and the output is digital voice data. In this step, data conversion is performed to save it as digital voice data.

[1018] Step 2:

[1019] The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi. Here, the input is digital audio data and the output is compressed audio data. In this step, a compression algorithm is used to reduce the data size and the data is sent to the server using a communication protocol.

[1020] Step 3:

[1021] The server records the received voice data and starts the voice recognition engine. Here, the input is compressed voice data and the output is text data. In this step, the voice recognition process is performed to convert the voice data into text. Specifically, Google's voice recognition API is used.

[1022] Step 4:

[1023] The server uses a generative AI method to summarize the converted text data, where the input is text data and the output is a summary text. In this step, a generative AI model (e.g., GPT-3) analyzes the input text and generates a summary.

[1024] Step 5:

[1025] The server compresses the generated summary text and sends it to the terminal. Here, the input is the summary text and the output is the compressed summary text. In this step, a compression algorithm is used to reduce the data size and the data is sent to the terminal using a communication protocol.

[1026] Step 6:

[1027] The terminal converts the received summary text into speech data through a speech synthesis engine. The input is the summary text and the output is speech data. Specifically, the text is converted into speech using the pyttsx3 engine.

[1028] Step 7:

[1029] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the worker. Here, the input is voice data, and the output is the voice that the worker receives through bone conduction. In this step, the voice data is transmitted without using the ears using bone conduction technology.

[1030] Step 8:

[1031] The worker understands the summarized audio received through bone conduction technology and gives instructions to the automated factory equipment. Here, the input is the audio via bone conduction, and the output is audio instructions to the automated factory equipment. In this step, the worker performs the actual work based on the audio instructions.

[1032] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1033] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, a bone conduction unit, and an emotion engine. This system can be implemented using a wearable device such as glasses worn by a user. Specific program processing of this system is described in detail below.

[1034] Collection and transmission of voice data

[1035] 1. The device activates the built-in microphone and collects surrounding audio in real time. The collected audio data is temporarily stored in a buffer.

[1036] 2. The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing if necessary.

[1037] 3. The device sends the compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are pre-set.

[1038] Speech-to-text conversion and summary generation

[1039] 4. The server saves the received voice data in a specific directory and prepares to pass it to the voice recognition engine.

[1040] 5. The server invokes the speech recognition engine to convert the audio data to text data. Here, the speech recognition algorithm analyzes the audio data and generates corresponding text.

[1041] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[1042] 6. The server starts the generation AI means and receives the text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[1043] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[1044] Recognizing user emotions with an emotion engine

[1045] 7. The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it to the emotion engine.

[1046] 8. The emotion engine analyzes the collected data and recognizes the user's current emotional state (e.g., excited, relaxed, anxious, etc.).

[1047] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[1048] Emotion regulation and transmission of summary data

[1049] 9. The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[1050] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[1051] 10. The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[1052] Speech synthesis of summary text and transmission via bone conduction

[1053] 11. The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[1054] Example: Generate the summary text "progress report" as machine speech "progress report."

[1055] 12. The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[1056] User Feedback

[1057] 13. Users can understand the summary audio received through bone conduction and efficiently grasp the content of the conversation.

[1058] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[1059] This system allows users to not only efficiently receive speech recognition and summarized information, but also receive optimal feedback according to their emotional state, enabling them to quickly and accurately understand the content of conversations and achieve smooth communication.

[1060] The processing flow will be explained below.

[1061] Step 1:

[1062] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[1063] Step 2:

[1064] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[1065] Step 3:

[1066] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are preset.

[1067] Step 4:

[1068] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[1069] Step 5:

[1070] The server activates a speech recognition engine to convert the speech data into text data, which then analyzes the speech and generates corresponding text.

[1071] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[1072] Step 6:

[1073] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[1074] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[1075] Step 7:

[1076] The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it on to the emotion engine.

[1077] Step 8:

[1078] The emotion engine analyzes the collected biometric data to recognize the user's current emotional state, which can be classified as excited, relaxed, anxious, etc.

[1079] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[1080] Step 9:

[1081] The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[1082] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[1083] Step 10:

[1084] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[1085] Step 11:

[1086] The device passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[1087] Example: Generate the summary text "progress report" as machine speech "progress report."

[1088] Step 12:

[1089] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[1090] Step 13:

[1091] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[1092] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[1093] Example 2

[1094] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1095] Conventional speech recognition systems only convert speech data into text, and have the problem of being unable to provide information that takes into account the user's emotional state. Furthermore, the collection of speech data, summary generation, emotion recognition, and speech synthesis are all performed separately, which reduces the overall efficiency of the system. Therefore, there is a need for an integrated system that can reduce the burden on users and provide information quickly and accurately.

[1096] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a processor that converts voice data into text, a generative model that summarizes the text data, an emotion recognition means that analyzes the user's emotional state, and a voice synthesis means that converts the summarized text data into voice data. This makes it possible to perform everything from voice collection to emotion analysis, summary generation, and voice synthesis in an integrated manner.

[1097] A "device for collecting audio data" is a device that has the function of collecting surrounding audio in real time and storing it in a buffer.

[1098] A "communication means" is a device that has the function of compressing collected voice data and sending it to a server using Bluetooth, Wi-Fi, etc.

[1099] "Processor" means a central processing unit that executes speech recognition algorithms within the server to convert speech data into text data.

[1100] A "generative model" is an artificial intelligence model that analyzes text data, extracts important content, and generates a summary.

[1101] The "emotion recognition means" is a device or algorithm for analyzing biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[1102] The "speech synthesis means" is a synthesis device that has the function of converting summarized text data into natural speech.

[1103] An "acoustic device" is a device that has the function of transmitting audio data to a user through bone conduction.

[1104] A "wearable device" is a portable device that can be worn by a user and that collects and communicates audio data.

[1105] This invention is a system that integrates voice recognition, generative AI, emotion recognition, voice synthesis, and voice transmission via bone conduction, and provides efficient information provision using a wearable device worn by the user.

[1106] The system includes the following main components:

[1107] 1. Device for collecting voice data: A microphone built into the wearable device collects the user's voice and surrounding voices in real time and stores the data in a buffer.

[1108] 2. Communication method: The collected voice data is converted into digital form, compressed if necessary, and then sent to the server via Bluetooth, Wi-Fi, etc.

[1109] 3. Processor: The central processing unit installed in the server converts the received voice data into text data using a voice recognition algorithm (e.g., Google Cloud Speech-to-Text API).

[1110] 4. Generative models: These include artificial intelligence models (e.g., GPT-4) that analyze text data, extract key content, and generate summaries. Generative models use prompts as input and generate corresponding summary text.

[1111] Example prompt: "I'll report on the progress of my next project."

[1112] 5. Emotion recognition means: Sensors built into smart glasses and other wearable devices collect biometric data such as the user's voice, facial expressions, and pulse rate, and then use emotion recognition algorithms to analyze the user's emotional state.

[1113] Example: Judging whether a user is "excited" based on their tone of voice and facial expression.

[1114] 6. Speech synthesis means: Includes a speech synthesis engine (e.g., Amazon Polly) for converting the summarized text data into natural-sounding speech, which then converts the summarized text into speech data and transmits it to the user.

[1115] Example: Generate the summary text "progress report" as machine speech "progress report."

[1116] 7. Acoustic Device: Includes devices that utilize bone conduction technology to transmit synthesized audio data directly to the user's inner ear through the skull, allowing the user to receive audio information without blocking their ears.

[1117] This system allows users to benefit from:

[1118] Not only can you receive voice recognition and summarized information efficiently, but you can also receive optimal feedback based on your emotional state.

[1119] The entire process, from collecting voice data to analyzing, summarizing, synthesizing voice, and finally transmitting it, is managed centrally, ensuring that information is provided quickly and accurately.

[1120] This system is particularly useful in situations where large amounts of information need to be efficiently processed and transmitted in real time, such as business meetings and educational settings. The present invention allows users to achieve smooth communication and information comprehension.

[1121] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1122] Step 1:

[1123] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[1124] Input: Ambient audio

[1125] Output: Buffered audio data

[1126] Specific operation: The microphone of the wearable device (such as smart glasses) is activated and collects the user's speech and surrounding sounds. The collected data is temporarily stored in a buffer in memory.

[1127] Step 2:

[1128] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[1129] Input: Buffered audio data

[1130] Output: Digital audio data (compressed)

[1131] Specific operation: A processor operates to convert analog audio data into digital, and then a codec is applied to compress the audio data.

[1132] Step 3:

[1133] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[1134] Input: Digital audio data (compressed)

[1135] Output: Audio data sent to the server

[1136] What happens: Compressed audio data is sent via Wi-Fi or Bluetooth modules and reaches the specified API endpoint on the server.

[1137] Step 4:

[1138] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[1139] Input: Audio data sent to the server

[1140] Output: Input file for speech recognition engine

[1141] Specific operation: The voice data is stored in a specific directory on the server, and the voice recognition engine is configured to refer to this file.

[1142] Step 5:

[1143] The server activates a speech recognition engine to convert the speech data into text data. The speech recognition algorithm analyzes the speech data and generates corresponding text.

[1144] Input: Input file for the speech recognition engine

[1145] Output: Text data

[1146] Specific operation: The process of converting speech to text is carried out by calling the Google Cloud Speech-to-Text API, etc. For example, speech saying "I will report on the progress of the next project" is converted to text saying "I will report on the progress of the next project."

[1147] Step 6:

[1148] The server launches the generative model and receives text data from the speech recognition engine. The generative AI analyzes the text data, extracts important content, and generates a summary.

[1149] Input: Text data

[1150] Output: Summary text

[1151] Specific operation: A GPT model (e.g., GPT-4) analyzes the text data "I will report on the progress of the next project" and summarizes it as "Project progress report."

[1152] Step 7:

[1153] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate and passes it to an emotion recognition means.

[1154] Input: User biometric data (voice, facial expression, pulse, etc.)

[1155] Output: Emotional state data

[1156] How it works: Sensors built into smart glasses or wearable devices collect biometric data from users and send it to emotion recognition algorithms.

[1157] Step 8:

[1158] An emotion recognition means analyzes the collected data and recognizes the user's current emotional state (eg, excited, relaxed, anxious, etc.).

[1159] Input: Biometric data

[1160] Output: Emotional state data

[1161] Specific operation: The emotion recognition algorithm analyzes the tone of voice and facial expressions to determine the user's emotional state, such as "excited" or "relaxed."

[1162] Step 9:

[1163] The server adjusts the content of the generated summary text based on the analysis results of the emotion recognition means, making the summary more concise if the user is emotionally excited.

[1164] Input: Summary text, emotional state data

[1165] Output: Adjusted summary text

[1166] Specific operation: The server adjusts the generated summary "Project Progress Report" to be more concise as "Progress Report" based on the emotional state data.

[1167] Step 10:

[1168] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[1169] Input: Adjusted summary text

[1170] Output: Summary text sent to terminal

[1171] What it does: The adjusted summary text is sent to your device using Bluetooth or Wi-Fi.

[1172] Step 11:

[1173] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[1174] Input: Adjusted summary text

[1175] Output: Synthesized speech data

[1176] What happens: A speech synthesis engine such as Amazon Polly converts the text "progress report" into machine-generated speech.

[1177] Step 12:

[1178] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[1179] Input: Synthetic speech data

[1180] Output: Transmission of audio information via bone conduction

[1181] Specific operation: The synthesized voice data is transmitted directly to the user's inner ear through the bone conduction module, and the user hears the voice.

[1182] Step 13:

[1183] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[1184] Input: Audio information via bone conduction

[1185] Output: Understood conversation

[1186] Specific operation: The user receives a summary audio such as "progress report" via bone conduction and quickly grasps the main points of the talk.

[1187] The above are the specific steps of the program processing of this system, which allows users to efficiently enjoy speech recognition, summary generation, emotion recognition, speech synthesis, and bone conduction communication.

[1188] (Application example 2)

[1189] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1190] Security operations require accurate and rapid understanding of on-site situations in real time and the most appropriate course of action. However, conventional methods are prone to information delays and miscommunication, making it difficult to respond immediately based on the urgency of the situation and the emotional state of the guards. To solve this problem, a system is needed that summarizes the on-site situation in real time and provides information that is appropriately adjusted according to the emotional state of the guards.

[1191] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a voice recognition engine means for converting voice data into text, a generation AI means for summarizing the voice recognition results, and an emotion engine means for collecting user emotion data and adjusting the summarized text data based on the emotion data. This makes it possible to immediately grasp the situation on-site and provide optimal information according to the user's emotional state.

[1192] A "voice recognition means" is a device or method that collects voice data in real time and converts it into digital voice data.

[1193] "Communication means" refers to a means for transmitting collected voice data to a server, and includes wireless communication technologies such as Bluetooth and Wi-Fi.

[1194] "Speech recognition engine means" refers to software or algorithms that analyze voice data and convert it into corresponding text data.

[1195] "Generative AI means" refers to means that use artificial intelligence technology to analyze converted text data, extract important content, and generate a summary.

[1196] The "speech synthesis means" refers to a means for converting summarized text data into speech data, and includes an algorithm for generating natural-sounding speech.

[1197] "Bone conduction means" refers to devices or technologies that transmit sound directly through the wearer's skull to the inner ear.

[1198] "Emotion engine means" refers to technology or devices that analyze biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[1199] The present invention is a security system that includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generative AI unit, a voice synthesis unit, a bone conduction unit, and an emotion engine unit. Specifically, the system uses a wearable device worn by a security guard. With this system, the security guard reports the situation on-site by voice, and the voice data is sent to a server for summarization and emotion analysis. Appropriately adjusted information is then provided to the security guard in real time via bone conduction.

[1200] The device activates its built-in microphone and collects audio from the site in real time. This audio data is temporarily stored in a buffer and converted into digital audio data. The compressed audio data is then sent to a server via communication methods such as Bluetooth or Wi-Fi.

[1201] The server converts the received voice data into text data using a voice recognition engine means. For example, voice data reported by a security guard that "a suspicious person has broken in" is converted into text data. Next, a generation AI means analyzes the text data, extracts important content, and generates a summary. For example, "a suspicious person has broken in" is summarized as "suspicious person has broken in." After that, an emotion engine means recognizes the user's emotional state from their voice and facial expression, and adjusts the summary text based on the emotion data. For example, if the user is in a tense state, the summary text is further simplified to "Emergency! Suspicious person has broken in."

[1202] The resulting adjusted summary text data is converted into natural-sounding speech using a speech synthesis device. This speech data is transmitted to the security guard via bone conduction. The security guard receives the speech information via bone conduction, allowing them to quickly grasp important information on the scene.

[1203] For example, when a security guard reports that "a suspicious person has entered the entrance," the system summarizes the voice data as "Suspicious person has entered," and after sensing the guard's state of tension, further simplifies it to "Emergency! Suspicious person has entered," and transmits it via bone conduction. In this way, the security guard can instantly take the most appropriate action depending on the situation.

[1204] Example prompt sentence:

[1205] "A suspicious person entered the entrance" Summarize the situation in one sentence: "A suspicious person entered the entrance"

[1206] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1207] Step 1:

[1208] The terminal activates the built-in microphone and collects audio from the scene in real time. The input is real-time audio data, and the output is digital audio data temporarily stored in a buffer.

[1209] Step 2:

[1210] The terminal converts the buffered audio data into a digital format and compresses it if necessary. The input is the audio data in the buffer, and the output is compressed digital audio data.

[1211] Step 3:

[1212] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The input is compressed digital audio data, and the output is audio data sent to the server.

[1213] Step 4:

[1214] The server stores the received voice data in a specific directory and prepares to pass it to the voice recognition engine means. The input is the voice data sent to the server, and the output is the voice data for the recognition engine to process.

[1215] Step 5:

[1216] The server converts the voice data into text data using a voice recognition engine means, where the input is the received voice data and the output is the converted text data.

[1217] Step 6:

[1218] The server uses generative AI methods to analyze the text data, extract important content, and generate a summary. The input is the converted text data, and the output is the summarized text data.

[1219] Step 7:

[1220] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate, and passes it to the emotion engine means. The input is the user's biometric data, and the output is data representing the user's emotional state.

[1221] Step 8:

[1222] The server uses an emotion engine means to analyze the collected data and recognize the user's current emotional state, where the input is the user's biometric data and the output is the emotional state data.

[1223] Step 9:

[1224] The server adjusts the content of the generated summary text based on the emotional state data. The input is the emotional state data and the summary text data, and the output is the adjusted summary text data.

[1225] Step 10:

[1226] The server compresses the adjusted summary text data and sends it to the terminal, where the input is the adjusted summary text data and the output is the compressed text data sent to the terminal.

[1227] Step 11:

[1228] The terminal passes the received summary text to a speech synthesis engine means for converting it into synthesized speech data, with the adjusted summary text data as input and the synthesized speech data as output.

[1229] Step 12:

[1230] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user. The input is the synthesized voice data, and the output is the transmission of voice through the bone conduction module.

[1231] Step 13:

[1232] The user understands the summary speech received through bone conduction and efficiently grasps the content of the conversation and the situation on the scene. The input is the speech data received through bone conduction, and the output is the understood information.

[1233] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1234] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1235] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1236] [Fourth embodiment]

[1237] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1238] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1239] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1240] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1241] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1242] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1243] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1244] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1245] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1246] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1247] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1248] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1249] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1250] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device, such as glasses, worn by a user. The program processing of this system is specifically described below.

[1251] Collection and transmission of voice data

[1252] 1. The device activates the built-in microphone and collects surrounding audio in real time.

[1253] 2. The device stores the collected voice data as digital data.

[1254] 3. The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi.

[1255] Speech-to-text conversion and summary generation

[1256] 4. The server records the received audio data.

[1257] 5. The server starts the speech recognition engine and converts the voice data into text data.

[1258] Example: The server analyzes user A's utterance "I will report on the progress of the next project" and converts it into the text "I will report on the progress of the next project."

[1259] 6. The server uses generative AI means to summarize the converted text data.

[1260] Example: Summarize the text "I will report on the progress of the next project" as "Project progress report."

[1261] Sending summary data and synthesizing speech

[1262] 7. The server compresses the generated summary text and sends it to the terminal.

[1263] 8. The terminal runs the summary text received through a speech synthesis engine and converts it into voice data.

[1264] Example: The device outputs the summary text "Project progress report" as synthesized speech.

[1265] Sound transmission through bone conduction

[1266] 9. The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[1267] Example: A user receives audio summaries such as "Project Status Report" via bone conduction without using their ears.

[1268] User Feedback

[1269] 10. Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[1270] Example: User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next conversation.

[1271] Addition of multilingual translation function

[1272] Furthermore, this system can be equipped with a multilingual translation function, specifically, a function to translate conversations in different languages ​​in real time and to convey the translation results to the user in a summarized form.

[1273] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[1274] The processing flow will be explained below.

[1275] Step 1:

[1276] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[1277] Step 2:

[1278] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[1279] Step 3:

[1280] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[1281] Step 4:

[1282] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[1283] Step 5:

[1284] The server invokes a speech recognition engine to convert the voice data into text data, where a speech recognition algorithm analyzes the voice data and generates corresponding text.

[1285] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[1286] Step 6:

[1287] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[1288] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[1289] Step 7:

[1290] The server compresses the generated summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[1291] Step 8:

[1292] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[1293] Example: Generate the summary text "Project progress report" as machine speech "Project progress report."

[1294] Step 9:

[1295] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[1296] Step 10:

[1297] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[1298] Example: User B receives a summary audio such as "Project Progress Report" and quickly grasps the main points of User A's speech.

[1299] Example 1

[1300] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1301] Conventional voice data collection and analysis systems are unable to collect, transmit, and analyze voice data in real time quickly enough, making it difficult for users to easily grasp information. Furthermore, when it comes to multilingual support, translation accuracy and real-time performance have been issues. The present invention aims to solve these problems.

[1302] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1303] In this invention, the server includes means for activating and collecting the built-in microphone in real time, means for saving the voice data as digital data, means for compressing the voice data, means for recording the received voice data, and means for compressing and transmitting the generated summary text. This enables the collection, analysis, summarization, and transmission of voice data in real time, and allows for rapid response to voice data in multiple languages.

[1304] "Speech recognition means" refers to a device or technology that collects voice data and captures the voice as digital data.

[1305] "Communication means" refers to the technology or protocol used to transmit collected audio data to the server, and generally includes wireless communication such as Bluetooth or Wi-Fi.

[1306] "Speech recognition engine means" refers to software or algorithms that analyze received voice data and convert it into text data.

[1307] "Generative AI methods" refer to artificial intelligence techniques that analyze and summarize text data, in particular using generative models.

[1308] "Speech synthesis means" refers to software or technology that converts text data into speech data.

[1309] "Bone conduction means" is a technology that transmits audio data as vibrations through the bones to deliver audio information to the user.

[1310] "Means for activating and collecting audio from a built-in microphone in real time" refers to a technology that activates a microphone built into a device and collects audio on the spot in real time.

[1311] "Means for storing audio data as digital data" refers to technology or devices that store collected audio data in digital form.

[1312] "Means for compressing audio data" refers to a compression algorithm or technique for reducing the file size of audio data.

[1313] "Means for recording received voice data" refers to the technology that allows the server to store the received voice data in a database so that it can be accessed later.

[1314] "Means for compressing and transmitting generated summary text" means a technology or algorithm for compressing the summary text generated by the generating AI means and transmitting it to the terminal.

[1315] A "wearable device" is a portable electronic device that can be worn by a user.

[1316] The present invention relates to a system including a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, and a bone conduction unit. This system can be implemented using a wearable device such as glasses worn by a user.

[1317] Collection and transmission of voice data

[1318] First, the device activates its built-in microphone to collect surrounding sounds in real time. For example, if the user is in a meeting, the device's microphone will record the speech of the meeting participants in real time. This collected audio data is then stored internally as digital data.

[1319] The device then compresses the collected audio data and sends it to a server via Bluetooth or Wi-Fi, for example in AAC format.

[1320] Speech-to-text conversion and summary generation

[1321] The server records the received voice data and invokes a speech recognition engine such as the Google Cloud Speech-to-Text API to convert the voice data into text data. For example, User A's utterance "I will report on the progress of the next project" is converted into text. The server then summarizes this text data using OpenAI's generative AI method. Specifically, the text "I will report on the progress of the next project" is summarized as "Project progress report."

[1322] Sending summary data and synthesizing speech

[1323] The server compresses the summarized text data using a compression algorithm such as GZIP and then transmits it back to the device via Bluetooth or Wi-Fi. The device then converts the received summary text into voice data using a speech synthesis engine such as Amazon Polly.

[1324] Sound transmission through bone conduction

[1325] The device then inputs the synthesized voice data into the bone conduction module and transmits it to the user. For example, the user can receive a summary of a "project progress report" without using their ears through bone conduction technology, which transmits sound information via the user's jawbone and skull.

[1326] User Feedback

[1327] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation. Specifically, User B can efficiently understand the main points of User A's speech through bone conduction, allowing them to smoothly move on to the next part of the conversation.

[1328] Addition of multilingual translation function

[1329] The system can also be equipped with a multilingual translation function, for example, to translate conversations in different languages ​​in real time and provide a summary of the translation results to the user.

[1330] Examples of prompt statements

[1331] User says: "I'll report on the progress of my next project."

[1332] Prompt for generative AI model: "Summarize the following text: Report on the progress of the following project."

[1333] In this way, the present invention enables users to quickly understand surrounding conversations and achieve smooth communication through a series of processes that efficiently recognize, summarize, and transmit voice data. In each of the above processing steps, the voice recognition means, generation AI means, bone conduction means, etc. work together to form a system that provides optimal feedback to the user.

[1334] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1335] Step 1:

[1336] The device activates the built-in microphone and collects surrounding audio in real time.

[1337] Input: Ambient audio.

[1338] Output: Collected audio data.

[1339] Specific operation: The glasses-type wearable device detects the user's voice input and activates the microphone. For example, the device records what is being said during a meeting.

[1340] Step 2:

[1341] The voice data collected by the device is stored as digital data.

[1342] Input: Collected audio data.

[1343] Output: Audio data in digital format.

[1344] Specific operation: The audio data is saved in the device's internal memory in MP3 format. For example, the collected audio data is temporarily saved in a buffer.

[1345] Step 3:

[1346] The device compresses the audio data it collects and sends it to the server via Bluetooth or Wi-Fi.

[1347] Input: Audio data in digital format.

[1348] Output: Compressed audio data, notification of completion of transmission to the server.

[1349] Specific operation: Using a compression algorithm such as AAC, the compressed audio data is sent to a server via Bluetooth or Wi-Fi. For example, the data is sent via a home Wi-Fi network.

[1350] Step 4:

[1351] The server records the received audio data.

[1352] Input: Compressed audio data.

[1353] Output: Recorded audio data.

[1354] Specific operation: The received voice data is stored in a database and managed with a timestamp and identification information. For example, the data is recorded in a relational database on the cloud.

[1355] Step 5:

[1356] The server activates a voice recognition engine and converts the voice data into text data.

[1357] Input: Recorded audio data.

[1358] Output: The converted text data.

[1359] Specific operation: The voice data is analyzed using the Google Cloud Speech-to-Text API and converted into text data. For example, the speech of User A is converted into the text "I will report on the progress of the next project."

[1360] Step 6:

[1361] The server uses generative AI means to summarize the converted text data.

[1362] Input: The converted text data.

[1363] Output: Summarized text data.

[1364] Specific operation: Send a prompt to a generative AI model (e.g., GPT-4) and obtain a summarized text. For example, input the prompt "Summarize the following text: Report on the progress of the following project" and obtain the summary "Project progress report."

[1365] Step 7:

[1366] The server compresses the generated summary text and transmits it to the terminal.

[1367] Input: Abstracted text data.

[1368] Output: Compressed summary text data.

[1369] Specific operation: Compress the summary text using the GZIP algorithm and send it to the terminal via the Wi-Fi module. For example, send the summary text through your home Wi-Fi network.

[1370] Step 8:

[1371] The summary text received by the terminal is passed through a speech synthesis engine and converted into voice data.

[1372] Input: Received summary text data.

[1373] Output: Audio data.

[1374] Specific operation: Calls a speech synthesis engine such as Amazon Polly and generates synthetic speech data from text. For example, it outputs the received text "Project progress report" as synthetic speech.

[1375] Step 9:

[1376] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user.

[1377] Input: Synthetic speech data.

[1378] Output: Audio information via bone conduction.

[1379] Specific operation: The synthesized voice data is sent to the bone conduction module, and the voice is transmitted to the user via the jawbone or skull. For example, the synthesized voice is transmitted as vibrations through the jawbone.

[1380] Step 10:

[1381] Users can understand the summary audio received through bone conduction technology and quickly grasp the flow of the conversation.

[1382] Input: Audio information via bone conduction.

[1383] Output: Understanding, next steps in the conversation.

[1384] Specific actions: The user understands the audio from the bone conduction speaker and considers the next statement or action based on the content. For example, the user understands a "project progress report" and performs the action of providing the necessary feedback.

[1385] (Application example 1)

[1386] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1387] Modern manufacturing demands improved work efficiency and safety. Collaboration between workers and robots within factories is particularly important, but language barriers and communication delays pose challenges. Current systems make it difficult for workers who speak different languages ​​to communicate with each other, reducing the efficiency of instruction transmission. Furthermore, as communication methods utilizing bone conduction technology have not yet been fully established, there is a need for intuitive and rapid instruction transmission.

[1388] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1389] In this invention, the server includes a speech recognition engine means for converting voice data into text, a generation AI means, and a speech synthesis means, which enable a voice instruction means for giving instructions to factory automation equipment using a visual support device worn by a worker.

[1390] A "voice recognition means" is a device or component that collects voice data and stores it as digital data.

[1391] "Communication means" refers to the technology and protocols used to transmit collected audio data to the server, including Bluetooth and Wi-Fi.

[1392] "Speech recognition engine means" refers to software or algorithms that run on the server and convert received voice data into text.

[1393] "Generative AI means" refers to artificial intelligence models or algorithms that analyze the converted text data and generate summaries or instructions.

[1394] "Speech synthesis means" refers to software or technology for converting the generated summary text data into speech data.

[1395] "Bone conduction means" refers to a device or module for transmitting audio data to a wearer using bone conduction technology.

[1396] A "visual support device" is a wearable device worn by a worker that provides visual and audio information. Specific examples include smart glasses.

[1397] "Factory automation equipment" refers to automated equipment and robots used in factories.

[1398] "Voice instruction means" refers to a system or technology that allows a worker to give instructions to automated factory equipment through voice using a visual assistance device.

[1399] The present invention relates to a voice guidance system for automated factory equipment using a visual assistance device worn by an operator. The system includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generation AI unit, a voice synthesis unit, a bone conduction unit, and a voice guidance unit.

[1400] Voice recognition means

[1401] First, the built-in microphone of the visual assistance device (e.g., smart glasses) is activated and collects the worker's voice in real time. The collected voice data is saved as digital data, allowing the worker's voice to be accurately captured.

[1402] communication means

[1403] The communication device then compresses the collected voice data and transmits it to a server via Bluetooth or Wi-Fi, efficiently transmitting large amounts of voice data.

[1404] Speech Recognition Engine Means

[1405] The server records the received voice data and activates a speech recognition engine. Specifically, it uses Google's speech recognition API or similar technology to convert the voice data into text data. For example, instructions such as "Please prepare to assemble the next part" are converted into text.

[1406] Generation AI means

[1407] The server inputs the converted text data into a generative AI means and generates a summary using a generative AI model (e.g., GPT-3). For example, the text "Please prepare to assemble the following parts" is summarized as "Prepare to assemble parts."

[1408] Voice synthesis means

[1409] The server sends the summarized text data to the visual support device, which then converts the summarized text into speech data using a speech synthesis engine (e.g., pyttsx3), generating a summary instruction as speech.

[1410] Bone conduction means

[1411] The visual assistance device inputs the generated voice data into the bone conduction module and transmits it to the worker, who can receive summary voice such as "Prepare for parts assembly" via bone conduction without using their ears.

[1412] Audio instruction tools

[1413] This allows workers wearing visual assistance devices to receive voice instructions and provide accurate and prompt voice guidance to automated factory equipment, which can, for example, facilitate smooth communication between workers who speak different languages ​​and improve the efficiency of instruction transmission.

[1414] Specific examples

[1415] Say: "Get ready to assemble the next part."

[1416] Generative AI summary: "Parts ready for assembly"

[1417] Bone conduction feedback: "Preparing to assemble parts"

[1418] Example prompts to input to a generative AI model:

[1419] Get ready to assemble the next part

[1420] In this invention, the voice recognition means, generation AI means, bone conduction means, etc. cooperate with each other at each step to provide optimal voice guidance to workers, which is expected to significantly improve work efficiency and safety in factories.

[1421] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1422] Step 1:

[1423] The terminal activates the built-in microphone and collects the worker's voice in real time. The collected voice data is saved as digital data. Here, the input is the worker's voice and the output is digital voice data. In this step, data conversion is performed to save it as digital voice data.

[1424] Step 2:

[1425] The device compresses the collected audio data and sends it to the server via Bluetooth or Wi-Fi. Here, the input is digital audio data and the output is compressed audio data. In this step, a compression algorithm is used to reduce the data size and the data is sent to the server using a communication protocol.

[1426] Step 3:

[1427] The server records the received voice data and starts the voice recognition engine. Here, the input is compressed voice data and the output is text data. In this step, the voice recognition process is performed to convert the voice data into text. Specifically, Google's voice recognition API is used.

[1428] Step 4:

[1429] The server uses a generative AI method to summarize the converted text data, where the input is text data and the output is a summary text. In this step, a generative AI model (e.g., GPT-3) analyzes the input text and generates a summary.

[1430] Step 5:

[1431] The server compresses the generated summary text and sends it to the terminal. Here, the input is the summary text and the output is the compressed summary text. In this step, a compression algorithm is used to reduce the data size and the data is sent to the terminal using a communication protocol.

[1432] Step 6:

[1433] The terminal converts the received summary text into speech data through a speech synthesis engine. The input is the summary text and the output is speech data. Specifically, the text is converted into speech using the pyttsx3 engine.

[1434] Step 7:

[1435] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the worker. Here, the input is voice data, and the output is the voice that the worker receives through bone conduction. In this step, the voice data is transmitted without using the ears using bone conduction technology.

[1436] Step 8:

[1437] The worker understands the summarized audio received through bone conduction technology and gives instructions to the automated factory equipment. Here, the input is the audio via bone conduction, and the output is audio instructions to the automated factory equipment. In this step, the worker performs the actual work based on the audio instructions.

[1438] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1439] The present invention relates to a system including a speech recognition unit, a communication unit, a speech recognition engine unit, a generation AI unit, a speech synthesis unit, a bone conduction unit, and an emotion engine. This system can be implemented using a wearable device such as glasses worn by a user. Specific program processing of this system is described in detail below.

[1440] Collection and transmission of voice data

[1441] 1. The device activates the built-in microphone and collects surrounding audio in real time. The collected audio data is temporarily stored in a buffer.

[1442] 2. The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing if necessary.

[1443] 3. The device sends the compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are pre-set.

[1444] Speech-to-text conversion and summary generation

[1445] 4. The server saves the received voice data in a specific directory and prepares to pass it to the voice recognition engine.

[1446] 5. The server invokes the speech recognition engine to convert the audio data to text data. Here, the speech recognition algorithm analyzes the audio data and generates corresponding text.

[1447] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[1448] 6. The server starts the generation AI means and receives the text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[1449] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[1450] Recognizing user emotions with an emotion engine

[1451] 7. The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it to the emotion engine.

[1452] 8. The emotion engine analyzes the collected data and recognizes the user's current emotional state (e.g., excited, relaxed, anxious, etc.).

[1453] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[1454] Emotion regulation and transmission of summary data

[1455] 9. The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[1456] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[1457] 10. The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[1458] Speech synthesis of summary text and transmission via bone conduction

[1459] 11. The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[1460] Example: Generate the summary text "progress report" as machine speech "progress report."

[1461] 12. The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[1462] User Feedback

[1463] 13. Users can understand the summary audio received through bone conduction and efficiently grasp the content of the conversation.

[1464] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[1465] This system allows users to not only efficiently receive speech recognition and summarized information, but also receive optimal feedback according to their emotional state, enabling them to quickly and accurately understand the content of conversations and achieve smooth communication.

[1466] The processing flow will be explained below.

[1467] Step 1:

[1468] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[1469] Step 2:

[1470] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[1471] Step 3:

[1472] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The destination server address and connection method are preset.

[1473] Step 4:

[1474] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[1475] Step 5:

[1476] The server activates a speech recognition engine to convert the speech data into text data, which then analyzes the speech and generates corresponding text.

[1477] Example: Converting the audio data "I will report on the progress of the next project" into the text "I will report on the progress of the next project."

[1478] Step 6:

[1479] The server activates the generation AI means and receives text data from the speech recognition engine. The generation AI analyzes the text data, extracts important content, and generates a summary.

[1480] Example: Summarize the text data "I will report on the progress of the next project" as "Project progress report."

[1481] Step 7:

[1482] The device collects biometric data such as the user's voice, facial expressions, and pulse rate and passes it on to the emotion engine.

[1483] Step 8:

[1484] The emotion engine analyzes the collected biometric data to recognize the user's current emotional state, which can be classified as excited, relaxed, anxious, etc.

[1485] Example: The emotion engine determines that User B is "excited" based on their tone of voice and facial expression.

[1486] Step 9:

[1487] The server adjusts the content of the generated summary text based on the analysis results of the emotion engine. If the user is emotionally charged, the summary will be made more concise.

[1488] Example: When User B is excited, he / she will shorten "project progress report" to "progress report."

[1489] Step 10:

[1490] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[1491] Step 11:

[1492] The device passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[1493] Example: Generate the summary text "progress report" as machine speech "progress report."

[1494] Step 12:

[1495] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[1496] Step 13:

[1497] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[1498] Example: User B receives a summary audio such as "progress report" and quickly grasps the main points of User A's speech.

[1499] Example 2

[1500] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1501] Conventional speech recognition systems only convert speech data into text, and have the problem of being unable to provide information that takes into account the user's emotional state. Furthermore, the collection of speech data, summary generation, emotion recognition, and speech synthesis are all performed separately, which reduces the overall efficiency of the system. Therefore, there is a need for an integrated system that can reduce the burden on users and provide information quickly and accurately.

[1502] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a processor that converts voice data into text, a generative model that summarizes the text data, an emotion recognition means that analyzes the user's emotional state, and a voice synthesis means that converts the summarized text data into voice data. This makes it possible to perform everything from voice collection to emotion analysis, summary generation, and voice synthesis in an integrated manner.

[1503] A "device for collecting audio data" is a device that has the function of collecting surrounding audio in real time and storing it in a buffer.

[1504] A "communication means" is a device that has the function of compressing collected voice data and sending it to a server using Bluetooth, Wi-Fi, etc.

[1505] "Processor" means a central processing unit that executes speech recognition algorithms within the server to convert speech data into text data.

[1506] A "generative model" is an artificial intelligence model that analyzes text data, extracts important content, and generates a summary.

[1507] The "emotion recognition means" is a device or algorithm for analyzing biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[1508] The "speech synthesis means" is a synthesis device that has the function of converting summarized text data into natural speech.

[1509] An "acoustic device" is a device that has the function of transmitting audio data to a user through bone conduction.

[1510] A "wearable device" is a portable device that can be worn by a user and that collects and communicates audio data.

[1511] This invention is a system that integrates voice recognition, generative AI, emotion recognition, voice synthesis, and voice transmission via bone conduction, and provides efficient information provision using a wearable device worn by the user.

[1512] The system includes the following main components:

[1513] 1. Device for collecting voice data: A microphone built into the wearable device collects the user's voice and surrounding voices in real time and stores the data in a buffer.

[1514] 2. Communication method: The collected voice data is converted into digital form, compressed if necessary, and then sent to the server via Bluetooth, Wi-Fi, etc.

[1515] 3. Processor: The central processing unit installed in the server converts the received voice data into text data using a voice recognition algorithm (e.g., Google Cloud Speech-to-Text API).

[1516] 4. Generative models: These include artificial intelligence models (e.g., GPT-4) that analyze text data, extract key content, and generate summaries. Generative models use prompts as input and generate corresponding summary text.

[1517] Example prompt: "I'll report on the progress of my next project."

[1518] 5. Emotion recognition means: Sensors built into smart glasses and other wearable devices collect biometric data such as the user's voice, facial expressions, and pulse rate, and then use emotion recognition algorithms to analyze the user's emotional state.

[1519] Example: Judging whether a user is "excited" based on their tone of voice and facial expression.

[1520] 6. Speech synthesis means: Includes a speech synthesis engine (e.g., Amazon Polly) for converting the summarized text data into natural-sounding speech, which then converts the summarized text into speech data and transmits it to the user.

[1521] Example: Generate the summary text "progress report" as machine speech "progress report."

[1522] 7. Acoustic Device: Includes devices that utilize bone conduction technology to transmit synthesized audio data directly to the user's inner ear through the skull, allowing the user to receive audio information without blocking their ears.

[1523] This system allows users to benefit from:

[1524] Not only can you receive voice recognition and summarized information efficiently, but you can also receive optimal feedback based on your emotional state.

[1525] The entire process, from collecting voice data to analyzing, summarizing, synthesizing voice, and finally transmitting it, is managed centrally, ensuring that information is provided quickly and accurately.

[1526] This system is particularly useful in situations where large amounts of information need to be efficiently processed and transmitted in real time, such as business meetings and educational settings. The present invention allows users to achieve smooth communication and information comprehension.

[1527] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1528] Step 1:

[1529] The device activates the built-in microphone and collects surrounding audio in real time, and the collected audio data is temporarily stored in a buffer.

[1530] Input: Ambient audio

[1531] Output: Buffered audio data

[1532] Specific operation: The microphone of the wearable device (such as smart glasses) is activated and collects the user's speech and surrounding sounds. The collected data is temporarily stored in a buffer in memory.

[1533] Step 2:

[1534] The terminal converts the voice data stored in the buffer into digital voice data and performs compression processing as necessary.

[1535] Input: Buffered audio data

[1536] Output: Digital audio data (compressed)

[1537] Specific operation: A processor operates to convert analog audio data into digital, and then a codec is applied to compress the audio data.

[1538] Step 3:

[1539] The device uses a communication method such as Bluetooth or Wi-Fi to send compressed audio data to the server. The destination server address and connection method are preset.

[1540] Input: Digital audio data (compressed)

[1541] Output: Audio data sent to the server

[1542] What happens: Compressed audio data is sent via Wi-Fi or Bluetooth modules and reaches the specified API endpoint on the server.

[1543] Step 4:

[1544] The server saves the received voice data in a specific directory and prepares it for passing to the voice recognition engine.

[1545] Input: Audio data sent to the server

[1546] Output: Input file for speech recognition engine

[1547] Specific operation: The voice data is stored in a specific directory on the server, and the voice recognition engine is configured to refer to this file.

[1548] Step 5:

[1549] The server activates a speech recognition engine to convert the speech data into text data. The speech recognition algorithm analyzes the speech data and generates corresponding text.

[1550] Input: Input file for the speech recognition engine

[1551] Output: Text data

[1552] Specific operation: The process of converting speech to text is carried out by calling the Google Cloud Speech-to-Text API, etc. For example, speech saying "I will report on the progress of the next project" is converted to text saying "I will report on the progress of the next project."

[1553] Step 6:

[1554] The server launches the generative model and receives text data from the speech recognition engine. The generative AI analyzes the text data, extracts important content, and generates a summary.

[1555] Input: Text data

[1556] Output: Summary text

[1557] Specific operation: A GPT model (e.g., GPT-4) analyzes the text data "I will report on the progress of the next project" and summarizes it as "Project progress report."

[1558] Step 7:

[1559] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate and passes it to an emotion recognition means.

[1560] Input: User biometric data (voice, facial expression, pulse, etc.)

[1561] Output: Emotional state data

[1562] How it works: Sensors built into smart glasses or wearable devices collect biometric data from users and send it to emotion recognition algorithms.

[1563] Step 8:

[1564] An emotion recognition means analyzes the collected data and recognizes the user's current emotional state (eg, excited, relaxed, anxious, etc.).

[1565] Input: Biometric data

[1566] Output: Emotional state data

[1567] Specific operation: The emotion recognition algorithm analyzes the tone of voice and facial expressions to determine the user's emotional state, such as "excited" or "relaxed."

[1568] Step 9:

[1569] The server adjusts the content of the generated summary text based on the analysis results of the emotion recognition means, making the summary more concise if the user is emotionally excited.

[1570] Input: Summary text, emotional state data

[1571] Output: Adjusted summary text

[1572] Specific operation: The server adjusts the generated summary "Project Progress Report" to be more concise as "Progress Report" based on the emotional state data.

[1573] Step 10:

[1574] The server compresses the adjusted summary text and sends it to the device, using Bluetooth or Wi-Fi as the communication method, just like the initial transmission.

[1575] Input: Adjusted summary text

[1576] Output: Summary text sent to terminal

[1577] What it does: The adjusted summary text is sent to your device using Bluetooth or Wi-Fi.

[1578] Step 11:

[1579] The terminal passes the received summary text to a speech synthesis engine, which converts it into synthetic speech data. The speech synthesis engine uses an algorithm to convert text data into natural-sounding speech.

[1580] Input: Adjusted summary text

[1581] Output: Synthesized speech data

[1582] What happens: A speech synthesis engine such as Amazon Polly converts the text "progress report" into machine-generated speech.

[1583] Step 12:

[1584] The device inputs the synthesized voice data into the bone conduction module, which transmits the voice directly to the wearer's inner ear through the skull.

[1585] Input: Synthetic speech data

[1586] Output: Transmission of audio information via bone conduction

[1587] Specific operation: The synthesized voice data is transmitted directly to the user's inner ear through the bone conduction module, and the user hears the voice.

[1588] Step 13:

[1589] The user can understand the summary speech received through bone conduction and efficiently grasp the content of the conversation.

[1590] Input: Audio information via bone conduction

[1591] Output: Understood conversation

[1592] Specific operation: The user receives a summary audio such as "progress report" via bone conduction and quickly grasps the main points of the talk.

[1593] The above are the specific steps of the program processing of this system, which allows users to efficiently enjoy speech recognition, summary generation, emotion recognition, speech synthesis, and bone conduction communication.

[1594] (Application example 2)

[1595] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1596] Security operations require accurate and rapid understanding of on-site situations in real time and the most appropriate course of action. However, conventional methods are prone to information delays and miscommunication, making it difficult to respond immediately based on the urgency of the situation and the emotional state of the guards. To solve this problem, a system is needed that summarizes the on-site situation in real time and provides information that is appropriately adjusted according to the emotional state of the guards.

[1597] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a voice recognition engine means for converting voice data into text, a generation AI means for summarizing the voice recognition results, and an emotion engine means for collecting user emotion data and adjusting the summarized text data based on the emotion data. This makes it possible to immediately grasp the situation on-site and provide optimal information according to the user's emotional state.

[1598] A "voice recognition means" is a device or method that collects voice data in real time and converts it into digital voice data.

[1599] "Communication means" refers to a means for transmitting collected voice data to a server, and includes wireless communication technologies such as Bluetooth and Wi-Fi.

[1600] "Speech recognition engine means" refers to software or algorithms that analyze voice data and convert it into corresponding text data.

[1601] "Generative AI means" refers to means that use artificial intelligence technology to analyze converted text data, extract important content, and generate a summary.

[1602] The "speech synthesis means" refers to a means for converting summarized text data into speech data, and includes an algorithm for generating natural-sounding speech.

[1603] "Bone conduction means" refers to devices or technologies that transmit sound directly through the wearer's skull to the inner ear.

[1604] "Emotion engine means" refers to technology or devices that analyze biometric data such as the user's voice, facial expression, and pulse rate to recognize the user's emotional state.

[1605] The present invention is a security system that includes a voice recognition unit, a communication unit, a voice recognition engine unit, a generative AI unit, a voice synthesis unit, a bone conduction unit, and an emotion engine unit. Specifically, the system uses a wearable device worn by a security guard. With this system, the security guard reports the situation on-site by voice, and the voice data is sent to a server for summarization and emotion analysis. Appropriately adjusted information is then provided to the security guard in real time via bone conduction.

[1606] The device activates its built-in microphone and collects audio from the site in real time. This audio data is temporarily stored in a buffer and converted into digital audio data. The compressed audio data is then sent to a server via communication methods such as Bluetooth or Wi-Fi.

[1607] The server converts the received voice data into text data using a voice recognition engine means. For example, voice data reported by a security guard that "a suspicious person has broken in" is converted into text data. Next, a generation AI means analyzes the text data, extracts important content, and generates a summary. For example, "a suspicious person has broken in" is summarized as "suspicious person has broken in." After that, an emotion engine means recognizes the user's emotional state from their voice and facial expression, and adjusts the summary text based on the emotion data. For example, if the user is in a tense state, the summary text is further simplified to "Emergency! Suspicious person has broken in."

[1608] The resulting adjusted summary text data is converted into natural-sounding speech using a speech synthesis device. This speech data is transmitted to the security guard via bone conduction. The security guard receives the speech information via bone conduction, allowing them to quickly grasp important information on the scene.

[1609] For example, when a security guard reports that "a suspicious person has entered the entrance," the system summarizes the voice data as "Suspicious person has entered," and after sensing the guard's state of tension, further simplifies it to "Emergency! Suspicious person has entered," and transmits it via bone conduction. In this way, the security guard can instantly take the most appropriate action depending on the situation.

[1610] Example prompt sentence:

[1611] "A suspicious person entered the entrance" Summarize the situation in one sentence: "A suspicious person entered the entrance"

[1612] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1613] Step 1:

[1614] The terminal activates the built-in microphone and collects audio from the scene in real time. The input is real-time audio data, and the output is digital audio data temporarily stored in a buffer.

[1615] Step 2:

[1616] The terminal converts the buffered audio data into a digital format and compresses it if necessary. The input is the audio data in the buffer, and the output is compressed digital audio data.

[1617] Step 3:

[1618] The device sends compressed audio data to the server using a communication method such as Bluetooth or Wi-Fi. The input is compressed digital audio data, and the output is audio data sent to the server.

[1619] Step 4:

[1620] The server stores the received voice data in a specific directory and prepares to pass it to the voice recognition engine means. The input is the voice data sent to the server, and the output is the voice data for the recognition engine to process.

[1621] Step 5:

[1622] The server converts the voice data into text data using a voice recognition engine means, where the input is the received voice data and the output is the converted text data.

[1623] Step 6:

[1624] The server uses generative AI methods to analyze the text data, extract important content, and generate a summary. The input is the converted text data, and the output is the summarized text data.

[1625] Step 7:

[1626] The terminal collects biometric data such as the user's voice, facial expression, and pulse rate, and passes it to the emotion engine means. The input is the user's biometric data, and the output is data representing the user's emotional state.

[1627] Step 8:

[1628] The server uses an emotion engine means to analyze the collected data and recognize the user's current emotional state, where the input is the user's biometric data and the output is the emotional state data.

[1629] Step 9:

[1630] The server adjusts the content of the generated summary text based on the emotional state data. The input is the emotional state data and the summary text data, and the output is the adjusted summary text data.

[1631] Step 10:

[1632] The server compresses the adjusted summary text data and sends it to the terminal, where the input is the adjusted summary text data and the output is the compressed text data sent to the terminal.

[1633] Step 11:

[1634] The terminal passes the received summary text to a speech synthesis engine means for converting it into synthesized speech data, with the adjusted summary text data as input and the synthesized speech data as output.

[1635] Step 12:

[1636] The terminal inputs the synthesized voice data into the bone conduction module and transmits it to the user. The input is the synthesized voice data, and the output is the transmission of voice through the bone conduction module.

[1637] Step 13:

[1638] The user understands the summary speech received through bone conduction and efficiently grasps the content of the conversation and the situation on the scene. The input is the speech data received through bone conduction, and the output is the understood information.

[1639] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1640] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1641] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1642] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1643] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1644] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1645] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1646] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1647] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1648] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1649] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1650] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1651] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1652] 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.

[1653] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1654] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1655] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1656] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1657] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1658] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1659] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1660] The following is further disclosed regarding the above embodiment.

[1661] (Claim 1)

[1662] a speech recognition means for collecting speech data;

[1663] a communication means for transmitting voice data to a server;

[1664] a speech recognition engine means for converting speech data into text in the server;

[1665] A generative AI means for summarizing speech recognition results;

[1666] a speech synthesis means for converting the summarized text data into speech data;

[1667] a bone conduction means for transmitting audio data to a wearer by bone conduction;

[1668] A system including:

[1669] (Claim 2)

[1670] 10. The system of claim 1, comprising a multilingual translation function.

[1671] (Claim 3)

[1672] 10. The system of claim 1, comprising a wearable device that collects and communicates audio data.

[1673] "Example 1"

[1674] (Claim 1)

[1675] a speech recognition means for collecting speech data;

[1676] a communication means for transmitting voice data to a server;

[1677] a speech recognition engine means for converting speech data into text in the server;

[1678] A generative AI means for summarizing speech recognition results;

[1679] a speech synthesis means for converting the summarized text data into speech data;

[1680] a bone conduction means for transmitting audio data to a wearer by bone conduction;

[1681] A means to activate and collect data from the built-in microphone in real time;

[1682] means for storing the audio data as digital data;

[1683] means for compressing audio data;

[1684] means for recording received audio data;

[1685] means for compressing and transmitting the generated summary text;

[1686] A system including:

[1687] (Claim 2)

[1688] 10. The system of claim 1, comprising a multilingual translation function.

[1689] (Claim 3)

[1690] 10. The system of claim 1, comprising a wearable device that collects and communicates audio data.

[1691] "Application Example 1"

[1692] (Claim 1)

[1693] a speech recognition means for collecting speech data;

[1694] a communication means for transmitting voice data to a server;

[1695] a speech recognition engine means for converting speech data into text in the server;

[1696] A generative AI means for summarizing speech recognition results;

[1697] a speech synthesis means for converting the summarized text data into speech data;

[1698] a bone conduction means for transmitting audio data to a wearer by bone conduction;

[1699] a voice instruction means for giving instructions to automated factory equipment using a visual assistance device worn by a worker;

[1700] A system including:

[1701] (Claim 2)

[1702] 10. The system of claim 1, comprising a multilingual translation function.

[1703] (Claim 3)

[1704] 10. The system of claim 1, further comprising a visual aid device that collects and communicates audio data.

[1705] "Example 2: Combining Emotion Engines"

[1706] (Claim 1)

[1707] a device for collecting voice data;

[1708] a communication means for transmitting voice data to a server;

[1709] a processor at the server for converting the voice data into text;

[1710] a generative model that summarizes text data;

[1711] emotion recognition means for analyzing the emotional state of a user;

[1712] a speech synthesis means for converting the summarized text data into speech data;

[1713] an acoustic device that transmits audio data to a wearer by bone conduction;

[1714] A system including:

[1715] (Claim 2)

[1716] 10. The system of claim 1, comprising a multilingual translation function.

[1717] (Claim 3)

[1718] 10. The system of claim 1, comprising a wearable device that collects and communicates audio data.

[1719] "Application example 2 when combining emotion engines"

[1720] (Claim 1)

[1721] a speech recognition means for collecting speech data;

[1722] a communication means for transmitting voice data to a server;

[1723] a speech recognition engine means for converting speech data into text in the server;

[1724] A generative AI means for summarizing speech recognition results;

[1725] a speech synthesis means for converting the summarized text data into speech data;

[1726] a bone conduction means for transmitting audio data to a wearer by bone conduction;

[1727] an emotion engine means for collecting emotion data of a user and adjusting the summary text data based on the emotion data;

[1728] A system including:

[1729] (Claim 2)

[1730] 10. The system of claim 1, comprising a multilingual translation function.

[1731] (Claim 3)

[1732] 10. The system of claim 1, comprising a wearable device that collects and communicates audio data. [Explanation of symbols]

[1733] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a speech recognition means for collecting speech data; a communication means for transmitting voice data to a server; a speech recognition engine means for converting speech data into text in the server; A generative AI means for summarizing speech recognition results; a speech synthesis means for converting the summarized text data into speech data; a bone conduction means for transmitting audio data to a wearer by bone conduction; A system including:

2. The system according to claim 1, further comprising a multilingual translation function.

3. The system of claim 1 , comprising a wearable device that collects and communicates audio data.

Images