System

Abstract

A system is provided.SOLUTION: A system comprising: means for acquiring voice data; means for converting the acquired voice data into text data; means for extracting and summarizing important points from the converted text data; means for generating the summarized text data as minutes in a format designated by a user; means for providing the generated minutes to the user and allowing the user to correct the minutes; and means for sending the corrected minutes to a person concerned.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] When creating minutes of a meeting, participants must concentrate on listening to what is being said, which can make it difficult to speak. Furthermore, important comments may be missed in the minutes because participants are unable to keep up with the pace of listening. As a result, creating accurate minutes requires a great deal of time and effort. Therefore, a method is needed to create accurate and prompt minutes while improving meeting efficiency. [Means for solving the problem]

[0005] In order to solve the above-mentioned problems, the present invention provides a system including means for acquiring voice data, means for converting the acquired voice data into text data, means for extracting and summarizing important points from the converted text data, means for generating the summarized text data as minutes in a format specified by the user, means for providing the generated minutes to the user so that the user can modify them, and means for sending the modified minutes to relevant parties.

[0006] First, audio during a meeting is acquired in real time using a means for acquiring audio data. Next, the audio is transcribed using a means for converting the acquired audio data into text data. After that, the text is summarized using a means for extracting and summarizing important points from the converted text data. Based on this summarized text data, minutes are created using a means for generating minutes in a format specified by the user.

[0007] Furthermore, by providing the generated minutes to the user and providing a means for the user to review and correct them, accurate minutes can be completed. Finally, by automatically sending the corrected minutes to the relevant parties, the minutes can be shared quickly and efficiently. In this way, the present invention reduces the burden of meetings and enables the efficient and accurate creation of minutes.

[0008] "Audio data" is data in a format that records the voice of a speaker occurring during a conference.

[0009] "Capturing means" refers to equipment or software capable of capturing and recording audio data in real time and inputting it into an electronic device for processing.

[0010] "Text data" is character information that has been converted by analyzing voice data.

[0011] "Means for converting" refers to speech recognition technology or software for analyzing voice data and converting it into text data.

[0012] "Key points" are the key points of information or statements that are particularly needed within the meeting content.

[0013] "Extraction means" refers to algorithms or software that identify and select important points from the converted text data.

[0014] A "summarization tool" is an algorithm or software that summarizes the extracted key points in a concise form.

[0015] "Minutes" are documents that record what was said at a meeting.

[0016] The "means for generating" is software for creating minutes in a specified format based on the summarized text data.

[0017] The "means for providing and amending" refers to a system or interface that displays the generated minutes to the user and allows the user to check and amend the contents.

[0018] The "means of sending" refers to a system or software that automatically sends the revised minutes to the relevant parties via email or other means. [Brief explanation of the drawings]

[0019] [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

[0020] 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.

[0021] First, the terms used in the following description will be explained.

[0022] 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).

[0023] 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.

[0024] 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.

[0025] 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.

[0026] 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."

[0027] [First embodiment]

[0028] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0029] 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.

[0030] 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).

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] 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.

[0037] 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.

[0038] 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.

[0039] 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."

[0040] The present invention is a series of systems that captures conference audio data in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0041] Acquiring audio data

[0042] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0043] Converting audio data into text data

[0044] The server converts the received voice data into text data in real time using a voice recognition engine, which has the ability to recognize voice and convert it into text with high accuracy.

[0045] Extraction and summary of key points

[0046] The converted text data is processed on a server, where a generative AI model is used to extract key points, which are then summarized into concise text using an algorithm.

[0047] Generate meeting minutes

[0048] Based on the summarized text data, the server generates a transcript according to the format specified by the user, which can be output in a standard or custom format.

[0049] Provision and correction of minutes

[0050] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. These corrections can be easily made using a dedicated interface.

[0051] Sending minutes

[0052] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting.

[0053] Specific examples

[0054] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The generative AI model then summarizes the text as "Today we will review the progress of the project." Based on the summary text, minutes are generated in a standard format. The user can review and edit the minutes, and they are eventually automatically sent to all parties involved.

[0055] In this way, the present invention significantly improves the efficiency of taking minutes at meetings, saving time and effort.

[0056] The processing flow will be explained below.

[0057] Step 1:

[0058] The terminal acquires voice data using a microphone at the same time as the conference starts, and transmits the voice data to the server in real time.

[0059] Step 2:

[0060] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[0061] Step 3:

[0062] The server inputs the converted text data into a generative AI model, which extracts key points from the text data. The generative AI model uses natural language processing techniques to identify important statements and content.

[0063] Step 4:

[0064] The server then uses a summarization algorithm to create a summary of the key information based on the extracted key points. This summarization algorithm is designed to achieve both accuracy and conciseness.

[0065] Step 5:

[0066] The server generates minutes from the summarized text data in a format specified by the user, and outputs the minutes in a standard or custom format.

[0067] Step 6:

[0068] The server provides the generated minutes to the user, and allows the user to check the contents of the minutes using a dedicated interface and correct them as necessary.

[0069] Step 7:

[0070] After the user completes the edits, the server automatically sends the final minutes to the relevant email addresses, allowing all meeting participants to quickly share accurate minutes.

[0071] Example 1

[0072] 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."

[0073] While it is important to create meeting minutes accurately and quickly, traditional methods rely on manual processes, which are time-consuming and labor-intensive. Furthermore, it is not easy to check and correct minutes in real time and share them with relevant parties immediately, which can lead to delays in information transmission. For these reasons, there was a need to improve the efficiency and accuracy of the minutes creation process.

[0074] 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.

[0075] In this invention, the server includes a means for transmitting voice data to the server in real time, a means for converting voice data into text data using a voice recognition engine, and a means for extracting key points using a generative AI model. This enables the voice data of a meeting to be converted into text in real time, key points to be efficiently extracted and summarized, and minutes to be generated quickly. Furthermore, the generated minutes can be reviewed and revised in real time by the user and immediately sent to relevant parties, preventing delays in information transmission and significantly improving post-meeting work efficiency.

[0076] "Audio data" refers to data in which audio information from meetings, interviews, etc. is recorded in digital format.

[0077] "Text data" refers to data of character information generated from voice data using voice recognition technology or the like.

[0078] A "server" is a computer system that provides resources and data over a network.

[0079] A "terminal" is a device operated by a user, and includes a computer, laptop, tablet, smartphone, etc.

[0080] A "user" is someone who uses the system, and includes conference participants and administrators.

[0081] A "voice recognition engine" is software or hardware that analyzes voice data and converts it into corresponding text data.

[0082] A "generative AI model" is an artificial intelligence model that uses artificial intelligence technology to analyze data and perform tasks such as extracting key points and summarizing.

[0083] "Minutes" are text documents that record the contents of a meeting or the like.

[0084] "Real-time" refers to data processing and information transmission occurring almost instantaneously, with minimal delay.

[0085] A "prompt sentence" is an input sentence given to a generative AI model that contains instructions for the model to perform a specific task.

[0086] A "summary" is a sentence that succinctly summarizes the contents of the text data.

[0087] An "email sending API" is an interface for sending email from a program.

[0088] The present invention is a system that captures audio data of a meeting in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0089] When a conference begins, the device captures audio data using its built-in microphone or an external microphone. The captured audio data is temporarily stored in a buffer and then sent to the server in real time. The audio data is sent using the TCP / IP protocol.

[0090] The server instantly converts the received voice data into text data using a "voice recognition engine." This voice recognition engine can be a general-purpose tool such as "Google Cloud Speech-to-Text." The server sends the voice data to the engine via an API and saves the returned text data.

[0091] The converted text data is then used on the server to extract important points using the generative AI model "GPT-4." The generative AI model then uses prompts to perform specific tasks. Specifically, a prompt containing the instruction "Please summarize the contents of today's meeting" is used. An example of a prompt is "Convert the audio data from the meeting into text data, summarize the key points, and create minutes. The contents of the meeting discussion are listed below."

[0092] The extracted key points are summarized into concise text using a summarization algorithm, and the server generates minutes based on the summary text. The format of the minutes can be specified by the user, and they can be output in formats such as Markdown, PDF, Word, etc. For example, the server can use a template engine to convert the summary text into Markdown format and output it as PDF using a library such as "PDFKit."

[0093] The generated minutes are provided to the user by the server through a dedicated web interface, where the user can review the contents of the minutes and make any necessary corrections. This web interface has a function for editing the minutes text in real time. After the user has completed the corrections, the minutes are automatically sent to all parties involved from the server using an email sending API (e.g., "SendGrid" or "Amazon SES").

[0094] As a concrete example, when a project team holds a progress meeting, a laptop or tablet picks up the audio of the meeting and sends it to a server in real time. Google Cloud Speech-to-Text converts the audio into text, such as "Today we will discuss the progress of the project," and the generative AI model GPT-4 summarizes this as "Today we will review the progress of the project." The summarized text is converted to Markdown format and output as a PDF. The user can review and edit these minutes via a web interface, and they are finally automatically sent to all parties involved.

[0095] Thus, the present invention efficiently and automatically processes minutes of meetings, thereby saving a great deal of time and effort.

[0096] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0097] Step 1:

[0098] When a conference starts, the device captures audio data using a built-in microphone or an external microphone. The audio data is temporarily stored in a buffer in real time and then sent to the server using the TCP / IP protocol.

[0099] Specific example of operation: A laptop installed in a conference room picks up speech using its built-in microphone and transmits it to a server via Wi-Fi.

[0100] Step 2:

[0101] The server converts the voice data it receives in real time into text data using a "voice recognition engine." The voice data (input) is analyzed by the voice recognition software and the corresponding text data (output) is generated.

[0102] Specific example of operation: The server sends voice data to the "Google Cloud Speech-to-Text" service, receives the results as text data, and stores them in a database.

[0103] Step 3:

[0104] The server extracts and summarizes the key points from the converted text data using a "generative AI model." The text data (input) is sent to the AI ​​model with a prompt, and a text (output) summarizing the key points is generated.

[0105] Specific example of operation: GPT-4 is sent a prompt saying, "Convert the audio data of the meeting into text data and summarize the key points," and the AI ​​model outputs the summary text, "Today we will review the progress of the project."

[0106] Step 4:

[0107] The server generates minutes based on the summarized text data. It converts the summarized text (input) into a specified format (output) using a template engine.

[0108] Specific example of operation: Convert the summary text into Markdown format and output it as PDF using the "PDFKit" library.

[0109] Step 5:

[0110] The server provides the generated minutes to the user through a dedicated web interface. The PDF file (input) is displayed via the web interface, and the user checks its contents.

[0111] Specific example of operation: A user opens a web browser, accesses a dedicated URL, and checks the displayed minutes.

[0112] Step 6:

[0113] Users can modify the contents of the minutes through a web interface by directly editing the displayed text (input) and saving it as modified text (output).

[0114] Example of how it works: A user edits typos or additional information on the web interface and clicks the "Save" button, which sends the corrections to the server.

[0115] Step 7:

[0116] The server automatically sends the revised minutes to the relevant parties. The revised PDF file (input) is sent to the relevant parties' email addresses using the email sending API.

[0117] Example of specific operation: The server attaches the corrected minutes and uses the "SendGrid" API to send a message to the relevant email address with the message "The meeting minutes are complete."

[0118] In this way, the system efficiently executes a series of processes from acquiring audio data of a meeting to generating, editing, and distributing minutes.

[0119] (Application example 1)

[0120] 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."

[0121] Conventional conferencing systems require manual processes, from collecting audio data to converting it into text, extracting key points, and creating and sending meeting minutes, resulting in inefficiency and requiring a great deal of time and effort. Furthermore, in-factory meetings also faced difficulties in generating and sharing work instructions and reports in real time, reducing work efficiency. Furthermore, there was a risk that important points would be overlooked.

[0122] 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.

[0123] In this invention, the server includes a means for acquiring voice data, a means for converting the acquired voice data into text data, and a means for extracting and summarizing important points from the converted text data. This automatically converts voice data into text, making it possible to summarize important points and generate minutes, work instructions, and reports. In addition, the generated documents can be modified and confirmed by the user, enabling highly accurate information sharing.

[0124] "Audio data" refers to audio recorded and stored in digital format from meetings, work sites, etc.

[0125] "Text data" is digital data that has been analyzed and converted into text information.

[0126] "Key points" are pieces of information that indicate important points or key points within the collected data.

[0127] A "generative AI model" is an artificial intelligence algorithm that generates or summarizes new information based on machine learning and data processing.

[0128] "Summarizing" means to concisely summarize a large amount of information and extract and organize only the main points and main points.

[0129] Minutes are documents that record the contents of meetings and discussions.

[0130] "Correction" refers to the process of changing the content of generated documents or data to improve their accuracy and clarity.

[0131] "Stakeholders" are people who need to share information or work instructions in a particular meeting or task.

[0132] A "work instruction" is a document that provides specific instructions on how to perform a particular task or work.

[0133] A "report" is a formal document used to record and report a particular situation or work process.

[0134] The system of this invention is designed to capture voice data from meetings and factories in real time, automatically convert it into text, and then extract and summarize important points to generate minutes, work instructions, and reports.The main components of this system include a terminal that captures the voice data, a server that processes the data, and a user interface for making corrections.

[0135] Hardware / Software Configuration

[0136] Device:

[0137] This device is equipped with a microphone to capture voice data during meetings or in factories. The voice data is captured in real time and sent to a server.

[0138] server:

[0139] Google Cloud Speech-to-Text is used to convert voice data into text data. This text data is processed internally on the server, and key points are extracted and summarized. OpenAI's GPT-3 is used as a generative AI model to generate summaries and instructions.

[0140] Data processing / data calculation:

[0141] 1. Acquisition and transmission of audio data:

[0142] It uses the device's microphone to collect voice data and transmits it to a server in real time.

[0143] 2. Audio to text conversion:

[0144] Using the Google Cloud Speech-to-Text API, voice data is converted into text with high accuracy.

[0145] 3. Extract and summarize key points:

[0146] The text data is analyzed internally on the server, and key points are extracted and summarized using OpenAI GPT-3's generative AI model.

[0147] 4. Generate minutes, work instructions, and reports:

[0148] Automatically generate meeting minutes, work instructions, and reports based on summarized text data, and output these documents in standard or customizable formats.

[0149] 5. User Modifications:

[0150] The generated document is presented to the user through a dedicated interface that allows for modification, and supports real-time modification.

[0151] Specific examples

[0152] For example, in a factory progress meeting, a device captures the audio data of the meeting in real time and sends it to a server. The server converts the audio data into text data such as "Today we will discuss the importance of maintaining machine A." Next, a generative AI model is used to summarize the text as "Maintenance of machine A is important." A work instruction is generated based on the summarized text. The user can then review and revise the work instruction, and it is finally automatically sent to all parties involved.

[0153] Example prompts to input to a generative AI model:

[0154] Summarize the following text:

[0155] Today we will discuss the importance of maintaining Machine A.

[0156] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0157] Step 1:

[0158] The terminal captures voice data in real time during meetings and factories. The terminal's built-in microphone collects the voice data in digital format. The collected voice data is then sent to a server in real time.

[0159] Input: Audio from meetings and factories

[0160] Output: Digital audio data (sent to server)

[0161] Step 2:

[0162] The server converts the acquired voice data into text data using the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives the returned text data.

[0163] Input: Digital audio data

[0164] Output: Text data (converted from audio)

[0165] Step 3:

[0166] The server analyzes the text data and extracts and summarizes key points using a generative AI model (OpenAI GPT-3). First, the text data is input to the generative AI model as a prompt sentence, and the summarized text is obtained.

[0167] Input: Text data (records of meetings and work)

[0168] Output: Summarized text data (key points)

[0169] Step 4:

[0170] The server generates meeting minutes, work instructions, and reports in a user-specified format based on the summarized text data, and the generated documents can be output in standard or customized formats.

[0171] Input: Summarized text data

[0172] Output: minutes, work instructions, reports (user-specified format)

[0173] Step 5:

[0174] The generated document is provided to the user from the server, who can review the document and make corrections as needed through a dedicated interface that supports real-time corrections.

[0175] Input: Minutes, work orders, reports

[0176] Output: Corrected document (user confirmation and correction)

[0177] Step 6:

[0178] The server automatically sends the corrected document to the relevant person's email address.

[0179] Input: revised document

[0180] Output: Send to relevant parties (via email)

[0181] Specifically, the system begins by capturing the audio of the meeting in real time and sending it to the server, after which a series of processes are automatically carried out, including converting the audio data into text, extracting important points, generating various documents, allowing the user to check and correct them, and finally sending them.

[0182] 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.

[0183] The present invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions.

[0184] Acquiring audio data

[0185] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0186] Converting audio data into text data

[0187] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[0188] Generating emotional data using an emotion engine

[0189] The server inputs the converted text data into an emotion engine to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of voice and the choice of words used, and emotion data is generated.

[0190] Extraction and summary of key points

[0191] The converted text data, along with the emotion data, is then fed into a generative AI model, which the server uses to extract key points and, if necessary, highlight parts that emphasize the user's emotions.The server then uses a summarization algorithm to compile the extracted key points into concise text.

[0192] Generate meeting minutes

[0193] Based on the summarized text data, the server generates minutes according to the format specified by the user, and the generated minutes can be output in a standard or custom format.

[0194] Provision and correction of minutes

[0195] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. Corrections can be easily made using a dedicated interface.

[0196] Sending minutes

[0197] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting ends.

[0198] Specific examples

[0199] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The emotion engine then analyzes emotions such as "excitement" or "anxiety" from the text data and summarizes it as "Today we will review the progress of the project. - (excited)." The parts that emphasize the user's emotions are highlighted and recorded in the minutes.

[0200] The server then generates minutes in a standard format based on this summary text. The user can then make any necessary corrections, and the minutes are finally sent automatically to all parties involved. In this way, the present invention reduces the burden on meetings and enables efficient and accurate minutes creation. Furthermore, the introduction of an emotion engine makes it possible to generate minutes that reflect the emotional state of participants, thereby providing a deeper understanding of the progress of the meeting.

[0201] The processing flow will be explained below.

[0202] Step 1:

[0203] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0204] Step 2:

[0205] The server inputs the received voice data into a speech recognition engine, which performs highly accurate speech recognition and converts the voice data into text. This speech recognition engine recognizes the voices of multiple speakers in the conference and transcribes each individual statement into text.

[0206] Step 3:

[0207] The server inputs the converted text data into an emotion engine, which analyzes the word choice, context, and speaker's tone in the text data to recognize the user's emotions in real time and generate emotion data. This emotion data includes emotion labels such as "happiness," "anxiety," and "excitement."

[0208] Step 4:

[0209] The server combines the generated emotion data with the text data and inputs it into a generative AI model, which extracts key points from the text and identifies parts that emphasize the user's emotions.

[0210] Step 5:

[0211] The server extracts key points and uses a summarization algorithm to create a concise summary text, highlighting parts that emphasize specific emotions based on the emotion data.

[0212] Step 6:

[0213] The server generates minutes from the summarized text data in a format specified by the user, and the generated minutes can be output in a standard or custom format.

[0214] Step 7:

[0215] The generated minutes are provided to the user from the server. The user can review the minutes using a dedicated interface and make corrections as necessary. The minutes also include highlights of the user's emotions, making it easier to understand the nuances of the meeting.

[0216] Step 8:

[0217] Once the user has completed the revisions, the server automatically sends the final minutes to the relevant parties' email addresses, including sentiment data and key points from the meeting, so that all meeting participants can quickly share them.

[0218] Specific examples

[0219] For example, suppose someone says, "Today we will discuss the progress of the project" during a project status meeting. The device captures this utterance and sends it to the server in real time. The server converts the speech into text data, "Today we will discuss the progress of the project." If the emotion engine identifies the emotion "excited," the server includes this emotion data. Next, the generative AI model extracts the key point, "Checking the progress of the project," and generates the summary text, "Today we will check the progress of the project. - (Excited)."

[0220] Based on this summary, minutes are created and provided to the user, who can then make any necessary corrections, and the final minutes are sent to all parties involved, allowing for quick sharing of important meeting content and related emotions.

[0221] Example 2

[0222] 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."

[0223] Conventional meeting minutes creation systems require a great deal of time and effort to capture audio data, convert it to text data, extract key points, and generate minutes, significantly reducing meeting efficiency. Furthermore, minutes created do not take into account the emotions of meeting participants, making it difficult to gain a deep understanding of the progress of a meeting.

[0224] The specification process by the specification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for transmitting voice data to the server in real time, means for inputting the received voice data into a voice recognition engine and converting it into text data, means for inputting the converted text data into an emotion engine and recognizing the user's emotion to generate emotion data, and means for inputting the generated emotion data and text data into a generative AI model and extracting and summarizing important points. This makes it possible to efficiently generate minutes of a meeting from the voice data in real time and provide accurate minutes that reflect the emotions of the participants.

[0225] "Audio data" refers to data that records audio in digital format.

[0226] A "server" refers to a computer system that sends, receives, and processes data over a network.

[0227] "Terminal" refers to a device such as a computer or smartphone that can be directly operated by a user.

[0228] A "voice recognition engine" refers to software or a system that analyzes voice data and converts it into text data.

[0229] "Text data" refers to data that records character information in digital format.

[0230] An "emotion engine" refers to software or a system that analyzes text data or voice tone to generate emotional data.

[0231] "Emotion data" refers to data that indicates the user's emotional state as analyzed by the emotion engine.

[0232] A "generative AI model" refers to an artificial intelligence model that performs natural language processing based on large amounts of data and generates new text.

[0233] "Extracting important points" refers to finding particularly important content or key points from text data or emotion data.

[0234] "Summarization" refers to shortening long text data to include only the main information.

[0235] Minutes are a document that records what was discussed at a meeting and can be referenced at a later date.

[0236] "Real-time" refers to a state in which data processing and communication occurs almost instantaneously.

[0237] "Providing to users" means making the data and information generated by the system accessible and usable by users.

[0238] "Modify" refers to changing the content of data or documents that have already been generated and stored.

[0239] "Send" refers to sending the generated or modified data or documents to other users or systems involved.

[0240] This invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions. This system functions efficiently, with the terminal, server, and user each playing their own role.

[0241] As soon as the conference starts, the device uses a microphone to capture audio data. The captured audio data is sent to the server in real time. Specifically, audio capture software installed on the device (e.g., a Windows audio recording application) records the audio in real time and sends it in a stream format to the server using HTTP or WebSocket.

[0242] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. In this conversion process, the server calls the API of the voice recognition engine, sends the voice data, and captures the returned text data.

[0243] The server then inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of the voice and the choice of words used, and emotion data is generated.

[0244] The server then inputs the generated emotion data and converted text data into a generative AI model (e.g., OpenAI GPT-3) to extract key points. The server then creates a prompt and sends it to the generative AI model. For example, the prompt might include something like, "Please summarize the following conversation. Please also consider sentiment analysis." The server then receives the response from the generative AI model, extracts key points, and summarizes them.

[0245] Based on the summarized text data, the server generates minutes in the format specified by the user. The generated minutes are output in a standard or custom format. The user can use a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. This correction process is intuitive and easy to perform.

[0246] The minutes, once they have been finalized, are automatically sent from the server to the email addresses of the relevant parties. Specifically, the server uses an email sending program (e.g., an SMTP server) to send the minutes to the specified recipients as an attachment.

[0247] Specific examples

[0248] For example, suppose a project team is holding a progress meeting. The device records the meeting audio in real time using its built-in microphone and sends it to the server via HTTPS. The server then uses the Google Cloud Speech-to-Text API to convert the audio into text data, such as "Today we will discuss the progress of the project." It then uses an emotion engine (IBM Watson Tone Analyzer) to recognize emotions such as "excitement" or "anxiety" from the text data. A prompt is then input into a generative AI model (OpenAI GPT-3), which extracts key points and summarizes them, taking into account the emotion data. For example, the summary might be something like "Today we will review the progress of the project. - (excited)." Based on the generated summary text, the server generates meeting minutes in standard PDF format or a custom format. The user can review the minutes in a web browser and make any necessary corrections. The server then automatically sends them to all parties involved via an SMTP server.

[0249] Prompt Sentence Examples

[0250] "Generate minutes by summarizing what was said in project team status meetings and adding sentiment analysis."

[0251] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0252] Step 1: Acquire audio data

[0253] The device starts capturing audio data using a microphone as soon as the conference starts. Audio can be recorded in real time using a built-in microphone or a connected microphone, or using audio capture software (e.g., Windows audio recording application). The input here is ambient audio, and the output is captured digital audio data.

[0254] Step 2: Sending audio data

[0255] The device sends the acquired audio data to the server in real time. The audio data is sent in stream format using HTTP or WebSocket. The input here is the audio data acquired in step 1, and the output is the audio data sent to the server.

[0256] Step 3: Convert audio data to text

[0257] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. Specifically, the server calls the voice recognition engine's API, sends the voice data, and captures the returned text data. The input here is the voice data sent in step 2, and the output is the converted text data.

[0258] Step 4: Generate emotion data

[0259] The server inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotion. The server then sends the text data to the emotion engine's API and obtains the returned emotion data. Specifically, emotion labels such as "positive" and "negative" are generated. The input here is the text data converted in step 3, and the output is the generated emotion data.

[0260] Step 5: Extract and summarize key points

[0261] The server inputs text data and emotion data using a generative AI model (e.g., OpenAI GPT-3), extracts key points, and summarizes them. The server creates a prompt and sends it to the generative AI model. The prompt uses something like, "Please summarize the following conversation. Please also consider emotion analysis." The server receives a response from the generative AI model, extracts key points, and summarizes them. The input here is emotion data and text data, and the output is summarized text data.

[0262] Step 6: Generate the transcript

[0263] The server generates minutes from the summarized text data in the format specified by the user. The server creates documents using a template engine and outputs them in Word or PDF format. The input here is the summarized text data from step 5, and the output is the generated minutes.

[0264] Step 7: Providing and correcting minutes

[0265] The generated minutes are first provided to the user from the server. The user uses a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. The user highlights the parts to be corrected and edits intuitively. The input here is the minutes generated in step 6, and the output is the corrected minutes.

[0266] Step 8: Sending the minutes

[0267] The final minutes, once they have been corrected, are automatically sent from the server to the email addresses of the relevant parties. The server uses an email sending program (e.g., an SMTP server) to send the minutes as an attachment to the specified recipient. The input here is the minutes corrected in step 7, and the output is the minutes sent to the relevant parties.

[0268] (Application example 2)

[0269] 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."

[0270] Improving communication efficiency and accurate information sharing among staff and managers are key challenges at logistics centers. In particular, there is a lack of means to manage work instructions and contact information in real time, and to extract and quickly share important information. It is also necessary to understand the emotional state of staff while they are working and respond appropriately. This is essential to improve work efficiency and reduce errors.

[0271] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0272] In this invention, the server includes means for acquiring voice data in real time, means for converting the acquired voice data into text data, and means for extracting important points from the converted text data and emotion data and summarizing them in real time, thereby making it possible to improve the efficiency of communication in a logistics center, quickly share important information, and grasp the emotional state of staff.

[0273] "Audio data" refers to information that records audio in digital format.

[0274] "Text data" is information obtained by converting voice data into a string of characters.

[0275] "Emotion data" is information obtained by analyzing the user's emotional state from voice data and text data.

[0276] A "summary" is information that extracts important points from the converted text data and summarizes them concisely.

[0277] A "work report" is a document that records the progress and status of work, generated based on summarized text data.

[0278] "User" means a person or organization that uses the system to acquire voice data or check or modify text data.

[0279] A "stakeholder" is a person or organization that needs to receive the generated work report.

[0280] "Real-time" refers to processing or reaction occurring immediately, without delay.

[0281] This is a system that realizes efficient management and information sharing of work instructions and contact information in a logistics center. This system performs a series of processes: acquires voice data, converts it into text data, extracts important points including emotional data, generates a work report in real time, and sends it to the relevant parties after user corrections.

[0282] Hardware and Software Configuration

[0283] 1. Audio data acquisition method

[0284] The device uses a microphone to capture voice data, which is then transmitted to a server in real time. The device can be a wearable device such as smart glasses or a head-mounted display.

[0285] 2. Means of converting audio data into text data

[0286] After receiving the voice data, the server converts it into text data using a voice recognition engine, such as the speech_recognition library.

[0287] 3. Means of generating emotion data

[0288] The server inputs the transformed text data into a sentiment analysis model to generate sentiment data, using the sentiment analysis pipeline in the transformers library.

[0289] 4. Extraction and summarization of key points

[0290] The server inputs emotion data and text data into a generative AI model to extract key points. The generative AI model uses GPT-3 or a similar model, and the generated prompt is set as follows:

[0291] "Summarize the text below, highlighting the key points"

[0292] for example:

[0293] This is a message from our shipping department. There is a slight delay in our current schedule. It is taking some time to sort out the packages, but we expect the delay to be resolved by tomorrow morning.

[0294] 5. Means of generating work reports

[0295] Based on the summarized text data, the server automatically generates work reports in real time, which can be output in standard or custom formats.

[0296] 6. Means of correction and provision

[0297] The generated work report is provided to the user via the server, and the user can make corrections via a web interface. The corrections are updated in real time.

[0298] 7. Means of delivery to interested parties

[0299] Once the final corrections are completed, the work report is automatically sent from the server to the relevant parties.

[0300] Specific examples

[0301] For example, when a new package arrives at a logistics center, staff can use a head-mounted display to input instructions by voice. This converts the voice into text in real time, and important points are extracted along with emotional data. This content is then generated in real time as a work report, which managers can immediately check and correct, and finally send to the relevant parties. This improves the efficiency of information transmission and the transparency of work.

[0302] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0303] Step 1:

[0304] The terminal uses a microphone to capture voice data. The voice data includes instructions and messages given during work at the logistics center. The terminal transmits the captured voice data to the server in real time. The input is the real-time voice data, and the output is the voice data transmitted to the server.

[0305] Step 2:

[0306] The server inputs the received voice data into a voice recognition engine. The voice recognition engine (e.g., speech_recognition library) is used to convert the voice data into text data. The input is the voice data sent to the server, and the output is text data.

[0307] Step 3:

[0308] The server inputs the transformed text data into a sentiment analysis model. The sentiment analysis model (e.g., the transformers library) is used to generate sentiment data from the text data. The input is text data, and the output is text data containing sentiment data.

[0309] Step 4:

[0310] The server inputs emotion data and text data into a generative AI model. Using a generative AI model (e.g., GPT-3), it extracts key points and generates a summary. The input is text data including emotion data, and the output is summarized text data. The generation prompt is as follows: "Please summarize the following text, highlighting the key points."

[0311] Step 5:

[0312] The server generates a work report based on the summarized text data. The generated work report is automatically formatted in a format specified by the user. The input is the summarized text data, and the output is a work report in a standard or custom format.

[0313] Step 6:

[0314] The server provides the generated work report to the user, who can then review the work report via a web interface and make real-time edits as needed. The input is the generated work report, and the output is the work report edited by the user.

[0315] Step 7:

[0316] The server sends the revised work report to the relevant parties. The revised report is automatically sent to the specified email address. The input is the work report revised by the user, and the output is the final work report sent to the relevant parties.

[0317] 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.

[0318] 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.

[0319] 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.

[0320] [Second embodiment]

[0321] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0322] 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.

[0323] 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).

[0324] 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.

[0325] 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.

[0326] 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).

[0327] 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.

[0328] 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.

[0329] 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.

[0330] 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.

[0331] 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.

[0332] 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."

[0333] The present invention is a series of systems that captures conference audio data in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0334] Acquiring audio data

[0335] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0336] Converting audio data into text data

[0337] The server converts the received voice data into text data in real time using a voice recognition engine, which has the ability to recognize voice and convert it into text with high accuracy.

[0338] Extraction and summary of key points

[0339] The converted text data is processed on a server, where a generative AI model is used to extract key points, which are then summarized into concise text using an algorithm.

[0340] Generate meeting minutes

[0341] Based on the summarized text data, the server generates a transcript according to the format specified by the user, which can be output in a standard or custom format.

[0342] Provision and correction of minutes

[0343] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. These corrections can be easily made using a dedicated interface.

[0344] Sending minutes

[0345] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting.

[0346] Specific examples

[0347] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The generative AI model then summarizes the text as "Today we will review the progress of the project." Based on the summary text, minutes are generated in a standard format. The user can review and edit the minutes, and they are eventually automatically sent to all parties involved.

[0348] In this way, the present invention significantly improves the efficiency of taking minutes at meetings, saving time and effort.

[0349] The processing flow will be explained below.

[0350] Step 1:

[0351] The terminal acquires voice data using a microphone at the same time as the conference starts, and transmits the voice data to the server in real time.

[0352] Step 2:

[0353] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[0354] Step 3:

[0355] The server inputs the converted text data into a generative AI model, which extracts key points from the text data. The generative AI model uses natural language processing techniques to identify important statements and content.

[0356] Step 4:

[0357] The server then uses a summarization algorithm to create a summary of the key information based on the extracted key points. This summarization algorithm is designed to achieve both accuracy and conciseness.

[0358] Step 5:

[0359] The server generates minutes from the summarized text data in a format specified by the user, and outputs the minutes in a standard or custom format.

[0360] Step 6:

[0361] The server provides the generated minutes to the user, and allows the user to check the contents of the minutes using a dedicated interface and correct them as necessary.

[0362] Step 7:

[0363] After the user completes the edits, the server automatically sends the final minutes to the relevant email addresses, allowing all meeting participants to quickly share accurate minutes.

[0364] Example 1

[0365] 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."

[0366] While it is important to create meeting minutes accurately and quickly, traditional methods rely on manual processes, which are time-consuming and labor-intensive. Furthermore, it is not easy to check and correct minutes in real time and share them with relevant parties immediately, which can lead to delays in information transmission. For these reasons, there was a need to improve the efficiency and accuracy of the minutes creation process.

[0367] 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.

[0368] In this invention, the server includes a means for transmitting voice data to the server in real time, a means for converting voice data into text data using a voice recognition engine, and a means for extracting key points using a generative AI model. This enables the voice data of a meeting to be converted into text in real time, key points to be efficiently extracted and summarized, and minutes to be generated quickly. Furthermore, the generated minutes can be reviewed and revised in real time by the user and immediately sent to relevant parties, preventing delays in information transmission and significantly improving post-meeting work efficiency.

[0369] "Audio data" refers to data in which audio information from meetings, interviews, etc. is recorded in digital format.

[0370] "Text data" refers to data of character information generated from voice data using voice recognition technology or the like.

[0371] A "server" is a computer system that provides resources and data over a network.

[0372] A "terminal" is a device operated by a user, and includes a computer, laptop, tablet, smartphone, etc.

[0373] A "user" is someone who uses the system, and includes conference participants and administrators.

[0374] A "voice recognition engine" is software or hardware that analyzes voice data and converts it into corresponding text data.

[0375] A "generative AI model" is an artificial intelligence model that uses artificial intelligence technology to analyze data and perform tasks such as extracting key points and summarizing.

[0376] "Minutes" are text documents that record the contents of a meeting or the like.

[0377] "Real-time" refers to data processing and information transmission occurring almost instantaneously, with minimal delay.

[0378] A "prompt sentence" is an input sentence given to a generative AI model that contains instructions for the model to perform a specific task.

[0379] A "summary" is a sentence that succinctly summarizes the contents of the text data.

[0380] An "email sending API" is an interface for sending email from a program.

[0381] The present invention is a system that captures audio data of a meeting in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0382] When a conference begins, the device captures audio data using its built-in microphone or an external microphone. The captured audio data is temporarily stored in a buffer and then sent to the server in real time. The audio data is sent using the TCP / IP protocol.

[0383] The server instantly converts the received voice data into text data using a "voice recognition engine." This voice recognition engine can be a general-purpose tool such as "Google Cloud Speech-to-Text." The server sends the voice data to the engine via an API and saves the returned text data.

[0384] The converted text data is then used on the server to extract important points using the generative AI model "GPT-4." The generative AI model then uses prompts to perform specific tasks. Specifically, a prompt containing the instruction "Please summarize the contents of today's meeting" is used. An example of a prompt is "Convert the audio data from the meeting into text data, summarize the key points, and create minutes. The contents of the meeting discussion are listed below."

[0385] The extracted key points are summarized into concise text using a summarization algorithm, and the server generates minutes based on the summary text. The format of the minutes can be specified by the user, and they can be output in formats such as Markdown, PDF, Word, etc. For example, the server can use a template engine to convert the summary text into Markdown format and output it as PDF using a library such as "PDFKit."

[0386] The generated minutes are provided to the user by the server through a dedicated web interface, where the user can review the contents of the minutes and make any necessary corrections. This web interface has a function for editing the minutes text in real time. After the user has completed the corrections, the minutes are automatically sent to all parties involved from the server using an email sending API (e.g., "SendGrid" or "Amazon SES").

[0387] As a concrete example, when a project team holds a progress meeting, a laptop or tablet picks up the audio of the meeting and sends it to a server in real time. Google Cloud Speech-to-Text converts the audio into text, such as "Today we will discuss the progress of the project," and the generative AI model GPT-4 summarizes this as "Today we will review the progress of the project." The summarized text is converted to Markdown format and output as a PDF. The user can review and edit these minutes via a web interface, and they are finally automatically sent to all parties involved.

[0388] Thus, the present invention efficiently and automatically processes minutes of meetings, thereby saving a great deal of time and effort.

[0389] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0390] Step 1:

[0391] When a conference starts, the device captures audio data using a built-in microphone or an external microphone. The audio data is temporarily stored in a buffer in real time and then sent to the server using the TCP / IP protocol.

[0392] Specific example of operation: A laptop installed in a conference room picks up speech using its built-in microphone and transmits it to a server via Wi-Fi.

[0393] Step 2:

[0394] The server converts the voice data it receives in real time into text data using a "voice recognition engine." The voice data (input) is analyzed by the voice recognition software and the corresponding text data (output) is generated.

[0395] Specific example of operation: The server sends voice data to the "Google Cloud Speech-to-Text" service, receives the results as text data, and stores them in a database.

[0396] Step 3:

[0397] The server extracts and summarizes the key points from the converted text data using a "generative AI model." The text data (input) is sent to the AI ​​model with a prompt, and a text (output) summarizing the key points is generated.

[0398] Specific example of operation: GPT-4 is sent a prompt saying, "Convert the audio data of the meeting into text data and summarize the key points," and the AI ​​model outputs the summary text, "Today we will review the progress of the project."

[0399] Step 4:

[0400] The server generates minutes based on the summarized text data. It converts the summarized text (input) into a specified format (output) using a template engine.

[0401] Specific example of operation: Convert the summary text into Markdown format and output it as PDF using the "PDFKit" library.

[0402] Step 5:

[0403] The server provides the generated minutes to the user through a dedicated web interface. The PDF file (input) is displayed via the web interface, and the user checks its contents.

[0404] Specific example of operation: A user opens a web browser, accesses a dedicated URL, and checks the displayed minutes.

[0405] Step 6:

[0406] Users can modify the contents of the minutes through a web interface by directly editing the displayed text (input) and saving it as modified text (output).

[0407] Example of how it works: A user edits typos or additional information on the web interface and clicks the "Save" button, which sends the corrections to the server.

[0408] Step 7:

[0409] The server automatically sends the revised minutes to the relevant parties. The revised PDF file (input) is sent to the relevant parties' email addresses using the email sending API.

[0410] Example of specific operation: The server attaches the corrected minutes and uses the "SendGrid" API to send a message to the relevant email address with the message "The meeting minutes are complete."

[0411] In this way, the system efficiently executes a series of processes from acquiring audio data of a meeting to generating, editing, and distributing minutes.

[0412] (Application example 1)

[0413] 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."

[0414] Conventional conferencing systems require manual processes, from collecting audio data to converting it into text, extracting key points, and creating and sending meeting minutes, resulting in inefficiency and requiring a great deal of time and effort. Furthermore, in-factory meetings also faced difficulties in generating and sharing work instructions and reports in real time, reducing work efficiency. Furthermore, there was a risk that important points would be overlooked.

[0415] 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.

[0416] In this invention, the server includes a means for acquiring voice data, a means for converting the acquired voice data into text data, and a means for extracting and summarizing important points from the converted text data. This automatically converts voice data into text, making it possible to summarize important points and generate minutes, work instructions, and reports. In addition, the generated documents can be modified and confirmed by the user, enabling highly accurate information sharing.

[0417] "Audio data" refers to audio recorded and stored in digital format from meetings, work sites, etc.

[0418] "Text data" is digital data that has been analyzed and converted into text information.

[0419] "Key points" are pieces of information that indicate important points or key points within the collected data.

[0420] A "generative AI model" is an artificial intelligence algorithm that generates or summarizes new information based on machine learning and data processing.

[0421] "Summarizing" means to concisely summarize a large amount of information and extract and organize only the main points and main points.

[0422] Minutes are documents that record the contents of meetings and discussions.

[0423] "Correction" refers to the process of changing the content of generated documents or data to improve their accuracy and clarity.

[0424] "Stakeholders" are people who need to share information or work instructions in a particular meeting or task.

[0425] A "work instruction" is a document that provides specific instructions on how to perform a particular task or work.

[0426] A "report" is a formal document used to record and report a particular situation or work process.

[0427] The system of this invention is designed to capture voice data from meetings and factories in real time, automatically convert it into text, and then extract and summarize important points to generate minutes, work instructions, and reports.The main components of this system include a terminal that captures the voice data, a server that processes the data, and a user interface for making corrections.

[0428] Hardware / Software Configuration

[0429] Device:

[0430] This device is equipped with a microphone to capture voice data during meetings or in factories. The voice data is captured in real time and sent to a server.

[0431] server:

[0432] Google Cloud Speech-to-Text is used to convert voice data into text data. This text data is processed internally on the server, and key points are extracted and summarized. OpenAI's GPT-3 is used as a generative AI model to generate summaries and instructions.

[0433] Data processing / data calculation:

[0434] 1. Acquisition and transmission of audio data:

[0435] It uses the device's microphone to collect voice data and transmits it to a server in real time.

[0436] 2. Audio to text conversion:

[0437] Using the Google Cloud Speech-to-Text API, voice data is converted into text with high accuracy.

[0438] 3. Extract and summarize key points:

[0439] The text data is analyzed internally on the server, and key points are extracted and summarized using OpenAI GPT-3's generative AI model.

[0440] 4. Generate minutes, work instructions, and reports:

[0441] Automatically generate meeting minutes, work instructions, and reports based on summarized text data, and output these documents in standard or customizable formats.

[0442] 5. User Modifications:

[0443] The generated document is presented to the user through a dedicated interface that allows for modification, and supports real-time modification.

[0444] Specific examples

[0445] For example, in a factory progress meeting, a device captures the audio data of the meeting in real time and sends it to a server. The server converts the audio data into text data such as "Today we will discuss the importance of maintaining machine A." Next, a generative AI model is used to summarize the text as "Maintenance of machine A is important." A work instruction is generated based on the summarized text. The user can then review and revise the work instruction, and it is finally automatically sent to all parties involved.

[0446] Example prompts to input to a generative AI model:

[0447] Summarize the following text:

[0448] Today we will discuss the importance of maintaining Machine A.

[0449] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0450] Step 1:

[0451] The terminal captures voice data in real time during meetings and factories. The terminal's built-in microphone collects the voice data in digital format. The collected voice data is then sent to a server in real time.

[0452] Input: Audio from meetings and factories

[0453] Output: Digital audio data (sent to server)

[0454] Step 2:

[0455] The server converts the acquired voice data into text data using the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives the returned text data.

[0456] Input: Digital audio data

[0457] Output: Text data (converted from audio)

[0458] Step 3:

[0459] The server analyzes the text data and extracts and summarizes key points using a generative AI model (OpenAI GPT-3). First, the text data is input to the generative AI model as a prompt sentence, and the summarized text is obtained.

[0460] Input: Text data (records of meetings and work)

[0461] Output: Summarized text data (key points)

[0462] Step 4:

[0463] The server generates meeting minutes, work instructions, and reports in a user-specified format based on the summarized text data, and the generated documents can be output in standard or customized formats.

[0464] Input: Summarized text data

[0465] Output: minutes, work instructions, reports (user-specified format)

[0466] Step 5:

[0467] The generated document is provided to the user from the server, who can review the document and make corrections as needed through a dedicated interface that supports real-time corrections.

[0468] Input: Minutes, work orders, reports

[0469] Output: Corrected document (user confirmation and correction)

[0470] Step 6:

[0471] The server automatically sends the corrected document to the relevant person's email address.

[0472] Input: revised document

[0473] Output: Send to relevant parties (via email)

[0474] Specifically, the system begins by capturing the audio of the meeting in real time and sending it to the server, after which a series of processes are automatically carried out, including converting the audio data into text, extracting important points, generating various documents, allowing the user to check and correct them, and finally sending them.

[0475] 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.

[0476] The present invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions.

[0477] Acquiring audio data

[0478] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0479] Converting audio data into text data

[0480] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[0481] Generating emotional data using an emotion engine

[0482] The server inputs the converted text data into an emotion engine to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of voice and the choice of words used, and emotion data is generated.

[0483] Extraction and summary of key points

[0484] The converted text data, along with the emotion data, is then fed into a generative AI model, which the server uses to extract key points and, if necessary, highlight parts that emphasize the user's emotions.The server then uses a summarization algorithm to compile the extracted key points into concise text.

[0485] Generate meeting minutes

[0486] Based on the summarized text data, the server generates minutes according to the format specified by the user, and the generated minutes can be output in a standard or custom format.

[0487] Provision and correction of minutes

[0488] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. Corrections can be easily made using a dedicated interface.

[0489] Sending minutes

[0490] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting ends.

[0491] Specific examples

[0492] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The emotion engine then analyzes emotions such as "excitement" or "anxiety" from the text data and summarizes it as "Today we will review the progress of the project. - (excited)." The parts that emphasize the user's emotions are highlighted and recorded in the minutes.

[0493] The server then generates minutes in a standard format based on this summary text. The user can then make any necessary corrections, and the minutes are finally sent automatically to all parties involved. In this way, the present invention reduces the burden on meetings and enables efficient and accurate minutes creation. Furthermore, the introduction of an emotion engine makes it possible to generate minutes that reflect the emotional state of participants, thereby providing a deeper understanding of the progress of the meeting.

[0494] The processing flow will be explained below.

[0495] Step 1:

[0496] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0497] Step 2:

[0498] The server inputs the received voice data into a speech recognition engine, which performs highly accurate speech recognition and converts the voice data into text. This speech recognition engine recognizes the voices of multiple speakers in the conference and transcribes each individual statement into text.

[0499] Step 3:

[0500] The server inputs the converted text data into an emotion engine, which analyzes the word choice, context, and speaker's tone in the text data to recognize the user's emotions in real time and generate emotion data. This emotion data includes emotion labels such as "happiness," "anxiety," and "excitement."

[0501] Step 4:

[0502] The server combines the generated emotion data with the text data and inputs it into a generative AI model, which extracts key points from the text and identifies parts that emphasize the user's emotions.

[0503] Step 5:

[0504] The server extracts key points and uses a summarization algorithm to create a concise summary text, highlighting parts that emphasize specific emotions based on the emotion data.

[0505] Step 6:

[0506] The server generates minutes from the summarized text data in a format specified by the user, and the generated minutes can be output in a standard or custom format.

[0507] Step 7:

[0508] The generated minutes are provided to the user from the server. The user can review the minutes using a dedicated interface and make corrections as necessary. The minutes also include highlights of the user's emotions, making it easier to understand the nuances of the meeting.

[0509] Step 8:

[0510] Once the user has completed the revisions, the server automatically sends the final minutes to the relevant parties' email addresses, including sentiment data and key points from the meeting, so that all meeting participants can quickly share them.

[0511] Specific examples

[0512] For example, suppose someone says, "Today we will discuss the progress of the project" during a project status meeting. The device captures this utterance and sends it to the server in real time. The server converts the speech into text data, "Today we will discuss the progress of the project." If the emotion engine identifies the emotion "excited," the server includes this emotion data. Next, the generative AI model extracts the key point, "Checking the progress of the project," and generates the summary text, "Today we will check the progress of the project. - (Excited)."

[0513] Based on this summary, minutes are created and provided to the user, who can then make any necessary corrections, and the final minutes are sent to all parties involved, allowing for quick sharing of important meeting content and related emotions.

[0514] Example 2

[0515] 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."

[0516] Conventional meeting minutes creation systems require a great deal of time and effort to capture audio data, convert it to text data, extract key points, and generate minutes, significantly reducing meeting efficiency. Furthermore, minutes created do not take into account the emotions of meeting participants, making it difficult to gain a deep understanding of the progress of a meeting.

[0517] The specification process by the specification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for transmitting voice data to the server in real time, means for inputting the received voice data into a voice recognition engine and converting it into text data, means for inputting the converted text data into an emotion engine and recognizing the user's emotion to generate emotion data, and means for inputting the generated emotion data and text data into a generative AI model and extracting and summarizing important points. This makes it possible to efficiently generate minutes of a meeting from the voice data in real time and provide accurate minutes that reflect the emotions of the participants.

[0518] "Audio data" refers to data that records audio in digital format.

[0519] A "server" refers to a computer system that sends, receives, and processes data over a network.

[0520] "Terminal" refers to a device such as a computer or smartphone that can be directly operated by a user.

[0521] A "voice recognition engine" refers to software or a system that analyzes voice data and converts it into text data.

[0522] "Text data" refers to data that records character information in digital format.

[0523] An "emotion engine" refers to software or a system that analyzes text data or voice tone to generate emotional data.

[0524] "Emotion data" refers to data that indicates the user's emotional state as analyzed by the emotion engine.

[0525] A "generative AI model" refers to an artificial intelligence model that performs natural language processing based on large amounts of data and generates new text.

[0526] "Extracting important points" refers to finding particularly important content or key points from text data or emotion data.

[0527] "Summarization" refers to shortening long text data to include only the main information.

[0528] Minutes are a document that records what was discussed at a meeting and can be referenced at a later date.

[0529] "Real-time" refers to a state in which data processing and communication occurs almost instantaneously.

[0530] "Providing to users" means making the data and information generated by the system accessible and usable by users.

[0531] "Modify" refers to changing the content of data or documents that have already been generated and stored.

[0532] "Send" refers to sending the generated or modified data or documents to other users or systems involved.

[0533] This invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions. This system functions efficiently, with the terminal, server, and user each playing their own role.

[0534] As soon as the conference starts, the device uses a microphone to capture audio data. The captured audio data is sent to the server in real time. Specifically, audio capture software installed on the device (e.g., a Windows audio recording application) records the audio in real time and sends it in a stream format to the server using HTTP or WebSocket.

[0535] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. In this conversion process, the server calls the API of the voice recognition engine, sends the voice data, and captures the returned text data.

[0536] The server then inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of the voice and the choice of words used, and emotion data is generated.

[0537] The server then inputs the generated emotion data and converted text data into a generative AI model (e.g., OpenAI GPT-3) to extract key points. The server then creates a prompt and sends it to the generative AI model. For example, the prompt might include something like, "Please summarize the following conversation. Please also consider sentiment analysis." The server then receives the response from the generative AI model, extracts key points, and summarizes them.

[0538] Based on the summarized text data, the server generates minutes in the format specified by the user. The generated minutes are output in a standard or custom format. The user can use a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. This correction process is intuitive and easy to perform.

[0539] The minutes, once they have been finalized, are automatically sent from the server to the email addresses of the relevant parties. Specifically, the server uses an email sending program (e.g., an SMTP server) to send the minutes to the specified recipients as an attachment.

[0540] Specific examples

[0541] For example, suppose a project team is holding a progress meeting. The device records the meeting audio in real time using its built-in microphone and sends it to the server via HTTPS. The server then uses the Google Cloud Speech-to-Text API to convert the audio into text data, such as "Today we will discuss the progress of the project." It then uses an emotion engine (IBM Watson Tone Analyzer) to recognize emotions such as "excitement" or "anxiety" from the text data. A prompt is then input into a generative AI model (OpenAI GPT-3), which extracts key points and summarizes them, taking into account the emotion data. For example, the summary might be something like "Today we will review the progress of the project. - (excited)." Based on the generated summary text, the server generates meeting minutes in standard PDF format or a custom format. The user can review the minutes in a web browser and make any necessary corrections. The server then automatically sends them to all parties involved via an SMTP server.

[0542] Prompt Sentence Examples

[0543] "Generate minutes by summarizing what was said in project team status meetings and adding sentiment analysis."

[0544] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0545] Step 1: Acquire audio data

[0546] The device starts capturing audio data using a microphone as soon as the conference starts. Audio can be recorded in real time using a built-in microphone or a connected microphone, or using audio capture software (e.g., Windows audio recording application). The input here is ambient audio, and the output is captured digital audio data.

[0547] Step 2: Sending audio data

[0548] The device sends the acquired audio data to the server in real time. The audio data is sent in stream format using HTTP or WebSocket. The input here is the audio data acquired in step 1, and the output is the audio data sent to the server.

[0549] Step 3: Convert audio data to text

[0550] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. Specifically, the server calls the voice recognition engine's API, sends the voice data, and captures the returned text data. The input here is the voice data sent in step 2, and the output is the converted text data.

[0551] Step 4: Generate emotion data

[0552] The server inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotion. The server then sends the text data to the emotion engine's API and obtains the returned emotion data. Specifically, emotion labels such as "positive" and "negative" are generated. The input here is the text data converted in step 3, and the output is the generated emotion data.

[0553] Step 5: Extract and summarize key points

[0554] The server inputs text data and emotion data using a generative AI model (e.g., OpenAI GPT-3), extracts key points, and summarizes them. The server creates a prompt and sends it to the generative AI model. The prompt uses something like, "Please summarize the following conversation. Please also consider emotion analysis." The server receives a response from the generative AI model, extracts key points, and summarizes them. The input here is emotion data and text data, and the output is summarized text data.

[0555] Step 6: Generate the transcript

[0556] The server generates minutes from the summarized text data in the format specified by the user. The server creates documents using a template engine and outputs them in Word or PDF format. The input here is the summarized text data from step 5, and the output is the generated minutes.

[0557] Step 7: Providing and correcting minutes

[0558] The generated minutes are first provided to the user from the server. The user uses a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. The user highlights the parts to be corrected and edits intuitively. The input here is the minutes generated in step 6, and the output is the corrected minutes.

[0559] Step 8: Sending the minutes

[0560] The final minutes, once they have been corrected, are automatically sent from the server to the email addresses of the relevant parties. The server uses an email sending program (e.g., an SMTP server) to send the minutes as an attachment to the specified recipient. The input here is the minutes corrected in step 7, and the output is the minutes sent to the relevant parties.

[0561] (Application example 2)

[0562] 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."

[0563] Improving communication efficiency and accurate information sharing among staff and managers are key challenges at logistics centers. In particular, there is a lack of means to manage work instructions and contact information in real time, and to extract and quickly share important information. It is also necessary to understand the emotional state of staff while they are working and respond appropriately. This is essential to improve work efficiency and reduce errors.

[0564] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0565] In this invention, the server includes means for acquiring voice data in real time, means for converting the acquired voice data into text data, and means for extracting important points from the converted text data and emotion data and summarizing them in real time, thereby making it possible to improve the efficiency of communication in a logistics center, quickly share important information, and grasp the emotional state of staff.

[0566] "Audio data" refers to information that records audio in digital format.

[0567] "Text data" is information obtained by converting voice data into a string of characters.

[0568] "Emotion data" is information obtained by analyzing the user's emotional state from voice data and text data.

[0569] A "summary" is information that extracts important points from the converted text data and summarizes them concisely.

[0570] A "work report" is a document that records the progress and status of work, generated based on summarized text data.

[0571] "User" means a person or organization that uses the system to acquire voice data or check or modify text data.

[0572] A "stakeholder" is a person or organization that needs to receive the generated work report.

[0573] "Real-time" refers to processing or reaction occurring immediately, without delay.

[0574] This is a system that realizes efficient management and information sharing of work instructions and contact information in a logistics center. This system performs a series of processes: acquires voice data, converts it into text data, extracts important points including emotional data, generates a work report in real time, and sends it to the relevant parties after user corrections.

[0575] Hardware and Software Configuration

[0576] 1. Audio data acquisition method

[0577] The device uses a microphone to capture voice data, which is then transmitted to a server in real time. The device can be a wearable device such as smart glasses or a head-mounted display.

[0578] 2. Means of converting audio data into text data

[0579] After receiving the voice data, the server converts it into text data using a voice recognition engine, such as the speech_recognition library.

[0580] 3. Means of generating emotion data

[0581] The server inputs the transformed text data into a sentiment analysis model to generate sentiment data, using the sentiment analysis pipeline in the transformers library.

[0582] 4. Extraction and summarization of key points

[0583] The server inputs emotion data and text data into a generative AI model to extract key points. The generative AI model uses GPT-3 or a similar model, and the generated prompt is set as follows:

[0584] "Summarize the text below, highlighting the key points"

[0585] for example:

[0586] This is a message from our shipping department. There is a slight delay in our current schedule. It is taking some time to sort out the packages, but we expect the delay to be resolved by tomorrow morning.

[0587] 5. Means of generating work reports

[0588] Based on the summarized text data, the server automatically generates work reports in real time, which can be output in standard or custom formats.

[0589] 6. Means of correction and provision

[0590] The generated work report is provided to the user via the server, and the user can make corrections via a web interface. The corrections are updated in real time.

[0591] 7. Means of delivery to interested parties

[0592] Once the final corrections are completed, the work report is automatically sent from the server to the relevant parties.

[0593] Specific examples

[0594] For example, when a new package arrives at a logistics center, staff can use a head-mounted display to input instructions by voice. This converts the voice into text in real time, and important points are extracted along with emotional data. This content is then generated in real time as a work report, which managers can immediately check and correct, and finally send to the relevant parties. This improves the efficiency of information transmission and the transparency of work.

[0595] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0596] Step 1:

[0597] The terminal uses a microphone to capture voice data. The voice data includes instructions and messages given during work at the logistics center. The terminal transmits the captured voice data to the server in real time. The input is the real-time voice data, and the output is the voice data transmitted to the server.

[0598] Step 2:

[0599] The server inputs the received voice data into a voice recognition engine. The voice recognition engine (e.g., speech_recognition library) is used to convert the voice data into text data. The input is the voice data sent to the server, and the output is text data.

[0600] Step 3:

[0601] The server inputs the transformed text data into a sentiment analysis model. The sentiment analysis model (e.g., the transformers library) is used to generate sentiment data from the text data. The input is text data, and the output is text data containing sentiment data.

[0602] Step 4:

[0603] The server inputs emotion data and text data into a generative AI model. Using a generative AI model (e.g., GPT-3), it extracts key points and generates a summary. The input is text data including emotion data, and the output is summarized text data. The generation prompt is as follows: "Please summarize the following text, highlighting the key points."

[0604] Step 5:

[0605] The server generates a work report based on the summarized text data. The generated work report is automatically formatted in a format specified by the user. The input is the summarized text data, and the output is a work report in a standard or custom format.

[0606] Step 6:

[0607] The server provides the generated work report to the user, who can then review the work report via a web interface and make real-time edits as needed. The input is the generated work report, and the output is the work report edited by the user.

[0608] Step 7:

[0609] The server sends the revised work report to the relevant parties. The revised report is automatically sent to the specified email address. The input is the work report revised by the user, and the output is the final work report sent to the relevant parties.

[0610] 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.

[0611] 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.

[0612] 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.

[0613] [Third embodiment]

[0614] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0615] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0616] 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).

[0617] 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.

[0618] 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.

[0619] 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).

[0620] 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.

[0621] 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.

[0622] 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.

[0623] 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.

[0624] 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.

[0625] 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."

[0626] The present invention is a series of systems that captures conference audio data in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0627] Acquiring audio data

[0628] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0629] Converting audio data into text data

[0630] The server converts the received voice data into text data in real time using a voice recognition engine, which has the ability to recognize voice and convert it into text with high accuracy.

[0631] Extraction and summary of key points

[0632] The converted text data is processed on a server, where a generative AI model is used to extract key points, which are then summarized into concise text using an algorithm.

[0633] Generate meeting minutes

[0634] Based on the summarized text data, the server generates a transcript according to the format specified by the user, which can be output in a standard or custom format.

[0635] Provision and correction of minutes

[0636] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. These corrections can be easily made using a dedicated interface.

[0637] Sending minutes

[0638] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting.

[0639] Specific examples

[0640] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The generative AI model then summarizes the text as "Today we will review the progress of the project." Based on the summary text, minutes are generated in a standard format. The user can review and edit the minutes, and they are eventually automatically sent to all parties involved.

[0641] In this way, the present invention significantly improves the efficiency of taking minutes at meetings, saving time and effort.

[0642] The processing flow will be explained below.

[0643] Step 1:

[0644] The terminal acquires voice data using a microphone at the same time as the conference starts, and transmits the voice data to the server in real time.

[0645] Step 2:

[0646] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[0647] Step 3:

[0648] The server inputs the converted text data into a generative AI model, which extracts key points from the text data. The generative AI model uses natural language processing techniques to identify important statements and content.

[0649] Step 4:

[0650] The server then uses a summarization algorithm to create a summary of the key information based on the extracted key points. This summarization algorithm is designed to achieve both accuracy and conciseness.

[0651] Step 5:

[0652] The server generates minutes from the summarized text data in a format specified by the user, and outputs the minutes in a standard or custom format.

[0653] Step 6:

[0654] The server provides the generated minutes to the user, and allows the user to check the contents of the minutes using a dedicated interface and correct them as necessary.

[0655] Step 7:

[0656] After the user completes the edits, the server automatically sends the final minutes to the relevant email addresses, allowing all meeting participants to quickly share accurate minutes.

[0657] Example 1

[0658] 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."

[0659] While it is important to create meeting minutes accurately and quickly, traditional methods rely on manual processes, which are time-consuming and labor-intensive. Furthermore, it is not easy to check and correct minutes in real time and share them with relevant parties immediately, which can lead to delays in information transmission. For these reasons, there was a need to improve the efficiency and accuracy of the minutes creation process.

[0660] 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.

[0661] In this invention, the server includes a means for transmitting voice data to the server in real time, a means for converting voice data into text data using a voice recognition engine, and a means for extracting key points using a generative AI model. This enables the voice data of a meeting to be converted into text in real time, key points to be efficiently extracted and summarized, and minutes to be generated quickly. Furthermore, the generated minutes can be reviewed and revised in real time by the user and immediately sent to relevant parties, preventing delays in information transmission and significantly improving post-meeting work efficiency.

[0662] "Audio data" refers to data in which audio information from meetings, interviews, etc. is recorded in digital format.

[0663] "Text data" refers to data of character information generated from voice data using voice recognition technology or the like.

[0664] A "server" is a computer system that provides resources and data over a network.

[0665] A "terminal" is a device operated by a user, and includes a computer, laptop, tablet, smartphone, etc.

[0666] A "user" is someone who uses the system, and includes conference participants and administrators.

[0667] A "voice recognition engine" is software or hardware that analyzes voice data and converts it into corresponding text data.

[0668] A "generative AI model" is an artificial intelligence model that uses artificial intelligence technology to analyze data and perform tasks such as extracting key points and summarizing.

[0669] "Minutes" are text documents that record the contents of a meeting or the like.

[0670] "Real-time" refers to data processing and information transmission occurring almost instantaneously, with minimal delay.

[0671] A "prompt sentence" is an input sentence given to a generative AI model that contains instructions for the model to perform a specific task.

[0672] A "summary" is a sentence that succinctly summarizes the contents of the text data.

[0673] An "email sending API" is an interface for sending email from a program.

[0674] The present invention is a system that captures audio data of a meeting in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0675] When a conference begins, the device captures audio data using its built-in microphone or an external microphone. The captured audio data is temporarily stored in a buffer and then sent to the server in real time. The audio data is sent using the TCP / IP protocol.

[0676] The server instantly converts the received voice data into text data using a "voice recognition engine." This voice recognition engine can be a general-purpose tool such as "Google Cloud Speech-to-Text." The server sends the voice data to the engine via an API and saves the returned text data.

[0677] The converted text data is then used on the server to extract important points using the generative AI model "GPT-4." The generative AI model then uses prompts to perform specific tasks. Specifically, a prompt containing the instruction "Please summarize the contents of today's meeting" is used. An example of a prompt is "Convert the audio data from the meeting into text data, summarize the key points, and create minutes. The contents of the meeting discussion are listed below."

[0678] The extracted key points are summarized into concise text using a summarization algorithm, and the server generates minutes based on the summary text. The format of the minutes can be specified by the user, and they can be output in formats such as Markdown, PDF, Word, etc. For example, the server can use a template engine to convert the summary text into Markdown format and output it as PDF using a library such as "PDFKit."

[0679] The generated minutes are provided to the user by the server through a dedicated web interface, where the user can review the contents of the minutes and make any necessary corrections. This web interface has a function for editing the minutes text in real time. After the user has completed the corrections, the minutes are automatically sent to all parties involved from the server using an email sending API (e.g., "SendGrid" or "Amazon SES").

[0680] As a concrete example, when a project team holds a progress meeting, a laptop or tablet picks up the audio of the meeting and sends it to a server in real time. Google Cloud Speech-to-Text converts the audio into text, such as "Today we will discuss the progress of the project," and the generative AI model GPT-4 summarizes this as "Today we will review the progress of the project." The summarized text is converted to Markdown format and output as a PDF. The user can review and edit these minutes via a web interface, and they are finally automatically sent to all parties involved.

[0681] Thus, the present invention efficiently and automatically processes minutes of meetings, thereby saving a great deal of time and effort.

[0682] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0683] Step 1:

[0684] When a conference starts, the device captures audio data using a built-in microphone or an external microphone. The audio data is temporarily stored in a buffer in real time and then sent to the server using the TCP / IP protocol.

[0685] Specific example of operation: A laptop installed in a conference room picks up speech using its built-in microphone and transmits it to a server via Wi-Fi.

[0686] Step 2:

[0687] The server converts the voice data it receives in real time into text data using a "voice recognition engine." The voice data (input) is analyzed by the voice recognition software and the corresponding text data (output) is generated.

[0688] Specific example of operation: The server sends voice data to the "Google Cloud Speech-to-Text" service, receives the results as text data, and stores them in a database.

[0689] Step 3:

[0690] The server extracts and summarizes the key points from the converted text data using a "generative AI model." The text data (input) is sent to the AI ​​model with a prompt, and a text (output) summarizing the key points is generated.

[0691] Specific example of operation: GPT-4 is sent a prompt saying, "Convert the audio data of the meeting into text data and summarize the key points," and the AI ​​model outputs the summary text, "Today we will review the progress of the project."

[0692] Step 4:

[0693] The server generates minutes based on the summarized text data. It converts the summarized text (input) into a specified format (output) using a template engine.

[0694] Specific example of operation: Convert the summary text into Markdown format and output it as PDF using the "PDFKit" library.

[0695] Step 5:

[0696] The server provides the generated minutes to the user through a dedicated web interface. The PDF file (input) is displayed via the web interface, and the user checks its contents.

[0697] Specific example of operation: A user opens a web browser, accesses a dedicated URL, and checks the displayed minutes.

[0698] Step 6:

[0699] Users can modify the contents of the minutes through a web interface by directly editing the displayed text (input) and saving it as modified text (output).

[0700] Example of how it works: A user edits typos or additional information on the web interface and clicks the "Save" button, which sends the corrections to the server.

[0701] Step 7:

[0702] The server automatically sends the revised minutes to the relevant parties. The revised PDF file (input) is sent to the relevant parties' email addresses using the email sending API.

[0703] Example of specific operation: The server attaches the corrected minutes and uses the "SendGrid" API to send a message to the relevant email address with the message "The meeting minutes are complete."

[0704] In this way, the system efficiently executes a series of processes from acquiring audio data of a meeting to generating, editing, and distributing minutes.

[0705] (Application example 1)

[0706] 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."

[0707] Conventional conferencing systems require manual processes, from collecting audio data to converting it into text, extracting key points, and creating and sending meeting minutes, resulting in inefficiency and requiring a great deal of time and effort. Furthermore, in-factory meetings also faced difficulties in generating and sharing work instructions and reports in real time, reducing work efficiency. Furthermore, there was a risk that important points would be overlooked.

[0708] 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.

[0709] In this invention, the server includes a means for acquiring voice data, a means for converting the acquired voice data into text data, and a means for extracting and summarizing important points from the converted text data. This automatically converts voice data into text, making it possible to summarize important points and generate minutes, work instructions, and reports. In addition, the generated documents can be modified and confirmed by the user, enabling highly accurate information sharing.

[0710] "Audio data" refers to audio recorded and stored in digital format from meetings, work sites, etc.

[0711] "Text data" is digital data that has been analyzed and converted into text information.

[0712] "Key points" are pieces of information that indicate important points or key points within the collected data.

[0713] A "generative AI model" is an artificial intelligence algorithm that generates or summarizes new information based on machine learning and data processing.

[0714] "Summarizing" means to concisely summarize a large amount of information and extract and organize only the main points and main points.

[0715] Minutes are documents that record the contents of meetings and discussions.

[0716] "Correction" refers to the process of changing the content of generated documents or data to improve their accuracy and clarity.

[0717] "Stakeholders" are people who need to share information or work instructions in a particular meeting or task.

[0718] A "work instruction" is a document that provides specific instructions on how to perform a particular task or work.

[0719] A "report" is a formal document used to record and report a particular situation or work process.

[0720] The system of this invention is designed to capture voice data from meetings and factories in real time, automatically convert it into text, and then extract and summarize important points to generate minutes, work instructions, and reports.The main components of this system include a terminal that captures the voice data, a server that processes the data, and a user interface for making corrections.

[0721] Hardware / Software Configuration

[0722] Device:

[0723] This device is equipped with a microphone to capture voice data during meetings or in factories. The voice data is captured in real time and sent to a server.

[0724] server:

[0725] Google Cloud Speech-to-Text is used to convert voice data into text data. This text data is processed internally on the server, and key points are extracted and summarized. OpenAI's GPT-3 is used as a generative AI model to generate summaries and instructions.

[0726] Data processing / data calculation:

[0727] 1. Acquisition and transmission of audio data:

[0728] It uses the device's microphone to collect voice data and transmits it to a server in real time.

[0729] 2. Audio to text conversion:

[0730] Using the Google Cloud Speech-to-Text API, voice data is converted into text with high accuracy.

[0731] 3. Extract and summarize key points:

[0732] The text data is analyzed internally on the server, and key points are extracted and summarized using OpenAI GPT-3's generative AI model.

[0733] 4. Generate minutes, work instructions, and reports:

[0734] Automatically generate meeting minutes, work instructions, and reports based on summarized text data, and output these documents in standard or customizable formats.

[0735] 5. User Modifications:

[0736] The generated document is presented to the user through a dedicated interface that allows for modification, and supports real-time modification.

[0737] Specific examples

[0738] For example, in a factory progress meeting, a device captures the audio data of the meeting in real time and sends it to a server. The server converts the audio data into text data such as "Today we will discuss the importance of maintaining machine A." Next, a generative AI model is used to summarize the text as "Maintenance of machine A is important." A work instruction is generated based on the summarized text. The user can then review and revise the work instruction, and it is finally automatically sent to all parties involved.

[0739] Example prompts to input to a generative AI model:

[0740] Summarize the following text:

[0741] Today we will discuss the importance of maintaining Machine A.

[0742] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0743] Step 1:

[0744] The terminal captures voice data in real time during meetings and factories. The terminal's built-in microphone collects the voice data in digital format. The collected voice data is then sent to a server in real time.

[0745] Input: Audio from meetings and factories

[0746] Output: Digital audio data (sent to server)

[0747] Step 2:

[0748] The server converts the acquired voice data into text data using the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives the returned text data.

[0749] Input: Digital audio data

[0750] Output: Text data (converted from audio)

[0751] Step 3:

[0752] The server analyzes the text data and extracts and summarizes key points using a generative AI model (OpenAI GPT-3). First, the text data is input to the generative AI model as a prompt sentence, and the summarized text is obtained.

[0753] Input: Text data (records of meetings and work)

[0754] Output: Summarized text data (key points)

[0755] Step 4:

[0756] The server generates meeting minutes, work instructions, and reports in a user-specified format based on the summarized text data, and the generated documents can be output in standard or customized formats.

[0757] Input: Summarized text data

[0758] Output: minutes, work instructions, reports (user-specified format)

[0759] Step 5:

[0760] The generated document is provided to the user from the server, who can review the document and make corrections as needed through a dedicated interface that supports real-time corrections.

[0761] Input: Minutes, work orders, reports

[0762] Output: Corrected document (user confirmation and correction)

[0763] Step 6:

[0764] The server automatically sends the corrected document to the relevant person's email address.

[0765] Input: revised document

[0766] Output: Send to relevant parties (via email)

[0767] Specifically, the system begins by capturing the audio of the meeting in real time and sending it to the server, after which a series of processes are automatically carried out, including converting the audio data into text, extracting important points, generating various documents, allowing the user to check and correct them, and finally sending them.

[0768] 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.

[0769] The present invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions.

[0770] Acquiring audio data

[0771] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0772] Converting audio data into text data

[0773] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[0774] Generating emotional data using an emotion engine

[0775] The server inputs the converted text data into an emotion engine to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of voice and the choice of words used, and emotion data is generated.

[0776] Extraction and summary of key points

[0777] The converted text data, along with the emotion data, is then fed into a generative AI model, which the server uses to extract key points and, if necessary, highlight parts that emphasize the user's emotions.The server then uses a summarization algorithm to compile the extracted key points into concise text.

[0778] Generate meeting minutes

[0779] Based on the summarized text data, the server generates minutes according to the format specified by the user, and the generated minutes can be output in a standard or custom format.

[0780] Provision and correction of minutes

[0781] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. Corrections can be easily made using a dedicated interface.

[0782] Sending minutes

[0783] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting ends.

[0784] Specific examples

[0785] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The emotion engine then analyzes emotions such as "excitement" or "anxiety" from the text data and summarizes it as "Today we will review the progress of the project. - (excited)." The parts that emphasize the user's emotions are highlighted and recorded in the minutes.

[0786] The server then generates minutes in a standard format based on this summary text. The user can then make any necessary corrections, and the minutes are finally sent automatically to all parties involved. In this way, the present invention reduces the burden on meetings and enables efficient and accurate minutes creation. Furthermore, the introduction of an emotion engine makes it possible to generate minutes that reflect the emotional state of participants, thereby providing a deeper understanding of the progress of the meeting.

[0787] The processing flow will be explained below.

[0788] Step 1:

[0789] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0790] Step 2:

[0791] The server inputs the received voice data into a speech recognition engine, which performs highly accurate speech recognition and converts the voice data into text. This speech recognition engine recognizes the voices of multiple speakers in the conference and transcribes each individual statement into text.

[0792] Step 3:

[0793] The server inputs the converted text data into an emotion engine, which analyzes the word choice, context, and speaker's tone in the text data to recognize the user's emotions in real time and generate emotion data. This emotion data includes emotion labels such as "happiness," "anxiety," and "excitement."

[0794] Step 4:

[0795] The server combines the generated emotion data with the text data and inputs it into a generative AI model, which extracts key points from the text and identifies parts that emphasize the user's emotions.

[0796] Step 5:

[0797] The server extracts key points and uses a summarization algorithm to create a concise summary text, highlighting parts that emphasize specific emotions based on the emotion data.

[0798] Step 6:

[0799] The server generates minutes from the summarized text data in a format specified by the user, and the generated minutes can be output in a standard or custom format.

[0800] Step 7:

[0801] The generated minutes are provided to the user from the server. The user can review the minutes using a dedicated interface and make corrections as necessary. The minutes also include highlights of the user's emotions, making it easier to understand the nuances of the meeting.

[0802] Step 8:

[0803] Once the user has completed the revisions, the server automatically sends the final minutes to the relevant parties' email addresses, including sentiment data and key points from the meeting, so that all meeting participants can quickly share them.

[0804] Specific examples

[0805] For example, suppose someone says, "Today we will discuss the progress of the project" during a project status meeting. The device captures this utterance and sends it to the server in real time. The server converts the speech into text data, "Today we will discuss the progress of the project." If the emotion engine identifies the emotion "excited," the server includes this emotion data. Next, the generative AI model extracts the key point, "Checking the progress of the project," and generates the summary text, "Today we will check the progress of the project. - (Excited)."

[0806] Based on this summary, minutes are created and provided to the user, who can then make any necessary corrections, and the final minutes are sent to all parties involved, allowing for quick sharing of important meeting content and related emotions.

[0807] Example 2

[0808] 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."

[0809] Conventional meeting minutes creation systems require a great deal of time and effort to capture audio data, convert it to text data, extract key points, and generate minutes, significantly reducing meeting efficiency. Furthermore, minutes created do not take into account the emotions of meeting participants, making it difficult to gain a deep understanding of the progress of a meeting.

[0810] The specification process by the specification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for transmitting voice data to the server in real time, means for inputting the received voice data into a voice recognition engine and converting it into text data, means for inputting the converted text data into an emotion engine and recognizing the user's emotion to generate emotion data, and means for inputting the generated emotion data and text data into a generative AI model and extracting and summarizing important points. This makes it possible to efficiently generate minutes of a meeting from the voice data in real time and provide accurate minutes that reflect the emotions of the participants.

[0811] "Audio data" refers to data that records audio in digital format.

[0812] A "server" refers to a computer system that sends, receives, and processes data over a network.

[0813] "Terminal" refers to a device such as a computer or smartphone that can be directly operated by a user.

[0814] A "voice recognition engine" refers to software or a system that analyzes voice data and converts it into text data.

[0815] "Text data" refers to data that records character information in digital format.

[0816] An "emotion engine" refers to software or a system that analyzes text data or voice tone to generate emotional data.

[0817] "Emotion data" refers to data that indicates the user's emotional state as analyzed by the emotion engine.

[0818] A "generative AI model" refers to an artificial intelligence model that performs natural language processing based on large amounts of data and generates new text.

[0819] "Extracting important points" refers to finding particularly important content or key points from text data or emotion data.

[0820] "Summarization" refers to shortening long text data to include only the main information.

[0821] Minutes are a document that records what was discussed at a meeting and can be referenced at a later date.

[0822] "Real-time" refers to a state in which data processing and communication occurs almost instantaneously.

[0823] "Providing to users" means making the data and information generated by the system accessible and usable by users.

[0824] "Modify" refers to changing the content of data or documents that have already been generated and stored.

[0825] "Send" refers to sending the generated or modified data or documents to other users or systems involved.

[0826] This invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions. This system functions efficiently, with the terminal, server, and user each playing their own role.

[0827] As soon as the conference starts, the device uses a microphone to capture audio data. The captured audio data is sent to the server in real time. Specifically, audio capture software installed on the device (e.g., a Windows audio recording application) records the audio in real time and sends it in a stream format to the server using HTTP or WebSocket.

[0828] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. In this conversion process, the server calls the API of the voice recognition engine, sends the voice data, and captures the returned text data.

[0829] The server then inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of the voice and the choice of words used, and emotion data is generated.

[0830] The server then inputs the generated emotion data and converted text data into a generative AI model (e.g., OpenAI GPT-3) to extract key points. The server then creates a prompt and sends it to the generative AI model. For example, the prompt might include something like, "Please summarize the following conversation. Please also consider sentiment analysis." The server then receives the response from the generative AI model, extracts key points, and summarizes them.

[0831] Based on the summarized text data, the server generates minutes in the format specified by the user. The generated minutes are output in a standard or custom format. The user can use a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. This correction process is intuitive and easy to perform.

[0832] The minutes, once they have been finalized, are automatically sent from the server to the email addresses of the relevant parties. Specifically, the server uses an email sending program (e.g., an SMTP server) to send the minutes to the specified recipients as an attachment.

[0833] Specific examples

[0834] For example, suppose a project team is holding a progress meeting. The device records the meeting audio in real time using its built-in microphone and sends it to the server via HTTPS. The server then uses the Google Cloud Speech-to-Text API to convert the audio into text data, such as "Today we will discuss the progress of the project." It then uses an emotion engine (IBM Watson Tone Analyzer) to recognize emotions such as "excitement" or "anxiety" from the text data. A prompt is then input into a generative AI model (OpenAI GPT-3), which extracts key points and summarizes them, taking into account the emotion data. For example, the summary might be something like "Today we will review the progress of the project. - (excited)." Based on the generated summary text, the server generates meeting minutes in standard PDF format or a custom format. The user can review the minutes in a web browser and make any necessary corrections. The server then automatically sends them to all parties involved via an SMTP server.

[0835] Prompt Sentence Examples

[0836] "Generate minutes by summarizing what was said in project team status meetings and adding sentiment analysis."

[0837] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0838] Step 1: Acquire audio data

[0839] The device starts capturing audio data using a microphone as soon as the conference starts. Audio can be recorded in real time using a built-in microphone or a connected microphone, or using audio capture software (e.g., Windows audio recording application). The input here is ambient audio, and the output is captured digital audio data.

[0840] Step 2: Sending audio data

[0841] The device sends the acquired audio data to the server in real time. The audio data is sent in stream format using HTTP or WebSocket. The input here is the audio data acquired in step 1, and the output is the audio data sent to the server.

[0842] Step 3: Convert audio data to text

[0843] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. Specifically, the server calls the voice recognition engine's API, sends the voice data, and captures the returned text data. The input here is the voice data sent in step 2, and the output is the converted text data.

[0844] Step 4: Generate emotion data

[0845] The server inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotion. The server then sends the text data to the emotion engine's API and obtains the returned emotion data. Specifically, emotion labels such as "positive" and "negative" are generated. The input here is the text data converted in step 3, and the output is the generated emotion data.

[0846] Step 5: Extract and summarize key points

[0847] The server inputs text data and emotion data using a generative AI model (e.g., OpenAI GPT-3), extracts key points, and summarizes them. The server creates a prompt and sends it to the generative AI model. The prompt uses something like, "Please summarize the following conversation. Please also consider emotion analysis." The server receives a response from the generative AI model, extracts key points, and summarizes them. The input here is emotion data and text data, and the output is summarized text data.

[0848] Step 6: Generate the transcript

[0849] The server generates minutes from the summarized text data in the format specified by the user. The server creates documents using a template engine and outputs them in Word or PDF format. The input here is the summarized text data from step 5, and the output is the generated minutes.

[0850] Step 7: Providing and correcting minutes

[0851] The generated minutes are first provided to the user from the server. The user uses a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. The user highlights the parts to be corrected and edits intuitively. The input here is the minutes generated in step 6, and the output is the corrected minutes.

[0852] Step 8: Sending the minutes

[0853] The final minutes, once they have been corrected, are automatically sent from the server to the email addresses of the relevant parties. The server uses an email sending program (e.g., an SMTP server) to send the minutes as an attachment to the specified recipient. The input here is the minutes corrected in step 7, and the output is the minutes sent to the relevant parties.

[0854] (Application example 2)

[0855] 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."

[0856] Improving communication efficiency and accurate information sharing among staff and managers are key challenges at logistics centers. In particular, there is a lack of means to manage work instructions and contact information in real time, and to extract and quickly share important information. It is also necessary to understand the emotional state of staff while they are working and respond appropriately. This is essential to improve work efficiency and reduce errors.

[0857] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0858] In this invention, the server includes means for acquiring voice data in real time, means for converting the acquired voice data into text data, and means for extracting important points from the converted text data and emotion data and summarizing them in real time, thereby making it possible to improve the efficiency of communication in a logistics center, quickly share important information, and grasp the emotional state of staff.

[0859] "Audio data" refers to information that records audio in digital format.

[0860] "Text data" is information obtained by converting voice data into a string of characters.

[0861] "Emotion data" is information obtained by analyzing the user's emotional state from voice data and text data.

[0862] A "summary" is information that extracts important points from the converted text data and summarizes them concisely.

[0863] A "work report" is a document that records the progress and status of work, generated based on summarized text data.

[0864] "User" means a person or organization that uses the system to acquire voice data or check or modify text data.

[0865] A "stakeholder" is a person or organization that needs to receive the generated work report.

[0866] "Real-time" refers to processing or reaction occurring immediately, without delay.

[0867] This is a system that realizes efficient management and information sharing of work instructions and contact information in a logistics center. This system performs a series of processes: acquires voice data, converts it into text data, extracts important points including emotional data, generates a work report in real time, and sends it to the relevant parties after user corrections.

[0868] Hardware and Software Configuration

[0869] 1. Audio data acquisition method

[0870] The device uses a microphone to capture voice data, which is then transmitted to a server in real time. The device can be a wearable device such as smart glasses or a head-mounted display.

[0871] 2. Means of converting audio data into text data

[0872] After receiving the voice data, the server converts it into text data using a voice recognition engine, such as the speech_recognition library.

[0873] 3. Means of generating emotion data

[0874] The server inputs the transformed text data into a sentiment analysis model to generate sentiment data, using the sentiment analysis pipeline in the transformers library.

[0875] 4. Extraction and summarization of key points

[0876] The server inputs emotion data and text data into a generative AI model to extract key points. The generative AI model uses GPT-3 or a similar model, and the generated prompt is set as follows:

[0877] "Summarize the text below, highlighting the key points"

[0878] for example:

[0879] This is a message from our shipping department. There is a slight delay in our current schedule. It is taking some time to sort out the packages, but we expect the delay to be resolved by tomorrow morning.

[0880] 5. Means of generating work reports

[0881] Based on the summarized text data, the server automatically generates work reports in real time, which can be output in standard or custom formats.

[0882] 6. Means of correction and provision

[0883] The generated work report is provided to the user via the server, and the user can make corrections via a web interface. The corrections are updated in real time.

[0884] 7. Means of delivery to interested parties

[0885] Once the final corrections are completed, the work report is automatically sent from the server to the relevant parties.

[0886] Specific examples

[0887] For example, when a new package arrives at a logistics center, staff can use a head-mounted display to input instructions by voice. This converts the voice into text in real time, and important points are extracted along with emotional data. This content is then generated in real time as a work report, which managers can immediately check and correct, and finally send to the relevant parties. This improves the efficiency of information transmission and the transparency of work.

[0888] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0889] Step 1:

[0890] The terminal uses a microphone to capture voice data. The voice data includes instructions and messages given during work at the logistics center. The terminal transmits the captured voice data to the server in real time. The input is the real-time voice data, and the output is the voice data transmitted to the server.

[0891] Step 2:

[0892] The server inputs the received voice data into a voice recognition engine. The voice recognition engine (e.g., speech_recognition library) is used to convert the voice data into text data. The input is the voice data sent to the server, and the output is text data.

[0893] Step 3:

[0894] The server inputs the transformed text data into a sentiment analysis model. The sentiment analysis model (e.g., the transformers library) is used to generate sentiment data from the text data. The input is text data, and the output is text data containing sentiment data.

[0895] Step 4:

[0896] The server inputs emotion data and text data into a generative AI model. Using a generative AI model (e.g., GPT-3), it extracts key points and generates a summary. The input is text data including emotion data, and the output is summarized text data. The generation prompt is as follows: "Please summarize the following text, highlighting the key points."

[0897] Step 5:

[0898] The server generates a work report based on the summarized text data. The generated work report is automatically formatted in a format specified by the user. The input is the summarized text data, and the output is a work report in a standard or custom format.

[0899] Step 6:

[0900] The server provides the generated work report to the user, who can then review the work report via a web interface and make real-time edits as needed. The input is the generated work report, and the output is the work report edited by the user.

[0901] Step 7:

[0902] The server sends the revised work report to the relevant parties. The revised report is automatically sent to the specified email address. The input is the work report revised by the user, and the output is the final work report sent to the relevant parties.

[0903] 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.

[0904] 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.

[0905] 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.

[0906] [Fourth embodiment]

[0907] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0908] 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.

[0909] 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).

[0910] 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.

[0911] 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.

[0912] 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).

[0913] 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.

[0914] 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.

[0915] 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.

[0916] 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.

[0917] 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.

[0918] 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.

[0919] 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."

[0920] The present invention is a series of systems that captures conference audio data in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0921] Acquiring audio data

[0922] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[0923] Converting audio data into text data

[0924] The server converts the received voice data into text data in real time using a voice recognition engine, which has the ability to recognize voice and convert it into text with high accuracy.

[0925] Extraction and summary of key points

[0926] The converted text data is processed on a server, where a generative AI model is used to extract key points, which are then summarized into concise text using an algorithm.

[0927] Generate meeting minutes

[0928] Based on the summarized text data, the server generates a transcript according to the format specified by the user, which can be output in a standard or custom format.

[0929] Provision and correction of minutes

[0930] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. These corrections can be easily made using a dedicated interface.

[0931] Sending minutes

[0932] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting.

[0933] Specific examples

[0934] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The generative AI model then summarizes the text as "Today we will review the progress of the project." Based on the summary text, minutes are generated in a standard format. The user can review and edit the minutes, and they are eventually automatically sent to all parties involved.

[0935] In this way, the present invention significantly improves the efficiency of taking minutes at meetings, saving time and effort.

[0936] The processing flow will be explained below.

[0937] Step 1:

[0938] The terminal acquires voice data using a microphone at the same time as the conference starts, and transmits the voice data to the server in real time.

[0939] Step 2:

[0940] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[0941] Step 3:

[0942] The server inputs the converted text data into a generative AI model, which extracts key points from the text data. The generative AI model uses natural language processing techniques to identify important statements and content.

[0943] Step 4:

[0944] The server then uses a summarization algorithm to create a summary of the key information based on the extracted key points. This summarization algorithm is designed to achieve both accuracy and conciseness.

[0945] Step 5:

[0946] The server generates minutes from the summarized text data in a format specified by the user, and outputs the minutes in a standard or custom format.

[0947] Step 6:

[0948] The server provides the generated minutes to the user, and allows the user to check the contents of the minutes using a dedicated interface and correct them as necessary.

[0949] Step 7:

[0950] After the user completes the edits, the server automatically sends the final minutes to the relevant email addresses, allowing all meeting participants to quickly share accurate minutes.

[0951] Example 1

[0952] 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."

[0953] While it is important to create meeting minutes accurately and quickly, traditional methods rely on manual processes, which are time-consuming and labor-intensive. Furthermore, it is not easy to check and correct minutes in real time and share them with relevant parties immediately, which can lead to delays in information transmission. For these reasons, there was a need to improve the efficiency and accuracy of the minutes creation process.

[0954] 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.

[0955] In this invention, the server includes a means for transmitting voice data to the server in real time, a means for converting voice data into text data using a voice recognition engine, and a means for extracting key points using a generative AI model. This enables the voice data of a meeting to be converted into text in real time, key points to be efficiently extracted and summarized, and minutes to be generated quickly. Furthermore, the generated minutes can be reviewed and revised in real time by the user and immediately sent to relevant parties, preventing delays in information transmission and significantly improving post-meeting work efficiency.

[0956] "Audio data" refers to data in which audio information from meetings, interviews, etc. is recorded in digital format.

[0957] "Text data" refers to data of character information generated from voice data using voice recognition technology or the like.

[0958] A "server" is a computer system that provides resources and data over a network.

[0959] A "terminal" is a device operated by a user, and includes a computer, laptop, tablet, smartphone, etc.

[0960] A "user" is someone who uses the system, and includes conference participants and administrators.

[0961] A "voice recognition engine" is software or hardware that analyzes voice data and converts it into corresponding text data.

[0962] A "generative AI model" is an artificial intelligence model that uses artificial intelligence technology to analyze data and perform tasks such as extracting key points and summarizing.

[0963] "Minutes" are text documents that record the contents of a meeting or the like.

[0964] "Real-time" refers to data processing and information transmission occurring almost instantaneously, with minimal delay.

[0965] A "prompt sentence" is an input sentence given to a generative AI model that contains instructions for the model to perform a specific task.

[0966] A "summary" is a sentence that succinctly summarizes the contents of the text data.

[0967] An "email sending API" is an interface for sending email from a program.

[0968] The present invention is a system that captures audio data of a meeting in real time, converts the audio into text data, summarizes important points, and generates, edits, and sends minutes of the meeting.

[0969] When a conference begins, the device captures audio data using its built-in microphone or an external microphone. The captured audio data is temporarily stored in a buffer and then sent to the server in real time. The audio data is sent using the TCP / IP protocol.

[0970] The server instantly converts the received voice data into text data using a "voice recognition engine." This voice recognition engine can be a general-purpose tool such as "Google Cloud Speech-to-Text." The server sends the voice data to the engine via an API and saves the returned text data.

[0971] The converted text data is then used on the server to extract important points using the generative AI model "GPT-4." The generative AI model then uses prompts to perform specific tasks. Specifically, a prompt containing the instruction "Please summarize the contents of today's meeting" is used. An example of a prompt is "Convert the audio data from the meeting into text data, summarize the key points, and create minutes. The contents of the meeting discussion are listed below."

[0972] The extracted key points are summarized into concise text using a summarization algorithm, and the server generates minutes based on the summary text. The format of the minutes can be specified by the user, and they can be output in formats such as Markdown, PDF, Word, etc. For example, the server can use a template engine to convert the summary text into Markdown format and output it as PDF using a library such as "PDFKit."

[0973] The generated minutes are provided to the user by the server through a dedicated web interface, where the user can review the contents of the minutes and make any necessary corrections. This web interface has a function for editing the minutes text in real time. After the user has completed the corrections, the minutes are automatically sent to all parties involved from the server using an email sending API (e.g., "SendGrid" or "Amazon SES").

[0974] As a concrete example, when a project team holds a progress meeting, a laptop or tablet picks up the audio of the meeting and sends it to a server in real time. Google Cloud Speech-to-Text converts the audio into text, such as "Today we will discuss the progress of the project," and the generative AI model GPT-4 summarizes this as "Today we will review the progress of the project." The summarized text is converted to Markdown format and output as a PDF. The user can review and edit these minutes via a web interface, and they are finally automatically sent to all parties involved.

[0975] Thus, the present invention efficiently and automatically processes minutes of meetings, thereby saving a great deal of time and effort.

[0976] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0977] Step 1:

[0978] When a conference starts, the device captures audio data using a built-in microphone or an external microphone. The audio data is temporarily stored in a buffer in real time and then sent to the server using the TCP / IP protocol.

[0979] Specific example of operation: A laptop installed in a conference room picks up speech using its built-in microphone and transmits it to a server via Wi-Fi.

[0980] Step 2:

[0981] The server converts the voice data it receives in real time into text data using a "voice recognition engine." The voice data (input) is analyzed by the voice recognition software and the corresponding text data (output) is generated.

[0982] Specific example of operation: The server sends voice data to the "Google Cloud Speech-to-Text" service, receives the results as text data, and stores them in a database.

[0983] Step 3:

[0984] The server extracts and summarizes the key points from the converted text data using a "generative AI model." The text data (input) is sent to the AI ​​model with a prompt, and a text (output) summarizing the key points is generated.

[0985] Specific example of operation: GPT-4 is sent a prompt saying, "Convert the audio data of the meeting into text data and summarize the key points," and the AI ​​model outputs the summary text, "Today we will review the progress of the project."

[0986] Step 4:

[0987] The server generates minutes based on the summarized text data. It converts the summarized text (input) into a specified format (output) using a template engine.

[0988] Specific example of operation: Convert the summary text into Markdown format and output it as PDF using the "PDFKit" library.

[0989] Step 5:

[0990] The server provides the generated minutes to the user through a dedicated web interface. The PDF file (input) is displayed via the web interface, and the user checks its contents.

[0991] Specific example of operation: A user opens a web browser, accesses a dedicated URL, and checks the displayed minutes.

[0992] Step 6:

[0993] Users can modify the contents of the minutes through a web interface by directly editing the displayed text (input) and saving it as modified text (output).

[0994] Example of how it works: A user edits typos or additional information on the web interface and clicks the "Save" button, which sends the corrections to the server.

[0995] Step 7:

[0996] The server automatically sends the revised minutes to the relevant parties. The revised PDF file (input) is sent to the relevant parties' email addresses using the email sending API.

[0997] Example of specific operation: The server attaches the corrected minutes and uses the "SendGrid" API to send a message to the relevant email address with the message "The meeting minutes are complete."

[0998] In this way, the system efficiently executes a series of processes from acquiring audio data of a meeting to generating, editing, and distributing minutes.

[0999] (Application example 1)

[1000] 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."

[1001] Conventional conferencing systems require manual processes, from collecting audio data to converting it into text, extracting key points, and creating and sending meeting minutes, resulting in inefficiency and requiring a great deal of time and effort. Furthermore, in-factory meetings also faced difficulties in generating and sharing work instructions and reports in real time, reducing work efficiency. Furthermore, there was a risk that important points would be overlooked.

[1002] 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.

[1003] In this invention, the server includes a means for acquiring voice data, a means for converting the acquired voice data into text data, and a means for extracting and summarizing important points from the converted text data. This automatically converts voice data into text, making it possible to summarize important points and generate minutes, work instructions, and reports. In addition, the generated documents can be modified and confirmed by the user, enabling highly accurate information sharing.

[1004] "Audio data" refers to audio recorded and stored in digital format from meetings, work sites, etc.

[1005] "Text data" is digital data that has been analyzed and converted into text information.

[1006] "Key points" are pieces of information that indicate important points or key points within the collected data.

[1007] A "generative AI model" is an artificial intelligence algorithm that generates or summarizes new information based on machine learning and data processing.

[1008] "Summarizing" means to concisely summarize a large amount of information and extract and organize only the main points and main points.

[1009] Minutes are documents that record the contents of meetings and discussions.

[1010] "Correction" refers to the process of changing the content of generated documents or data to improve their accuracy and clarity.

[1011] "Stakeholders" are people who need to share information or work instructions in a particular meeting or task.

[1012] A "work instruction" is a document that provides specific instructions on how to perform a particular task or work.

[1013] A "report" is a formal document used to record and report a particular situation or work process.

[1014] The system of this invention is designed to capture voice data from meetings and factories in real time, automatically convert it into text, and then extract and summarize important points to generate minutes, work instructions, and reports.The main components of this system include a terminal that captures the voice data, a server that processes the data, and a user interface for making corrections.

[1015] Hardware / Software Configuration

[1016] Device:

[1017] This device is equipped with a microphone to capture voice data during meetings or in factories. The voice data is captured in real time and sent to a server.

[1018] server:

[1019] Google Cloud Speech-to-Text is used to convert voice data into text data. This text data is processed internally on the server, and key points are extracted and summarized. OpenAI's GPT-3 is used as a generative AI model to generate summaries and instructions.

[1020] Data processing / data calculation:

[1021] 1. Acquisition and transmission of audio data:

[1022] It uses the device's microphone to collect voice data and transmits it to a server in real time.

[1023] 2. Audio to text conversion:

[1024] Using the Google Cloud Speech-to-Text API, voice data is converted into text with high accuracy.

[1025] 3. Extract and summarize key points:

[1026] The text data is analyzed internally on the server, and key points are extracted and summarized using OpenAI GPT-3's generative AI model.

[1027] 4. Generate minutes, work instructions, and reports:

[1028] Automatically generate meeting minutes, work instructions, and reports based on summarized text data, and output these documents in standard or customizable formats.

[1029] 5. User Modifications:

[1030] The generated document is presented to the user through a dedicated interface that allows for modification, and supports real-time modification.

[1031] Specific examples

[1032] For example, in a factory progress meeting, a device captures the audio data of the meeting in real time and sends it to a server. The server converts the audio data into text data such as "Today we will discuss the importance of maintaining machine A." Next, a generative AI model is used to summarize the text as "Maintenance of machine A is important." A work instruction is generated based on the summarized text. The user can then review and revise the work instruction, and it is finally automatically sent to all parties involved.

[1033] Example prompts to input to a generative AI model:

[1034] Summarize the following text:

[1035] Today we will discuss the importance of maintaining Machine A.

[1036] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1037] Step 1:

[1038] The terminal captures voice data in real time during meetings and factories. The terminal's built-in microphone collects the voice data in digital format. The collected voice data is then sent to a server in real time.

[1039] Input: Audio from meetings and factories

[1040] Output: Digital audio data (sent to server)

[1041] Step 2:

[1042] The server converts the acquired voice data into text data using the Google Cloud Speech-to-Text API. The server sends the voice data to the API and receives the returned text data.

[1043] Input: Digital audio data

[1044] Output: Text data (converted from audio)

[1045] Step 3:

[1046] The server analyzes the text data and extracts and summarizes key points using a generative AI model (OpenAI GPT-3). First, the text data is input to the generative AI model as a prompt sentence, and the summarized text is obtained.

[1047] Input: Text data (records of meetings and work)

[1048] Output: Summarized text data (key points)

[1049] Step 4:

[1050] The server generates meeting minutes, work instructions, and reports in a user-specified format based on the summarized text data, and the generated documents can be output in standard or customized formats.

[1051] Input: Summarized text data

[1052] Output: minutes, work instructions, reports (user-specified format)

[1053] Step 5:

[1054] The generated document is provided to the user from the server, who can review the document and make corrections as needed through a dedicated interface that supports real-time corrections.

[1055] Input: Minutes, work orders, reports

[1056] Output: Corrected document (user confirmation and correction)

[1057] Step 6:

[1058] The server automatically sends the corrected document to the relevant person's email address.

[1059] Input: revised document

[1060] Output: Send to relevant parties (via email)

[1061] Specifically, the system begins by capturing the audio of the meeting in real time and sending it to the server, after which a series of processes are automatically carried out, including converting the audio data into text, extracting important points, generating various documents, allowing the user to check and correct them, and finally sending them.

[1062] 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.

[1063] The present invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions.

[1064] Acquiring audio data

[1065] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[1066] Converting audio data into text data

[1067] The server inputs the received voice data into a voice recognition engine, which converts the voice data into text data. This voice recognition engine performs highly accurate voice recognition in real time.

[1068] Generating emotional data using an emotion engine

[1069] The server inputs the converted text data into an emotion engine to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of voice and the choice of words used, and emotion data is generated.

[1070] Extraction and summary of key points

[1071] The converted text data, along with the emotion data, is then fed into a generative AI model, which the server uses to extract key points and, if necessary, highlight parts that emphasize the user's emotions.The server then uses a summarization algorithm to compile the extracted key points into concise text.

[1072] Generate meeting minutes

[1073] Based on the summarized text data, the server generates minutes according to the format specified by the user, and the generated minutes can be output in a standard or custom format.

[1074] Provision and correction of minutes

[1075] The generated minutes are first provided to the user from the server. The user can then check the contents of the minutes and make any necessary corrections. Corrections can be easily made using a dedicated interface.

[1076] Sending minutes

[1077] The final minutes, once edited, are automatically sent from the server to the email addresses of the relevant parties, allowing users to share accurate and edited minutes with the relevant parties immediately after the meeting ends.

[1078] Specific examples

[1079] For example, suppose a project team is holding a progress meeting. The device captures the audio of the meeting and sends it to the server in real time. The server converts the audio into text data such as "Today we will discuss the progress of the project." The emotion engine then analyzes emotions such as "excitement" or "anxiety" from the text data and summarizes it as "Today we will review the progress of the project. - (excited)." The parts that emphasize the user's emotions are highlighted and recorded in the minutes.

[1080] The server then generates minutes in a standard format based on this summary text. The user can then make any necessary corrections, and the minutes are finally sent automatically to all parties involved. In this way, the present invention reduces the burden on meetings and enables efficient and accurate minutes creation. Furthermore, the introduction of an emotion engine makes it possible to generate minutes that reflect the emotional state of participants, thereby providing a deeper understanding of the progress of the meeting.

[1081] The processing flow will be explained below.

[1082] Step 1:

[1083] When the conference starts, the terminal uses a microphone to capture audio data, which is then transmitted to the server in real time.

[1084] Step 2:

[1085] The server inputs the received voice data into a speech recognition engine, which performs highly accurate speech recognition and converts the voice data into text. This speech recognition engine recognizes the voices of multiple speakers in the conference and transcribes each individual statement into text.

[1086] Step 3:

[1087] The server inputs the converted text data into an emotion engine, which analyzes the word choice, context, and speaker's tone in the text data to recognize the user's emotions in real time and generate emotion data. This emotion data includes emotion labels such as "happiness," "anxiety," and "excitement."

[1088] Step 4:

[1089] The server combines the generated emotion data with the text data and inputs it into a generative AI model, which extracts key points from the text and identifies parts that emphasize the user's emotions.

[1090] Step 5:

[1091] The server extracts key points and uses a summarization algorithm to create a concise summary text, highlighting parts that emphasize specific emotions based on the emotion data.

[1092] Step 6:

[1093] The server generates minutes from the summarized text data in a format specified by the user, and the generated minutes can be output in a standard or custom format.

[1094] Step 7:

[1095] The generated minutes are provided to the user from the server. The user can review the minutes using a dedicated interface and make corrections as necessary. The minutes also include highlights of the user's emotions, making it easier to understand the nuances of the meeting.

[1096] Step 8:

[1097] Once the user has completed the revisions, the server automatically sends the final minutes to the relevant parties' email addresses, including sentiment data and key points from the meeting, so that all meeting participants can quickly share them.

[1098] Specific examples

[1099] For example, suppose someone says, "Today we will discuss the progress of the project" during a project status meeting. The device captures this utterance and sends it to the server in real time. The server converts the speech into text data, "Today we will discuss the progress of the project." If the emotion engine identifies the emotion "excited," the server includes this emotion data. Next, the generative AI model extracts the key point, "Checking the progress of the project," and generates the summary text, "Today we will check the progress of the project. - (Excited)."

[1100] Based on this summary, minutes are created and provided to the user, who can then make any necessary corrections, and the final minutes are sent to all parties involved, allowing for quick sharing of important meeting content and related emotions.

[1101] Example 2

[1102] 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."

[1103] Conventional meeting minutes creation systems require a great deal of time and effort to capture audio data, convert it to text data, extract key points, and generate minutes, significantly reducing meeting efficiency. Furthermore, minutes created do not take into account the emotions of meeting participants, making it difficult to gain a deep understanding of the progress of a meeting.

[1104] The specification process by the specification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for transmitting voice data to the server in real time, means for inputting the received voice data into a voice recognition engine and converting it into text data, means for inputting the converted text data into an emotion engine and recognizing the user's emotion to generate emotion data, and means for inputting the generated emotion data and text data into a generative AI model and extracting and summarizing important points. This makes it possible to efficiently generate minutes of a meeting from the voice data in real time and provide accurate minutes that reflect the emotions of the participants.

[1105] "Audio data" refers to data that records audio in digital format.

[1106] A "server" refers to a computer system that sends, receives, and processes data over a network.

[1107] "Terminal" refers to a device such as a computer or smartphone that can be directly operated by a user.

[1108] A "voice recognition engine" refers to software or a system that analyzes voice data and converts it into text data.

[1109] "Text data" refers to data that records character information in digital format.

[1110] An "emotion engine" refers to software or a system that analyzes text data or voice tone to generate emotional data.

[1111] "Emotion data" refers to data that indicates the user's emotional state as analyzed by the emotion engine.

[1112] A "generative AI model" refers to an artificial intelligence model that performs natural language processing based on large amounts of data and generates new text.

[1113] "Extracting important points" refers to finding particularly important content or key points from text data or emotion data.

[1114] "Summarization" refers to shortening long text data to include only the main information.

[1115] Minutes are a document that records what was discussed at a meeting and can be referenced at a later date.

[1116] "Real-time" refers to a state in which data processing and communication occurs almost instantaneously.

[1117] "Providing to users" means making the data and information generated by the system accessible and usable by users.

[1118] "Modify" refers to changing the content of data or documents that have already been generated and stored.

[1119] "Send" refers to sending the generated or modified data or documents to other users or systems involved.

[1120] This invention combines a series of systems that captures meeting audio data in real time, converts it into text data, extracts important points, and generates, edits, and sends minutes of the meeting, with an emotion engine that recognizes the user's emotions. This system functions efficiently, with the terminal, server, and user each playing their own role.

[1121] As soon as the conference starts, the device uses a microphone to capture audio data. The captured audio data is sent to the server in real time. Specifically, audio capture software installed on the device (e.g., a Windows audio recording application) records the audio in real time and sends it in a stream format to the server using HTTP or WebSocket.

[1122] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. In this conversion process, the server calls the API of the voice recognition engine, sends the voice data, and captures the returned text data.

[1123] The server then inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotions. This emotion recognition is performed by analyzing the tone of the voice and the choice of words used, and emotion data is generated.

[1124] The server then inputs the generated emotion data and converted text data into a generative AI model (e.g., OpenAI GPT-3) to extract key points. The server then creates a prompt and sends it to the generative AI model. For example, the prompt might include something like, "Please summarize the following conversation. Please also consider sentiment analysis." The server then receives the response from the generative AI model, extracts key points, and summarizes them.

[1125] Based on the summarized text data, the server generates minutes in the format specified by the user. The generated minutes are output in a standard or custom format. The user can use a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. This correction process is intuitive and easy to perform.

[1126] The minutes, once they have been finalized, are automatically sent from the server to the email addresses of the relevant parties. Specifically, the server uses an email sending program (e.g., an SMTP server) to send the minutes to the specified recipients as an attachment.

[1127] Specific examples

[1128] For example, suppose a project team is holding a progress meeting. The device records the meeting audio in real time using its built-in microphone and sends it to the server via HTTPS. The server then uses the Google Cloud Speech-to-Text API to convert the audio into text data, such as "Today we will discuss the progress of the project." It then uses an emotion engine (IBM Watson Tone Analyzer) to recognize emotions such as "excitement" or "anxiety" from the text data. A prompt is then input into a generative AI model (OpenAI GPT-3), which extracts key points and summarizes them, taking into account the emotion data. For example, the summary might be something like "Today we will review the progress of the project. - (excited)." Based on the generated summary text, the server generates meeting minutes in standard PDF format or a custom format. The user can review the minutes in a web browser and make any necessary corrections. The server then automatically sends them to all parties involved via an SMTP server.

[1129] Prompt Sentence Examples

[1130] "Generate minutes by summarizing what was said in project team status meetings and adding sentiment analysis."

[1131] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1132] Step 1: Acquire audio data

[1133] The device starts capturing audio data using a microphone as soon as the conference starts. Audio can be recorded in real time using a built-in microphone or a connected microphone, or using audio capture software (e.g., Windows audio recording application). The input here is ambient audio, and the output is captured digital audio data.

[1134] Step 2: Sending audio data

[1135] The device sends the acquired audio data to the server in real time. The audio data is sent in stream format using HTTP or WebSocket. The input here is the audio data acquired in step 1, and the output is the audio data sent to the server.

[1136] Step 3: Convert audio data to text

[1137] The server inputs the received voice data into a voice recognition engine (e.g., Google Cloud Speech-to-Text API) and converts the voice data into text data. Specifically, the server calls the voice recognition engine's API, sends the voice data, and captures the returned text data. The input here is the voice data sent in step 2, and the output is the converted text data.

[1138] Step 4: Generate emotion data

[1139] The server inputs the converted text data into an emotion engine (e.g., IBM Watson Tone Analyzer) to recognize the user's emotion. The server then sends the text data to the emotion engine's API and obtains the returned emotion data. Specifically, emotion labels such as "positive" and "negative" are generated. The input here is the text data converted in step 3, and the output is the generated emotion data.

[1140] Step 5: Extract and summarize key points

[1141] The server inputs text data and emotion data using a generative AI model (e.g., OpenAI GPT-3), extracts key points, and summarizes them. The server creates a prompt and sends it to the generative AI model. The prompt uses something like, "Please summarize the following conversation. Please also consider emotion analysis." The server receives a response from the generative AI model, extracts key points, and summarizes them. The input here is emotion data and text data, and the output is summarized text data.

[1142] Step 6: Generate the transcript

[1143] The server generates minutes from the summarized text data in the format specified by the user. The server creates documents using a template engine and outputs them in Word or PDF format. The input here is the summarized text data from step 5, and the output is the generated minutes.

[1144] Step 7: Providing and correcting minutes

[1145] The generated minutes are first provided to the user from the server. The user uses a dedicated interface (e.g., an editing screen on a browser) to check the contents of the minutes and make corrections as necessary. The user highlights the parts to be corrected and edits intuitively. The input here is the minutes generated in step 6, and the output is the corrected minutes.

[1146] Step 8: Sending the minutes

[1147] The final minutes, once they have been corrected, are automatically sent from the server to the email addresses of the relevant parties. The server uses an email sending program (e.g., an SMTP server) to send the minutes as an attachment to the specified recipient. The input here is the minutes corrected in step 7, and the output is the minutes sent to the relevant parties.

[1148] (Application example 2)

[1149] 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."

[1150] Improving communication efficiency and accurate information sharing among staff and managers are key challenges at logistics centers. In particular, there is a lack of means to manage work instructions and contact information in real time, and to extract and quickly share important information. It is also necessary to understand the emotional state of staff while they are working and respond appropriately. This is essential to improve work efficiency and reduce errors.

[1151] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1152] In this invention, the server includes means for acquiring voice data in real time, means for converting the acquired voice data into text data, and means for extracting important points from the converted text data and emotion data and summarizing them in real time, thereby making it possible to improve the efficiency of communication in a logistics center, quickly share important information, and grasp the emotional state of staff.

[1153] "Audio data" refers to information that records audio in digital format.

[1154] "Text data" is information obtained by converting voice data into a string of characters.

[1155] "Emotion data" is information obtained by analyzing the user's emotional state from voice data and text data.

[1156] A "summary" is information that extracts important points from the converted text data and summarizes them concisely.

[1157] A "work report" is a document that records the progress and status of work, generated based on summarized text data.

[1158] "User" means a person or organization that uses the system to acquire voice data or check or modify text data.

[1159] A "stakeholder" is a person or organization that needs to receive the generated work report.

[1160] "Real-time" refers to processing or reaction occurring immediately, without delay.

[1161] This is a system that realizes efficient management and information sharing of work instructions and contact information in a logistics center. This system performs a series of processes: acquires voice data, converts it into text data, extracts important points including emotional data, generates a work report in real time, and sends it to the relevant parties after user corrections.

[1162] Hardware and Software Configuration

[1163] 1. Audio data acquisition method

[1164] The device uses a microphone to capture voice data, which is then transmitted to a server in real time. The device can be a wearable device such as smart glasses or a head-mounted display.

[1165] 2. Means of converting audio data into text data

[1166] After receiving the voice data, the server converts it into text data using a voice recognition engine, such as the speech_recognition library.

[1167] 3. Means of generating emotion data

[1168] The server inputs the transformed text data into a sentiment analysis model to generate sentiment data, using the sentiment analysis pipeline in the transformers library.

[1169] 4. Extraction and summarization of key points

[1170] The server inputs emotion data and text data into a generative AI model to extract key points. The generative AI model uses GPT-3 or a similar model, and the generated prompt is set as follows:

[1171] "Summarize the text below, highlighting the key points"

[1172] for example:

[1173] This is a message from our shipping department. There is a slight delay in our current schedule. It is taking some time to sort out the packages, but we expect the delay to be resolved by tomorrow morning.

[1174] 5. Means of generating work reports

[1175] Based on the summarized text data, the server automatically generates work reports in real time, which can be output in standard or custom formats.

[1176] 6. Means of correction and provision

[1177] The generated work report is provided to the user via the server, and the user can make corrections via a web interface. The corrections are updated in real time.

[1178] 7. Means of delivery to interested parties

[1179] Once the final corrections are completed, the work report is automatically sent from the server to the relevant parties.

[1180] Specific examples

[1181] For example, when a new package arrives at a logistics center, staff can use a head-mounted display to input instructions by voice. This converts the voice into text in real time, and important points are extracted along with emotional data. This content is then generated in real time as a work report, which managers can immediately check and correct, and finally send to the relevant parties. This improves the efficiency of information transmission and the transparency of work.

[1182] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1183] Step 1:

[1184] The terminal uses a microphone to capture voice data. The voice data includes instructions and messages given during work at the logistics center. The terminal transmits the captured voice data to the server in real time. The input is the real-time voice data, and the output is the voice data transmitted to the server.

[1185] Step 2:

[1186] The server inputs the received voice data into a voice recognition engine. The voice recognition engine (e.g., speech_recognition library) is used to convert the voice data into text data. The input is the voice data sent to the server, and the output is text data.

[1187] Step 3:

[1188] The server inputs the transformed text data into a sentiment analysis model. The sentiment analysis model (e.g., the transformers library) is used to generate sentiment data from the text data. The input is text data, and the output is text data containing sentiment data.

[1189] Step 4:

[1190] The server inputs emotion data and text data into a generative AI model. Using a generative AI model (e.g., GPT-3), it extracts key points and generates a summary. The input is text data including emotion data, and the output is summarized text data. The generation prompt is as follows: "Please summarize the following text, highlighting the key points."

[1191] Step 5:

[1192] The server generates a work report based on the summarized text data. The generated work report is automatically formatted in a format specified by the user. The input is the summarized text data, and the output is a work report in a standard or custom format.

[1193] Step 6:

[1194] The server provides the generated work report to the user, who can then review the work report via a web interface and make real-time edits as needed. The input is the generated work report, and the output is the work report edited by the user.

[1195] Step 7:

[1196] The server sends the revised work report to the relevant parties. The revised report is automatically sent to the specified email address. The input is the work report revised by the user, and the output is the final work report sent to the relevant parties.

[1197] 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.

[1198] 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.

[1199] 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.

[1200] 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.

[1201] 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.

[1202] 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.

[1203] 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).

[1204] 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.

[1205] 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."

[1206] 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.

[1207] 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).

[1208] 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.

[1209] 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.

[1210] 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.

[1211] 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.

[1212] 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.

[1213] 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.

[1214] 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.

[1215] 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.

[1216] 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.

[1217] 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.

[1218] The following is further disclosed regarding the above embodiment.

[1219] (Claim 1)

[1220] means for acquiring audio data;

[1221] A means for converting the acquired voice data into text data;

[1222] A means for extracting and summarizing important points from the converted text data;

[1223] means for generating the summarized text data as minutes in a format designated by the user;

[1224] A means for providing the generated minutes to a user and allowing the user to modify the minutes;

[1225] a means of sending the amended minutes to the relevant parties;

[1226] A system including:

[1227] (Claim 2)

[1228] a means for acquiring audio data in real time;

[1229] A means for converting the acquired voice data into text data in real time;

[1230] A means for extracting and summarizing key points from the converted text data in real time;

[1231] A means for generating the summarized text data as minutes in a format designated by the user in real time;

[1232] A means for providing the generated minutes to a user in real time and allowing the user to modify the minutes in real time;

[1233] A means to send corrected minutes to relevant parties in real time,

[1234] 10. The system of claim 1, comprising:

[1235] (Claim 3)

[1236] A means for temporarily storing the generated minutes so that the user can check them;

[1237] means for accepting user modifications to the stored minutes;

[1238] means for sending the minutes with the user's modifications to the relevant parties;

[1239] 10. The system of claim 1, comprising:

[1240] "Example 1"

[1241] (Claim 1)

[1242] means for acquiring audio data;

[1243] A means for converting the acquired voice data into text data;

[1244] A means for extracting and summarizing important points from the converted text data;

[1245] means for generating the summarized text data as minutes in a format designated by the user;

[1246] A means for providing the generated minutes to a user and allowing the user to modify the minutes;

[1247] a means of sending the amended minutes to the relevant parties;

[1248] means for transmitting the audio data to a server in real time;

[1249] A means to extract key points using generative AI models;

[1250] means for converting voice data into text data using a voice recognition engine;

[1251] A system including:

[1252] (Claim 2)

[1253] a means for acquiring audio data in real time;

[1254] A means for converting the acquired voice data into text data in real time;

[1255] A means for extracting and summarizing key points from the converted text data in real time;

[1256] A means for generating the summarized text data as minutes in a format designated by the user in real time;

[1257] A means for providing the generated minutes to a user in real time and allowing the user to modify the minutes in real time;

[1258] A means to send corrected minutes to relevant parties in real time,

[1259] A means to extract key points in real time using generative AI models;

[1260] 10. The system of claim 1, comprising:

[1261] (Claim 3)

[1262] A means for temporarily storing the generated minutes so that the user can check them;

[1263] means for accepting user modifications to the stored minutes;

[1264] means for sending the minutes with the user's modifications to the relevant parties;

[1265] A dedicated interface allows users to check and edit minutes,

[1266] 10. The system of claim 1, comprising:

[1267] "Application Example 1"

[1268] (Claim 1)

[1269] means for acquiring audio data;

[1270] A means for converting the acquired voice data into text data;

[1271] A means for extracting and summarizing important points from the converted text data;

[1272] means for generating the summarized text data as minutes in a format designated by the user;

[1273] A means for providing the generated minutes to a user and allowing the user to modify the minutes;

[1274] a means of sending the amended minutes to the relevant parties;

[1275] A means for automatically generating work instructions and reports using the acquired voice data;

[1276] A means to use generative AI models to summarize key points before sending to stakeholders;

[1277] A system including:

[1278] (Claim 2)

[1279] a means for acquiring audio data in real time;

[1280] A means for converting the acquired voice data into text data in real time;

[1281] A means for extracting and summarizing key points from the converted text data in real time;

[1282] A means for generating the summarized text data as minutes in a format designated by the user in real time;

[1283] A means for providing the generated minutes to a user in real time and allowing the user to modify the minutes in real time;

[1284] A means to send corrected minutes to relevant parties in real time,

[1285] A means to automatically generate work instructions and reports using voice data in real time;

[1286] A means to use generative AI models to summarise key points in real time before sending to stakeholders;

[1287] 10. The system of claim 1, comprising:

[1288] (Claim 3)

[1289] A means for temporarily storing the generated minutes so that the user can check them;

[1290] means for accepting user modifications to the stored minutes;

[1291] means for sending the minutes with the user's modifications to the relevant parties;

[1292] A means to temporarily store work instructions and reports generated in real time, allowing users to check and modify them,

[1293] 10. The system of claim 1, comprising:

[1294] "Example 2: Combining Emotion Engines"

[1295] (Claim 1)

[1296] means for acquiring audio data;

[1297] A means for transmitting the acquired voice data to a server in real time;

[1298] A means for inputting the received voice data into a voice recognition engine and converting it into text data;

[1299] a means for inputting the converted text data into an emotion engine, recognizing the user's emotion, and generating emotion data;

[1300] A means of inputting the generated emotion data and text data into a generative AI model to extract and summarize key points;

[1301] means for generating the summarized text data as minutes in a format designated by the user;

[1302] A means for providing the generated minutes to a user and allowing the user to modify the minutes;

[1303] a means of sending the amended minutes to the relevant parties;

[1304] A system including:

[1305] (Claim 2)

[1306] A means for providing the generated minutes to a user in real time and allowing the user to modify the minutes in real time;

[1307] A means to send corrected minutes to relevant parties in real time,

[1308] 10. The system of claim 1, comprising:

[1309] (Claim 3)

[1310] A means for temporarily storing the generated minutes so that the user can check them;

[1311] means for accepting user modifications to the stored minutes;

[1312] means for sending the minutes with the user's modifications to the relevant parties;

[1313] 10. The system of claim 1, comprising:

[1314] "Application example 2 when combining emotion engines"

[1315] (Claim 1)

[1316] means for acquiring audio data;

[1317] A means for converting the acquired voice data into text data;

[1318] means for extracting key points from the converted text data and generating a summary including sentiment data;

[1319] means for generating a work report in a format designated by a user from the text data summarized in real time;

[1320] A means for providing the generated work report to a user and allowing the user to modify the work report;

[1321] a means for transmitting the revised work report to the interested parties;

[1322] ...

[1323] A system including:

[1324] (Claim 2)

[1325] a means for acquiring audio data in real time;

[1326] A means for converting the acquired voice data into text data in real time;

[1327] A means of extracting important points from the converted text data and sentiment data and summarizing them in real time;

[1328] means for generating the summarized text data as a work report in a format designated by the user in real time;

[1329] A means for providing the generated work report to a user in real time and allowing the user to modify the work report in real time;

[1330] A means for sending real-time revised work reports to stakeholders;

[1331] 10. The system of claim 1, comprising:

[1332] (Claim 3)

[1333] A means for saving the generated work report in real time so that the user can check it;

[1334] means for accepting user modifications to the stored work report in real time;

[1335] means for sending the work report containing the user's modifications to the relevant parties;

[1336] 10. The system of claim 1, comprising: [Explanation of symbols]

[1337] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for acquiring audio data; A means for converting the acquired voice data into text data; A means for extracting and summarizing important points from the converted text data; means for generating the summarized text data as minutes in a format designated by the user; A means for providing the generated minutes to a user and allowing the user to modify the minutes; a means of sending the amended minutes to the relevant parties; A system including:

2. a means for acquiring audio data in real time; A means for converting the acquired voice data into text data in real time; A means for extracting and summarizing key points from the converted text data in real time; A means for generating the summarized text data as minutes in a format designated by the user in real time; A means for providing the generated minutes to a user in real time and allowing the user to modify the minutes in real time; A means to send corrected minutes to relevant parties in real time, The system of claim 1 , comprising:

3. A means for temporarily storing the generated minutes so that the user can check them; means for accepting user modifications to the stored minutes; means for sending the minutes with the user's modifications to the relevant parties; The system of claim 1 , comprising:

Images