system

Abstract

We provide the system. [Solution] A means for collecting audio and screen data during a meeting, A generation means for converting the aforementioned audio data into text data, Correction means for standardizing technical terms in the aforementioned text data, A display means for visually displaying the corrected text data on the screen data, A storage and sharing means for saving and sharing the corrected text data, A system that includes this.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] With the increase in remote meetings, the need to accurately automatically generate meeting minutes has been growing. However, with conventional technologies, there are difficulties in the grammatical accuracy of Japanese and the uniformity of technical terms, making it difficult to create practical meeting minutes. Also, unique terms used in various specialized fields may be misconverted during the process of text conversion, leading to information loss and confusion, which has become a problem. There is a need for a method to efficiently solve such problems.

Means for Solving the Problems

[0005] This invention comprises a collection means for simultaneously collecting audio data and screen data, a generation means for converting audio data into text in real time using generation AI, a correction means for standardizing technical terms, and a display means for overlaying the correction results onto the screen. This system efficiently records meetings and accurately corrects and standardizes the technical terms used based on an internal database. It also includes a storage and sharing means for securely saving the corrected, accurate meeting minutes data and making it shareable with relevant parties. This prevents mistranslations and improves the efficiency of meeting minute creation.

[0006] "Collection means" refers to devices and methods for acquiring audio and screen data generated during a meeting.

[0007] "Generation means" refers to functions including generative AI used to convert collected audio data into text data.

[0008] "Correction measures" refer to the process of correcting and standardizing technical terms contained in the generated text data using existing databases.

[0009] "Display means" refers to a function that overlays the corrected text data onto the screen data so that the user can see it.

[0010] "Methods for saving and sharing" refer to methods and functions for securely saving the corrected, final text data and sharing it with relevant parties. [Brief explanation of the drawing]

[0011] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4]This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention]

[0012] Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings.

[0013] First, let's explain the terminology used in the following explanation.

[0014] In the following embodiments, the labeled processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0015] In the following embodiments, the labeled RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0016] In the following embodiments, the labeled storage is one or more non-volatile storage devices that store various programs, various parameters, and the like. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0017] In the following embodiments, the labeled communication I / F (Interface) is an interface including a communication processor, an antenna, and the like. The communication I / F manages communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0018] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0019] [First Embodiment]

[0020] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0021] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0022] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0023] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0024] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0025] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0026] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0027] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0028] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0029] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0030] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0031] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0032] This invention is embodied in a system that automatically and accurately generates meeting minutes. This system is characterized by its ability to simultaneously collect audio and screen data and to correct for technical terms.

[0033] This system consists of several main components. First, during a meeting, the terminal collects audio through a microphone and captures video of slides and documents displayed on the projector or PC screen. This allows various data from the meeting to be recorded in real time.

[0034] Next, the server receives the audio data and uses a generative AI to convert it into text data. At this stage, natural language processing techniques that take into account the characteristics of the Japanese language are used, and general grammatical corrections are made.

[0035] The generated text is further refined by the RAG module on the server. Here, the latest technical terms stored in the database are referenced, and identified words are corrected. For example, the term "AI" is corrected and standardized to "artificial intelligence."

[0036] This corrected text is returned to the device and overlaid on the captured screen data. The display format is adjusted for user readability, and the text is displayed in conjunction with the slide content, allowing for immediate review during the meeting.

[0037] The final meeting minutes data is stored on the server in a secure manner. It is then uploaded to a shared folder or appropriate platform so that those with access rights can access the data. At this stage, users receive notifications and can proactively distribute the data to relevant parties after making individual revisions or additions as needed.

[0038] As a concrete example, consider a scenario where a new proposal for product development is made at a specific meeting. A terminal collects the details of the proposal in real time, and meeting minutes are generated using standardized technical terminology, ensuring consistency and ease of understanding for future reference. This contributes to increased work efficiency and improved communication.

[0039] The following describes the processing flow.

[0040] Step 1:

[0041] The device collects audio data via the microphone at the start of the meeting and activates its screen capture function to record slides and materials displayed during the meeting. This ensures that all audio and visual information is simultaneously acquired as digital data.

[0042] Step 2:

[0043] The server receives audio data transmitted in real time and converts it into text data using a generative AI model. Here, speech recognition technology is used to transcribe the content of the audio into text, and basic Japanese grammatical corrections are applied.

[0044] Step 3:

[0045] The server sends the generated text to the RAG module, where it corrects any mistranslations of technical terms. RAG refers to the latest terminology standards in its database and replaces technical terms in the text with appropriate words as needed. This process involves natural language processing techniques.

[0046] Step 4:

[0047] The device displays the corrected text data overlaid on the screen footage captured during the meeting. The font and layout are adjusted to improve text readability, allowing the user to instantly review the corrected text while viewing the screen.

[0048] Step 5:

[0049] The server securely stores the final meeting minutes data after formatting it. Data storage is performed using encryption technology, and the files are uploaded to a shared folder accessible to all relevant parties.

[0050] Step 6:

[0051] Users receive notifications and review the meeting minutes on their devices. After making any necessary corrections or adding annotations, they can distribute the minutes to other attendees and relevant departments via email or internal communication tools.

[0052] (Example 1)

[0053] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0054] It is difficult to efficiently collect audio and visual information during meetings and automatically generate accurate and standardized meeting minutes based on that information. Furthermore, insufficient real-time information integration and consistent standardization of technical terms result in a lack of information consistency, making it inconvenient to refer to later.

[0055] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0056] In this invention, the server includes a conversion device for converting audio information into text information, an adjustment device for standardizing technical terms in the text information, and a management device for recording and distributing the adjusted text information. This enables the efficient and accurate generation and visualization of meeting minutes in real time during a meeting.

[0057] A "perceptual device" is a device for effectively collecting audio and visual information in real time during a meeting.

[0058] A "conversion device" is a device that quickly and accurately converts collected audio information into text information.

[0059] A "regulation device" is a device used to standardize and correct specialized terminology in textual information into a standardized expression.

[0060] A "representation device" is a device for interactively representing adjusted textual information on visible information.

[0061] A "management device" is a device for recording, distributing, and properly managing adjusted textual information.

[0062] This invention provides a system that effectively utilizes audio and visual information from meetings to automatically generate accurate meeting minutes in real time. Specific embodiments of the invention are described below.

[0063] During a meeting, the device collects audio information using its microphone. It also simultaneously captures visual information, such as slides and documents displayed on a projector or display device, using its connected camera and screen capture function. To record this information in high resolution, the device is equipped with optimal audio filtering and high-performance image processing software.

[0064] The server receives audio information transmitted from the terminal and converts it into text using a generative AI model. During this process, it utilizes advanced Japanese natural language processing techniques to grammatically refine the information extracted from the audio. The AI ​​model used by the server includes speech recognition technology.

[0065] The generated text information is standardized for technical terms by a server-side adjustment device. By referring to an internal database and unifying technical terms into appropriate expressions, meeting minutes with a consistent format are formed.

[0066] The adjusted text information is resent to the device and displayed overlaid on the captured visual information. The font and layout of this display are adjusted for improved readability for the user. Users can review the meeting minutes generated in real time via their device and provide feedback as needed.

[0067] The finalized meeting minutes are securely stored on the server and uploaded to cloud storage or a shared folder with appropriate access permissions. This allows users to easily share the meeting minutes later as needed.

[0068] As a concrete example, consider a meeting where new product development proposals are discussed. During this meeting, a terminal captures the proposal content, and the server standardizes the terminology, ensuring consistent information that can be easily referenced by anyone at a later date.

[0069] An example of a prompt to input into a generative AI model is the instruction, "Transcribe the meeting audio, correct for technical jargon, and generate meeting minutes." This is an important prompt for reliably converting audio information into text information and automatically creating standardized meeting minutes.

[0070] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0071] Step 1:

[0072] The device uses its microphone to collect audio information from the meeting. In addition, the device uses its connected camera and screen capture function to capture visual information such as meeting slides and documents in high resolution. The input is the audio and video of the slides during the meeting, and the output is a file containing these recordings in digital format. Specifically, the device performs noise reduction and compresses the video to efficiently store the data.

[0073] Step 2:

[0074] The terminal formats the collected audio as digital data and sends it to the server via a secure communication protocol. The input is the formatted audio data, and the output is the audio file sent to the server. Specifically, the terminal encrypts the data and performs security checks to prevent unauthorized access.

[0075] Step 3:

[0076] The server inputs the received audio data into a generating AI model, which then converts it into text information using speech recognition technology. The input is audio data, and the output is grammatically corrected text data. The server performs a process of dividing the audio into segments, converting them into phonemes, and generating string data.

[0077] Step 4:

[0078] The server compares technical terms in the generated text information with an internal database and uses a reconciliation device to standardize them. The input is text data obtained from the generating AI model, and the output is text data with the technical terms reconciled. Specifically, it performs conversion processing to standardize terms such as "AI" to "artificial intelligence."

[0079] Step 5:

[0080] The terminal receives adjusted text information and displays it on the screen, overlaid with captured video data. The input is adjusted text data sent from the server, and the output is the combined data displayed in a highly legible font and layout. Specifically, the terminal adjusts the display position and changes the font size to make the information easy for the user to understand.

[0081] Step 6:

[0082] The server securely stores the finalized textual information as the meeting minutes and uploads it to a shared folder or the cloud with appropriate access permissions. The input is the final textual data, and the output is an accessible digital record. Specifically, the server backs up the data and sends access notifications to the necessary users.

[0083] (Application Example 1)

[0084] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0085] In business environments such as data centers, efficient meeting management and accurate recording of meeting content are required. However, the diversity of specialized terminology and the inability to share information immediately are obstacles, contributing to decreased work efficiency. Therefore, a system is needed that accurately records meeting content, standardizes specialized terminology, and enables immediate information sharing between multiple terminals.

[0086] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0087] In this invention, the server includes collection means for acquiring audio and video data, generation means for converting audio data into text information, and correction means for standardizing terminology in the text information. This enables the generation of accurate meeting minutes in real time during a meeting and allows for immediate information sharing among multiple terminals.

[0088] "Collection means" refers to a device or method for acquiring audio and video data during a meeting.

[0089] "Generation means" refers to a process or system for converting collected audio data into text information.

[0090] "Correction means" refers to a function or technique for modifying technical terms or specific terminology in text information in order to standardize them.

[0091] "Display means" refers to a function or device for visually presenting corrected text information on video data.

[0092] "Storage and sharing means" refers to a procedure or device for storing corrected text information and sharing that information with those who have access rights.

[0093] "Integrated distribution means" refers to a function or method for integrating information collected from multiple terminals and distributing that information appropriately.

[0094] To implement this invention, it is necessary to establish a system that collects audio and video data in a conference environment and processes it in real time. The system's terminals use microphones and cameras to acquire audio and materials displayed on a projector or screen during the conference. This data is transmitted to a server via a network.

[0095] The server converts the collected audio data into text information using the Google® Cloud Speech-to-Text API. Because the converted text may not accurately represent technical terms, the server corrects the terminology within the text using the Hugging Face Transformers library. This correction process involves retrieving relevant terminology information from an internal data store and performing standardization. The corrected text information is then displayed on participants' devices via the Flask server. This displayed text information is visible to meeting participants and serves as an aid to facilitate understanding.

[0096] Furthermore, the server securely stores the corrected text information in cloud storage and allows users to share it by setting access permissions. This feature makes it easy to refer to the information as meeting minutes even after the meeting has ended. Because actions based on the meeting content can be taken quickly, work efficiency can be improved.

[0097] As a concrete example, consider a system operations meeting held in a data center. In this system, participants discuss new network solutions they propose, and audio and video are collected in real time, with meeting minutes generated immediately. This allows for a rapid and standardized understanding of the technical details discussed in the meeting.

[0098] An example of a prompt for a generative AI model is as follows:

[0099] Prompt: Accurately recognize the technical terms discussed during the meeting, correct the generated text information, and produce accurate meeting minutes in Japanese.

[0100] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0101] Step 1:

[0102] The device's microphone and camera capture audio and video during the meeting. Audio is received as an analog signal, and video as pixel data. The device converts this data into a digital format, performs initial processing, and then transmits it to the server via the network.

[0103] Step 2:

[0104] The server sends the received audio data to the Google Cloud Speech-to-Text API, converting the audio signal into text information. Audio data is provided as input, and a speech recognition algorithm produces text output as a string. This converted text information is stored on the server.

[0105] Step 3:

[0106] The server uses Hugging Face Transformers to correct specialized terminology in text information. Given generated text information as input, it converts relevant specialized terms into appropriately standardized terms. Specifically, it refers to an internal data store, performs a conversion process to applicable terms, and obtains corrected text information.

[0107] Step 4:

[0108] The server uses Flask to send corrected text information to the client's terminal and displays it integrated with the video data. Corrected text information is provided as input, and the output is displayed on the video in real time via the user's visual interface.

[0109] Step 5:

[0110] The user reviews the information displayed through the terminal interface and modifies or adds annotations to the meeting minutes as needed. The displayed information is provided as input, and the final meeting minutes data is generated through user interaction.

[0111] Step 6:

[0112] The server securely stores the corrected text information in cloud storage and shares it among users as needed. The final meeting minutes data is provided as input, stored under appropriate access controls, and made available to other users via links and notifications.

[0113] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0114] This invention aims to improve the accuracy and convenience of information by incorporating an emotion engine that recognizes user emotions in real time into a system for automatically generating meeting minutes with high accuracy. This system achieves effective meeting minute generation by collecting audio and screen data, converting it to text, correcting technical terms, displaying the text, saving and sharing data, and providing emotion analysis capabilities.

[0115] At the start of the meeting, the device collects audio via the microphone and video for capturing meeting materials and slides. Simultaneously, the emotion engine analyzes the audio data to determine the emotional state of participants based on their tone and tempo. The determined emotional information is used to deepen understanding of the meeting's progress through the meeting minutes.

[0116] The server uses an AI model to generate audio data and convert it into text data in real time. This text is grammatically corrected using natural language processing technology, and further standardization of technical terms is performed using the RAG module. By combining the transcribed data with sentiment data, a more comprehensive meeting transcript is generated that includes the context and emotional nuances of the conversation.

[0117] The corrected text is displayed on the device screen. The sentiment engine's analysis results are reflected here, and the style and emphasis of the text display on the screen may be dynamically adjusted. For example, statements indicating strong emotions by the user may have a larger font size to draw attention.

[0118] The final meeting minutes are stored on a server and provided to users with access rights via the cloud system or on-premises network. This allows attendees and relevant departments to quickly and smoothly share and utilize the minutes after the meeting.

[0119] As a concrete example, if a participant shows strong excitement during a new product announcement at a meeting, that emotion will be recorded. This information will be communicated to other stakeholders through the meeting minutes and used as a reference for new product development. In this way, by combining it with an emotion engine, it is possible to capture not only textual information but also the overall atmosphere of the meeting.

[0120] The following describes the processing flow.

[0121] Step 1:

[0122] The device uses a microphone at the start of the meeting to collect participants' voices in real time and simultaneously captures the meeting materials screen. This is a crucial step in enabling the recording of the entire meeting's audio and visual information in digital format.

[0123] Step 2:

[0124] The server acquires audio data transmitted from the terminal and converts it into text data in real time using a generative AI model. This process transforms speech into text information, guaranteeing grammatical accuracy.

[0125] Step 3:

[0126] The server sends the generated text to the RAG module, where it standardizes the technical terms within the text. Specifically, it compares identified terms with the company's internal database and replaces them with appropriate terms, thereby improving the consistency and clarity of the discussion.

[0127] Step 4:

[0128] The device uses an emotion engine to analyze the user's emotions from the audio data for the generated text. The analysis results are reflected in the text's display style; for example, emotionally rich statements that should be emphasized are visually highlighted with different fonts or colors.

[0129] Step 5:

[0130] The device combines corrected text and sentiment data, overlaying it onto the screen capture video. This display is designed to allow users to intuitively understand the key points and emotional nuances of the meeting.

[0131] Step 6:

[0132] The server securely stores the final text data and uploads it to cloud storage or an internal network accessible to all stakeholders. This allows users to easily access, review, and share the generated meeting minutes after the meeting.

[0133] (Example 2)

[0134] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0135] In generating meeting minutes, conventional technologies could transcribe and display speech, but they could not generate information that took into account the emotions of the participants, making it difficult to grasp the atmosphere of the meeting and the changes in emotions. Furthermore, there were challenges in maintaining consistency in technical terms and the real-time nature of the generated text, which could reduce the accuracy and usefulness of the information.

[0136] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0137] In this invention, the server includes collection means for collecting audio and video information, generation means for converting audio information into text information, correction means for standardizing specialized terminology in the text information, and analysis and integration means for analyzing emotional information and integrating it into the text information. This makes it possible not only to record the contents of a meeting as text information, but also to generate a three-dimensional meeting record that reflects the emotions of the participants. Furthermore, it is possible to provide highly accurate information in real time while maintaining consistency in specialized terminology.

[0138] "Audio information" refers to data about conversations and voices collected using microphones or other audio input devices.

[0139] "Visual information" refers to visual data such as meeting materials and slides, collected using cameras or other visual input devices.

[0140] "Collection means" refers to technical means and devices for acquiring audio and video information, thereby incorporating the data into the system.

[0141] "Generation means" refers to technical means or devices for converting collected audio information into text information, and involves using a generation AI model to perform text conversion.

[0142] "Correction means" refers to technical means or processes for standardizing and correcting specialized terminology and grammar contained in generated textual information.

[0143] "Display means" refers to technical devices or methods for visually displaying corrected text information on video information, thereby providing users with easily understandable information.

[0144] "Means of saving and sharing" refers to technical means or systems for saving corrected text information and securely sharing it with users who have access rights.

[0145] "Analysis and integration means" refers to technical means or processes for analyzing emotional information from audio and video information and integrating it into textual information.

[0146] This invention is a system for automatically generating meeting minutes, aiming to create a three-dimensional meeting record that includes emotional information by combining audio and video information. This system mainly consists of terminals and a server.

[0147] The device uses a microphone to collect audio and a camera to capture video information such as meeting materials and slides. This allows for a comprehensive recording of the meeting's content and situation. The collected audio and video information is transmitted to the server in real time.

[0148] The server uses a generative AI model to instantly convert audio information into text. The converted text is then grammatically corrected using natural language processing techniques. Furthermore, the RAG module is used as a correction tool to standardize specialized terminology. This improves the consistency and accuracy of the meeting minutes.

[0149] For emotion analysis, the server uses an emotion engine based on audio information to determine the emotional state of participants from their tone and tempo. This analysis is integrated into the text information, visually reflecting the atmosphere and activity level of the meeting. For example, it's possible to visually emphasize key statements by increasing the font size.

[0150] The final meeting minutes are securely stored on the server and shared with authorized users via the cloud system. This allows for quick and smooth information sharing after the meeting.

[0151] An example of a prompt might be, "As a prompt to input into the generation AI model, please summarize the sentiment analysis of participants regarding specific statements made during the meeting, and how those sentiments will be reflected in the meeting minutes." In this way, the entire system working together enables the generation of highly accurate meeting minutes.

[0152] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0153] Step 1:

[0154] The device uses a microphone to collect audio information. When the meeting begins, the device converts the audio into digital data in real time and sends it to the server. At this stage, the input is raw audio, and the output is digital audio data. Similarly, when collecting video information, the device uses a camera to capture visual information of the meeting and sends the digital video data to the server.

[0155] Step 2:

[0156] The server converts speech information into text information using a generative AI model. The server analyzes the digital speech data received as input and converts the utterances into text. Noise reduction and speaker identification are also performed during this process, and the output is text data. Furthermore, natural language processing techniques are applied to the converted text to perform grammatical correction.

[0157] Step 3:

[0158] The server uses the RAG module as a correction tool to standardize specialized terminology in textual information. The input is text data that has undergone natural language processing, and the RAG module compares it with a business-specific database and replaces it with standard terminology. The output here is consistent, standardized text data.

[0159] Step 4:

[0160] The server uses an emotion engine to analyze participants' emotions from audio information and integrate it into text information. The input consists of digital audio and video data, and the server determines emotional states from voice tone, tempo, and facial expressions. This analysis results in output data with emotional information added to the text.

[0161] Step 5:

[0162] The terminal receives integrated text data sent from the server and displays it visually on the screen. The input is text data containing emotional information, which is displayed through dynamic visual adjustments. For example, emphasized statements may be displayed with a larger font size. The output is visually emphasized text.

[0163] Step 6:

[0164] The server securely stores the final meeting minutes data and shares it with authorized users via the cloud. The input is the finalized meeting minutes data, which is encrypted and stored, and a shareable link is generated. The output is the meeting minutes data accessible to users.

[0165] (Application Example 2)

[0166] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0167] Manually creating meeting minutes in conferences and business meetings is time-consuming and may leave out important information. Furthermore, accurately recording and sharing participants' emotions and the atmosphere of the meeting is difficult. This can lead to a lack of information in post-meeting communication and decision-making. Moreover, accurate information retention and visualization of participants' emotions are also required in staff meetings within stores and in interactions with customers.

[0168] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0169] In this invention, the server includes a collection means for collecting information during a meeting, an analysis means for analyzing the emotions of participants, and a highlighting means for dynamically highlighting the embedded string data. This makes it possible to create three-dimensional meeting minutes that reflect the emotions of participants, even in meetings held in physical stores, while incorporating emotional information into meeting minutes that are instantly transcribed and corrected from audio.

[0170] "Collection means" refers to a device or function for collecting information such as audio data and video data during a meeting.

[0171] "Generation means" refers to a technology or device for converting collected audio data into string data.

[0172] "Correction means" refers to a method or process for standardizing and grammatically correcting technical terms in generated string data.

[0173] "Display means" refers to a method or apparatus for visually presenting corrected string data on a screen or display.

[0174] "Storage and provision means" refers to a function or mechanism that stores corrected string data in a storage device and makes it available to the user as needed.

[0175] "Analysis means" refers to technology for analyzing the emotional state of meeting participants from their voices and facial expressions and extracting that information.

[0176] "Embedding means" refers to a process or device for reflecting emotional information obtained through analysis into string data.

[0177] "Highlighting means" refers to a device or function for dynamically highlighting and displaying string data that incorporates emotional information.

[0178] The system realizing this invention first includes a means for efficiently collecting audio and video data during a meeting. A terminal collects audio data through a microphone and transmits it to a server in real time. The server uses a generative AI model to instantly convert the audio data into text data. As a result, the content spoken during the meeting is transcribed sequentially.

[0179] Next, the server uses natural language processing technology to grammatically correct the generated string data. The RAG process is used as a correction method to appropriately standardize technical terms within the text data based on internal information sources. Through this process, the string data becomes accurate and consistent.

[0180] As an analytical method, the emotional state of participants is grasped in real time from audio and video data. This utilizes voice analysis technology and facial recognition technology to analyze emotions from the tone of the user's voice and facial expressions.

[0181] The analyzed emotional information is embedded into string data using a built-in mechanism. This data is dynamically highlighted on the device, visually representing differences based on emotion. For example, statements expressing particularly heightened emotions can be made easier to distinguish from other information by increasing the font size or changing the color.

[0182] As a concrete example, in a meeting about product displays held within a physical store, the opinions expressed by staff members regarding new display designs are recorded along with their emotions. If a staff member expresses a positive opinion, that part of the text is highlighted for easier viewing.

[0183] An example of a prompt message would be, "Analyze staff reactions to the new product display and add sentiment data to the meeting minutes." This allows participants and relevant departments to review the meeting content in detail and with emotional nuances after the meeting.

[0184] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0185] Step 1:

[0186] The device collects audio data using a microphone during a meeting. The input is the voices of meeting participants, and the output is a digital audio file. This enables real-time collection of audio data.

[0187] Step 2:

[0188] The server converts the received audio data into text data using a generation AI model. The input is a digital audio file, and the output is text data. The conversion process instantly transcribes the audio content into text, and the recording of the meeting proceeds.

[0189] Step 3:

[0190] The server uses natural language processing techniques to correct the grammar of string data. The input is the converted string data, and the output is the corrected string data. Grammar correction improves the accuracy and readability of the data.

[0191] Step 4:

[0192] The server utilizes the RAG process to standardize technical terms within string data based on internal sources. The input is grammatically corrected string data, and the output is the corrected, standardized string data. This process ensures consistency in terminology.

[0193] Step 5:

[0194] The server uses audio and video data to analyze participants' emotions. The input is the collected audio and video data, and the output is the analyzed emotional information. This extracts the emotional nuances of the meeting.

[0195] Step 6:

[0196] The server incorporates the analyzed sentiment information into the corrected string data. The input consists of the sentiment information and the corrected string data, and the output is unified string data with the sentiment information incorporated. This integrates the data in a way that includes emotional nuances.

[0197] Step 7:

[0198] The device dynamically highlights embedded string data. The input is integrated string data, and the output is visually highlighted screen data. Specifically, this includes actions such as increasing the font size or changing the color of statements that express particularly strong emotions.

[0199] Step 8:

[0200] Users view highlighted text data on the screen and save it as meeting minutes. The input is the highlighted screen data, and the output is the saved meeting minutes file. Meeting minutes, reflecting sentiment data, are later used for decision-making and reference in other meetings.

[0201] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0202] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0203] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0204] [Second Embodiment]

[0205] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0206] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0207] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0208] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0209] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0210] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0211] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0212] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0213] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0214] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0215] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0216] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0217] This invention is embodied in a system that automatically and accurately generates meeting minutes. This system is characterized by its ability to simultaneously collect audio and screen data and to correct for technical terms.

[0218] This system consists of several main components. First, during a meeting, the terminal collects audio through a microphone and captures video of slides and documents displayed on the projector or PC screen. This allows various data from the meeting to be recorded in real time.

[0219] Next, the server receives the audio data and uses a generative AI to convert it into text data. At this stage, natural language processing techniques that take into account the characteristics of the Japanese language are used, and general grammatical corrections are made.

[0220] The generated text is further refined by the RAG module on the server. Here, the latest technical terms stored in the database are referenced, and identified words are corrected. For example, the term "AI" is corrected and standardized to "artificial intelligence."

[0221] This corrected text is returned to the device and overlaid on the captured screen data. The display format is adjusted for user readability, and the text is displayed in conjunction with the slide content, allowing for immediate review during the meeting.

[0222] The final meeting minutes data is stored on the server in a secure manner. It is then uploaded to a shared folder or appropriate platform so that those with access rights can access the data. At this stage, users receive notifications and can proactively distribute the data to relevant parties after making individual revisions or additions as needed.

[0223] As a concrete example, consider a scenario where a new proposal for product development is made at a specific meeting. A terminal collects the details of the proposal in real time, and meeting minutes are generated using standardized technical terminology, ensuring consistency and ease of understanding for future reference. This contributes to increased work efficiency and improved communication.

[0224] The following describes the processing flow.

[0225] Step 1:

[0226] The device collects audio data via the microphone at the start of the meeting and activates its screen capture function to record slides and materials displayed during the meeting. This ensures that all audio and visual information is simultaneously acquired as digital data.

[0227] Step 2:

[0228] The server receives audio data transmitted in real time and converts it into text data using a generative AI model. Here, speech recognition technology is used to transcribe the content of the audio into text, and basic Japanese grammatical corrections are applied.

[0229] Step 3:

[0230] The server sends the generated text to the RAG module, where it corrects any mistranslations of technical terms. RAG refers to the latest terminology standards in its database and replaces technical terms in the text with appropriate words as needed. This process involves natural language processing techniques.

[0231] Step 4:

[0232] The device displays the corrected text data overlaid on the screen footage captured during the meeting. The font and layout are adjusted to improve text readability, allowing the user to instantly review the corrected text while viewing the screen.

[0233] Step 5:

[0234] The server securely stores the final meeting minutes data after formatting it. Data storage is performed using encryption technology, and the files are uploaded to a shared folder accessible to all relevant parties.

[0235] Step 6:

[0236] Users receive notifications and review the meeting minutes on their devices. After making any necessary corrections or adding annotations, they can distribute the minutes to other attendees and relevant departments via email or internal communication tools.

[0237] (Example 1)

[0238] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0239] It is difficult to efficiently collect audio and visual information during meetings and automatically generate accurate and standardized meeting minutes based on that information. Furthermore, insufficient real-time information integration and consistent standardization of technical terms result in a lack of information consistency, making it inconvenient to refer to later.

[0240] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0241] In this invention, the server includes a conversion device for converting audio information into text information, an adjustment device for standardizing technical terms in the text information, and a management device for recording and distributing the adjusted text information. This enables the efficient and accurate generation and visualization of meeting minutes in real time during a meeting.

[0242] A "perceptual device" is a device for effectively collecting audio and visual information in real time during a meeting.

[0243] A "conversion device" is a device that quickly and accurately converts collected audio information into text information.

[0244] A "regulation device" is a device used to standardize and correct specialized terminology in textual information into a standardized expression.

[0245] A "representation device" is a device for interactively representing adjusted textual information on visible information.

[0246] A "management device" is a device for recording, distributing, and properly managing adjusted textual information.

[0247] This invention provides a system that effectively utilizes audio and visual information from meetings to automatically generate accurate meeting minutes in real time. Specific embodiments of the invention are described below.

[0248] During a meeting, the device collects audio information using its microphone. It also simultaneously captures visual information, such as slides and documents displayed on a projector or display device, using its connected camera and screen capture function. To record this information in high resolution, the device is equipped with optimal audio filtering and high-performance image processing software.

[0249] The server receives audio information transmitted from the terminal and converts it into text using a generative AI model. During this process, it utilizes advanced Japanese natural language processing techniques to grammatically refine the information extracted from the audio. The AI ​​model used by the server includes speech recognition technology.

[0250] The generated text information is standardized for technical terms by a server-side adjustment device. By referring to an internal database and unifying technical terms into appropriate expressions, meeting minutes with a consistent format are formed.

[0251] The adjusted text information is resent to the device and displayed overlaid on the captured visual information. The font and layout of this display are adjusted for improved readability for the user. Users can review the meeting minutes generated in real time via their device and provide feedback as needed.

[0252] The finalized meeting minutes are securely stored on the server and uploaded to cloud storage or a shared folder with appropriate access permissions. This allows users to easily share the meeting minutes later as needed.

[0253] As a concrete example, consider a meeting where new product development proposals are discussed. During this meeting, a terminal captures the proposal content, and the server standardizes the terminology, ensuring consistent information that can be easily referenced by anyone at a later date.

[0254] An example of a prompt to input into a generative AI model is the instruction, "Transcribe the meeting audio, correct for technical jargon, and generate meeting minutes." This is an important prompt for reliably converting audio information into text information and automatically creating standardized meeting minutes.

[0255] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0256] Step 1:

[0257] The device uses its microphone to collect audio information from the meeting. In addition, the device uses its connected camera and screen capture function to capture visual information such as meeting slides and documents in high resolution. The input is the audio and video of the slides during the meeting, and the output is a file containing these recordings in digital format. Specifically, the device performs noise reduction and compresses the video to efficiently store the data.

[0258] Step 2:

[0259] The terminal formats the collected audio as digital data and sends it to the server via a secure communication protocol. The input is the formatted audio data, and the output is the audio file sent to the server. Specifically, the terminal encrypts the data and performs security checks to prevent unauthorized access.

[0260] Step 3:

[0261] The server inputs the received audio data into a generating AI model, which then converts it into text information using speech recognition technology. The input is audio data, and the output is grammatically corrected text data. The server performs a process of dividing the audio into segments, converting them into phonemes, and generating string data.

[0262] Step 4:

[0263] The server compares technical terms in the generated text information with an internal database and uses a reconciliation device to standardize them. The input is text data obtained from the generating AI model, and the output is text data with the technical terms reconciled. Specifically, it performs conversion processing to standardize terms such as "AI" to "artificial intelligence."

[0264] Step 5:

[0265] The terminal receives adjusted text information and displays it on the screen, overlaid with captured video data. The input is adjusted text data sent from the server, and the output is the combined data displayed in a highly legible font and layout. Specifically, the terminal adjusts the display position and changes the font size to make the information easy for the user to understand.

[0266] Step 6:

[0267] The server securely stores the finalized textual information as the meeting minutes and uploads it to a shared folder or the cloud with appropriate access permissions. The input is the final textual data, and the output is an accessible digital record. Specifically, the server backs up the data and sends access notifications to the necessary users.

[0268] (Application Example 1)

[0269] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0270] In business environments such as data centers, efficient meeting management and accurate recording of meeting content are required. However, the diversity of specialized terminology and the inability to share information immediately are obstacles, contributing to decreased work efficiency. Therefore, a system is needed that accurately records meeting content, standardizes specialized terminology, and enables immediate information sharing between multiple terminals.

[0271] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0272] In this invention, the server includes collection means for acquiring audio and video data, generation means for converting audio data into text information, and correction means for standardizing terminology in the text information. This enables the generation of accurate meeting minutes in real time during a meeting and allows for immediate information sharing among multiple terminals.

[0273] "Collection means" refers to a device or method for acquiring audio and video data during a meeting.

[0274] "Generation means" refers to a process or system for converting collected audio data into text information.

[0275] "Correction means" refers to a function or technique for modifying technical terms or specific terminology in text information in order to standardize them.

[0276] "Display means" refers to a function or device for visually presenting corrected text information on video data.

[0277] "Storage and sharing means" refers to a procedure or device for storing corrected text information and sharing that information with those who have access rights.

[0278] "Integrated distribution means" refers to a function or method for integrating information collected from multiple terminals and distributing that information appropriately.

[0279] To implement this invention, it is necessary to establish a system that collects audio and video data in a conference environment and processes it in real time. The system's terminals use microphones and cameras to acquire audio and materials displayed on a projector or screen during the conference. This data is transmitted to a server via a network.

[0280] The server converts the collected audio data into text using the Google Cloud Speech-to-Text API. Because the converted text may not accurately represent technical terms, the server corrects the terminology within the text using the Hugging Face Transformers library. This correction process involves retrieving relevant terminology information from an internal data store and performing standardization. The corrected text is then displayed on participants' devices via the Flask server. This displayed text is visible to meeting participants and serves as an aid to facilitate understanding.

[0281] Furthermore, the server securely stores the corrected text information in cloud storage and allows users to share it by setting access permissions. This feature makes it easy to refer to the information as meeting minutes even after the meeting has ended. Because actions based on the meeting content can be taken quickly, work efficiency can be improved.

[0282] As a concrete example, consider a system operations meeting held in a data center. In this system, participants discuss new network solutions they propose, and audio and video are collected in real time, with meeting minutes generated immediately. This allows for a rapid and standardized understanding of the technical details discussed in the meeting.

[0283] Examples of prompt texts for generating AI models are as follows:

[0284] Prompt: Accurately recognize technical terms discussed during a meeting, correct the generated text information, and generate accurate Japanese meeting minutes.

[0285] The flow of specific processing in Application Example 1 will be described using FIG. 12.

[0286] Step 1:

[0287] The microphone and camera of the terminal acquire the audio and video during the meeting. The audio is obtained as an analog signal, and the video is obtained as pixel data. The terminal converts these data into a digital format, performs initial processing, and then transfers them to the server via the network.

[0288] Step 2:

[0289] The server sends the received audio data to the Google Cloud Speech-to-Text API to convert the audio signal into text information. Audio data is provided as input, and a text output as a character string is obtained by the speech recognition algorithm. This converted text information is retained within the server.

[0290] Step 3:

[0291] The server uses Hugging Face Transformers to perform a process of correcting the technical terms in the text information. The generated text information is provided as input, and the relevant technical terms are converted into appropriately standardized terms. Specifically, by referring to the internal data store, a conversion process to applicable terms is performed, and corrected text information is obtained.

[0292] Step 4:

[0293] The server uses Flask to send corrected text information to the client's terminal and displays it integrated with the video data. Corrected text information is provided as input, and the output is displayed on the video in real time via the user's visual interface.

[0294] Step 5:

[0295] The user reviews the information displayed through the terminal interface and modifies or adds annotations to the meeting minutes as needed. The displayed information is provided as input, and the final meeting minutes data is generated through user interaction.

[0296] Step 6:

[0297] The server securely stores the corrected text information in cloud storage and shares it among users as needed. The final meeting minutes data is provided as input, stored under appropriate access controls, and made available to other users via links and notifications.

[0298] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0299] This invention aims to improve the accuracy and convenience of information by incorporating an emotion engine that recognizes user emotions in real time into a system for automatically generating meeting minutes with high accuracy. This system achieves effective meeting minute generation by collecting audio and screen data, converting it to text, correcting technical terms, displaying the text, saving and sharing data, and providing emotion analysis capabilities.

[0300] At the start of the meeting, the device collects audio via the microphone and video for capturing meeting materials and slides. Simultaneously, the emotion engine analyzes the audio data to determine the emotional state of participants based on their tone and tempo. The determined emotional information is used to deepen understanding of the meeting's progress through the meeting minutes.

[0301] The server uses an AI model to generate audio data and convert it into text data in real time. This text is grammatically corrected using natural language processing technology, and further standardization of technical terms is performed using the RAG module. By combining the transcribed data with sentiment data, a more comprehensive meeting transcript is generated that includes the context and emotional nuances of the conversation.

[0302] The corrected text is displayed on the device screen. The sentiment engine's analysis results are reflected here, and the style and emphasis of the text display on the screen may be dynamically adjusted. For example, statements indicating strong emotions by the user may have a larger font size to draw attention.

[0303] The final meeting minutes are stored on a server and provided to users with access rights via the cloud system or on-premises network. This allows attendees and relevant departments to quickly and smoothly share and utilize the minutes after the meeting.

[0304] As a concrete example, if a participant shows strong excitement during a new product announcement at a meeting, that emotion will be recorded. This information will be communicated to other stakeholders through the meeting minutes and used as a reference for new product development. In this way, by combining it with an emotion engine, it is possible to capture not only textual information but also the overall atmosphere of the meeting.

[0305] The following describes the processing flow.

[0306] Step 1:

[0307] At the start of the meeting, the terminal uses the microphone to collect the voices of participants in real time and simultaneously captures the screen of the meeting materials. This operation is an important step to enable the recording of the entire meeting's audio and visual information in digital form.

[0308] Step 2:

[0309] The server acquires the voice data transmitted from the terminal and uses the generated AI model to convert it into text data in real time. This process transforms the voice into character information and ensures grammatical accuracy.

[0310] Step 3:

[0311] The server sends the generated text to the RAG module to standardize the technical terms in the text. Specifically, by comparing the identified terms with the in-house database and replacing them with appropriate terms, the consistency and clarity of the discussion are improved.

[0312] Step 4:

[0313] The terminal utilizes the emotion engine for the generated text to analyze the user's emotions from the voice data. The analysis results are reflected in the display style of the text. For example, emotionally rich statements that should be emphasized are visually emphasized with different fonts or colors.

[0314] Step 5:

[0315] The terminal combines the corrected text and the emotion data and overlays and displays them on the screen capture video. This display is designed to enable the user to intuitively understand the key points and emotional nuances of the meeting.

[0316] Step 6:

[0317] The server securely stores the final text data and uploads it to cloud storage or an internal network accessible to all stakeholders. This allows users to easily access, review, and share the generated meeting minutes after the meeting.

[0318] (Example 2)

[0319] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0320] In generating meeting minutes, conventional technologies could transcribe and display speech, but they could not generate information that took into account the emotions of the participants, making it difficult to grasp the atmosphere of the meeting and the changes in emotions. Furthermore, there were challenges in maintaining consistency in technical terms and the real-time nature of the generated text, which could reduce the accuracy and usefulness of the information.

[0321] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0322] In this invention, the server includes collection means for collecting audio and video information, generation means for converting audio information into text information, correction means for standardizing specialized terminology in the text information, and analysis and integration means for analyzing emotional information and integrating it into the text information. This makes it possible not only to record the contents of a meeting as text information, but also to generate a three-dimensional meeting record that reflects the emotions of the participants. Furthermore, it is possible to provide highly accurate information in real time while maintaining consistency in specialized terminology.

[0323] "Audio information" refers to data about conversations and voices collected using microphones or other audio input devices.

[0324] "Visual information" refers to visual data such as meeting materials and slides, collected using cameras or other visual input devices.

[0325] "Collection means" refers to technical means and devices for acquiring audio and video information, thereby incorporating the data into the system.

[0326] "Generation means" refers to technical means or devices for converting collected audio information into text information, and involves using a generation AI model to perform text conversion.

[0327] "Correction means" refers to technical means or processes for standardizing and correcting specialized terminology and grammar contained in generated textual information.

[0328] "Display means" refers to technical devices or methods for visually displaying corrected text information on video information, thereby providing users with easily understandable information.

[0329] "Means of saving and sharing" refers to technical means or systems for saving corrected text information and securely sharing it with users who have access rights.

[0330] "Analysis and integration means" refers to technical means or processes for analyzing emotional information from audio and video information and integrating it into textual information.

[0331] This invention is a system for automatically generating meeting minutes, aiming to create a three-dimensional meeting record that includes emotional information by combining audio and video information. This system mainly consists of terminals and a server.

[0332] The device uses a microphone to collect audio and a camera to capture video information such as meeting materials and slides. This allows for a comprehensive recording of the meeting's content and situation. The collected audio and video information is transmitted to the server in real time.

[0333] The server uses a generative AI model to instantly convert audio information into text. The converted text is then grammatically corrected using natural language processing techniques. Furthermore, the RAG module is used as a correction tool to standardize specialized terminology. This improves the consistency and accuracy of the meeting minutes.

[0334] For emotion analysis, the server uses an emotion engine based on audio information to determine the emotional state of participants from their tone and tempo. This analysis is integrated into the text information, visually reflecting the atmosphere and activity level of the meeting. For example, it's possible to visually emphasize key statements by increasing the font size.

[0335] The final meeting minutes are securely stored on the server and shared with authorized users via the cloud system. This allows for quick and smooth information sharing after the meeting.

[0336] An example of a prompt might be, "As a prompt to input into the generation AI model, please summarize the sentiment analysis of participants regarding specific statements made during the meeting, and how those sentiments will be reflected in the meeting minutes." In this way, the entire system working together enables the generation of highly accurate meeting minutes.

[0337] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0338] Step 1:

[0339] The device uses a microphone to collect audio information. When the meeting begins, the device converts the audio into digital data in real time and sends it to the server. At this stage, the input is raw audio, and the output is digital audio data. Similarly, when collecting video information, the device uses a camera to capture visual information of the meeting and sends the digital video data to the server.

[0340] Step 2:

[0341] The server converts speech information into text information using a generative AI model. The server analyzes the digital speech data received as input and converts the utterances into text. Noise reduction and speaker identification are also performed during this process, and the output is text data. Furthermore, natural language processing techniques are applied to the converted text to perform grammatical correction.

[0342] Step 3:

[0343] The server uses the RAG module as a correction tool to standardize specialized terminology in textual information. The input is text data that has undergone natural language processing, and the RAG module compares it with a business-specific database and replaces it with standard terminology. The output here is consistent, standardized text data.

[0344] Step 4:

[0345] The server uses an emotion engine to analyze participants' emotions from audio information and integrate it into text information. The input consists of digital audio and video data, and the server determines emotional states from voice tone, tempo, and facial expressions. This analysis results in output data with emotional information added to the text.

[0346] Step 5:

[0347] The terminal receives integrated text data sent from the server and displays it visually on the screen. The input is text data containing emotional information, which is displayed through dynamic visual adjustments. For example, emphasized statements may be displayed with a larger font size. The output is visually emphasized text.

[0348] Step 6:

[0349] The server securely stores the final meeting minutes data and shares it with authorized users via the cloud. The input is the finalized meeting minutes data, which is encrypted and stored, and a shareable link is generated. The output is the meeting minutes data accessible to users.

[0350] (Application Example 2)

[0351] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0352] Manually creating meeting minutes in conferences and business meetings is time-consuming and may leave out important information. Furthermore, accurately recording and sharing participants' emotions and the atmosphere of the meeting is difficult. This can lead to a lack of information in post-meeting communication and decision-making. Moreover, accurate information retention and visualization of participants' emotions are also required in staff meetings within stores and in interactions with customers.

[0353] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0354] In this invention, the server includes a collection means for collecting information during a meeting, an analysis means for analyzing the emotions of participants, and a highlighting means for dynamically highlighting the embedded string data. This makes it possible to create three-dimensional meeting minutes that reflect the emotions of participants, even in meetings held in physical stores, while incorporating emotional information into meeting minutes that are instantly transcribed and corrected from audio.

[0355] "Collection means" refers to a device or function for collecting information such as audio data and video data during a meeting.

[0356] "Generation means" refers to a technology or device for converting collected audio data into string data.

[0357] "Correction means" refers to a method or process for standardizing and grammatically correcting technical terms in generated string data.

[0358] "Display means" refers to a method or apparatus for visually presenting corrected string data on a screen or display.

[0359] "Storage and provision means" refers to a function or mechanism that stores corrected string data in a storage device and makes it available to the user as needed.

[0360] "Analysis means" refers to technology for analyzing the emotional state of meeting participants from their voices and facial expressions and extracting that information.

[0361] "Embedding means" refers to a process or device for reflecting emotional information obtained through analysis into string data.

[0362] "Highlighting means" refers to a device or function for dynamically highlighting and displaying string data that incorporates emotional information.

[0363] The system realizing this invention first includes a means for efficiently collecting audio and video data during a meeting. A terminal collects audio data through a microphone and transmits it to a server in real time. The server uses a generative AI model to instantly convert the audio data into text data. As a result, the content spoken during the meeting is transcribed sequentially.

[0364] Next, the server uses natural language processing technology to grammatically correct the generated string data. The RAG process is used as a correction method to appropriately standardize technical terms within the text data based on internal information sources. Through this process, the string data becomes accurate and consistent.

[0365] As an analytical method, the emotional state of participants is grasped in real time from audio and video data. This utilizes voice analysis technology and facial recognition technology to analyze emotions from the tone of the user's voice and facial expressions.

[0366] The analyzed emotional information is embedded into string data using a built-in mechanism. This data is dynamically highlighted on the device, visually representing differences based on emotion. For example, statements expressing particularly heightened emotions can be made easier to distinguish from other information by increasing the font size or changing the color.

[0367] As a concrete example, in a meeting about product displays held within a physical store, the opinions expressed by staff members regarding new display designs are recorded along with their emotions. If a staff member expresses a positive opinion, that part of the text is highlighted for easier viewing.

[0368] An example of a prompt message would be, "Analyze staff reactions to the new product display and add sentiment data to the meeting minutes." This allows participants and relevant departments to review the meeting content in detail and with emotional nuances after the meeting.

[0369] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0370] Step 1:

[0371] The device collects audio data using a microphone during a meeting. The input is the voices of meeting participants, and the output is a digital audio file. This enables real-time collection of audio data.

[0372] Step 2:

[0373] The server converts the received audio data into text data using a generation AI model. The input is a digital audio file, and the output is text data. The conversion process instantly transcribes the audio content into text, and the recording of the meeting proceeds.

[0374] Step 3:

[0375] The server uses natural language processing techniques to correct the grammar of string data. The input is the converted string data, and the output is the corrected string data. Grammar correction improves the accuracy and readability of the data.

[0376] Step 4:

[0377] The server utilizes the RAG process to standardize technical terms within string data based on internal sources. The input is grammatically corrected string data, and the output is the corrected, standardized string data. This process ensures consistency in terminology.

[0378] Step 5:

[0379] The server uses audio and video data to analyze participants' emotions. The input is the collected audio and video data, and the output is the analyzed emotional information. This extracts the emotional nuances of the meeting.

[0380] Step 6:

[0381] The server incorporates the analyzed sentiment information into the corrected string data. The input consists of the sentiment information and the corrected string data, and the output is unified string data with the sentiment information incorporated. This integrates the data in a way that includes emotional nuances.

[0382] Step 7:

[0383] The device dynamically highlights embedded string data. The input is integrated string data, and the output is visually highlighted screen data. Specifically, this includes actions such as increasing the font size or changing the color of statements that express particularly strong emotions.

[0384] Step 8:

[0385] Users view highlighted text data on the screen and save it as meeting minutes. The input is the highlighted screen data, and the output is the saved meeting minutes file. Meeting minutes, reflecting sentiment data, are later used for decision-making and reference in other meetings.

[0386] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0387] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0388] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0389] [Third Embodiment]

[0390] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0391] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0392] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0393] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0394] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0395] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0396] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0397] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0398] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0399] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0400] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0401] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0402] This invention is embodied in a system that automatically and accurately generates meeting minutes. This system is characterized by its ability to simultaneously collect audio and screen data and to correct for technical terms.

[0403] This system consists of several main components. First, during a meeting, the terminal collects audio through a microphone and captures video of slides and documents displayed on the projector or PC screen. This allows various data from the meeting to be recorded in real time.

[0404] Next, the server receives the audio data and uses a generative AI to convert it into text data. At this stage, natural language processing techniques that take into account the characteristics of the Japanese language are used, and general grammatical corrections are made.

[0405] The generated text is further refined by the RAG module on the server. Here, the latest technical terms stored in the database are referenced, and identified words are corrected. For example, the term "AI" is corrected and standardized to "artificial intelligence."

[0406] This corrected text is returned to the device and overlaid on the captured screen data. The display format is adjusted for user readability, and the text is displayed in conjunction with the slide content, allowing for immediate review during the meeting.

[0407] The final meeting minutes data is stored on the server in a secure manner. It is then uploaded to a shared folder or appropriate platform so that those with access rights can access the data. At this stage, users receive notifications and can proactively distribute the data to relevant parties after making individual revisions or additions as needed.

[0408] As a concrete example, consider a scenario where a new proposal for product development is made at a specific meeting. A terminal collects the details of the proposal in real time, and meeting minutes are generated using standardized technical terminology, ensuring consistency and ease of understanding for future reference. This contributes to increased work efficiency and improved communication.

[0409] The following describes the processing flow.

[0410] Step 1:

[0411] The device collects audio data via the microphone at the start of the meeting and activates its screen capture function to record slides and materials displayed during the meeting. This ensures that all audio and visual information is simultaneously acquired as digital data.

[0412] Step 2:

[0413] The server receives audio data transmitted in real time and converts it into text data using a generative AI model. Here, speech recognition technology is used to transcribe the content of the audio into text, and basic Japanese grammatical corrections are applied.

[0414] Step 3:

[0415] The server sends the generated text to the RAG module, where it corrects any mistranslations of technical terms. RAG refers to the latest terminology standards in its database and replaces technical terms in the text with appropriate words as needed. This process involves natural language processing techniques.

[0416] Step 4:

[0417] The device displays the corrected text data overlaid on the screen footage captured during the meeting. The font and layout are adjusted to improve text readability, allowing the user to instantly review the corrected text while viewing the screen.

[0418] Step 5:

[0419] The server securely stores the final meeting minutes data after formatting it. Data storage is performed using encryption technology, and the files are uploaded to a shared folder accessible to all relevant parties.

[0420] Step 6:

[0421] Users receive notifications and review the meeting minutes on their devices. After making any necessary corrections or adding annotations, they can distribute the minutes to other attendees and relevant departments via email or internal communication tools.

[0422] (Example 1)

[0423] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0424] It is difficult to efficiently collect audio and visual information during meetings and automatically generate accurate and standardized meeting minutes based on that information. Furthermore, insufficient real-time information integration and consistent standardization of technical terms result in a lack of information consistency, making it inconvenient to refer to later.

[0425] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0426] In this invention, the server includes a conversion device for converting audio information into text information, an adjustment device for standardizing technical terms in the text information, and a management device for recording and distributing the adjusted text information. This enables the efficient and accurate generation and visualization of meeting minutes in real time during a meeting.

[0427] A "perceptual device" is a device for effectively collecting audio and visual information in real time during a meeting.

[0428] A "conversion device" is a device that quickly and accurately converts collected audio information into text information.

[0429] A "regulation device" is a device used to standardize and correct specialized terminology in textual information into a standardized expression.

[0430] A "representation device" is a device for interactively representing adjusted textual information on visible information.

[0431] A "management device" is a device for recording, distributing, and properly managing adjusted textual information.

[0432] This invention provides a system that effectively utilizes audio and visual information from meetings to automatically generate accurate meeting minutes in real time. Specific embodiments of the invention are described below.

[0433] During a meeting, the device collects audio information using its microphone. It also simultaneously captures visual information, such as slides and documents displayed on a projector or display device, using its connected camera and screen capture function. To record this information in high resolution, the device is equipped with optimal audio filtering and high-performance image processing software.

[0434] The server receives audio information transmitted from the terminal and converts it into text using a generative AI model. During this process, it utilizes advanced Japanese natural language processing techniques to grammatically refine the information extracted from the audio. The AI ​​model used by the server includes speech recognition technology.

[0435] The generated text information is standardized for technical terms by a server-side adjustment device. By referring to an internal database and unifying technical terms into appropriate expressions, meeting minutes with a consistent format are formed.

[0436] The adjusted text information is resent to the device and displayed overlaid on the captured visual information. The font and layout of this display are adjusted for improved readability for the user. Users can review the meeting minutes generated in real time via their device and provide feedback as needed.

[0437] The finalized meeting minutes are securely stored on the server and uploaded to cloud storage or a shared folder with appropriate access permissions. This allows users to easily share the meeting minutes later as needed.

[0438] As a concrete example, consider a meeting where new product development proposals are discussed. During this meeting, a terminal captures the proposal content, and the server standardizes the terminology, ensuring consistent information that can be easily referenced by anyone at a later date.

[0439] An example of a prompt to input into a generative AI model is the instruction, "Transcribe the meeting audio, correct for technical jargon, and generate meeting minutes." This is an important prompt for reliably converting audio information into text information and automatically creating standardized meeting minutes.

[0440] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0441] Step 1:

[0442] The device uses its microphone to collect audio information from the meeting. In addition, the device uses its connected camera and screen capture function to capture visual information such as meeting slides and documents in high resolution. The input is the audio and video of the slides during the meeting, and the output is a file containing these recordings in digital format. Specifically, the device performs noise reduction and compresses the video to efficiently store the data.

[0443] Step 2:

[0444] The terminal formats the collected audio as digital data and sends it to the server via a secure communication protocol. The input is the formatted audio data, and the output is the audio file sent to the server. Specifically, the terminal encrypts the data and performs security checks to prevent unauthorized access.

[0445] Step 3:

[0446] The server inputs the received audio data into a generating AI model, which then converts it into text information using speech recognition technology. The input is audio data, and the output is grammatically corrected text data. The server performs a process of dividing the audio into segments, converting them into phonemes, and generating string data.

[0447] Step 4:

[0448] The server compares technical terms in the generated text information with an internal database and uses a reconciliation device to standardize them. The input is text data obtained from the generating AI model, and the output is text data with the technical terms reconciled. Specifically, it performs conversion processing to standardize terms such as "AI" to "artificial intelligence."

[0449] Step 5:

[0450] The terminal receives adjusted text information and displays it on the screen, overlaid with captured video data. The input is adjusted text data sent from the server, and the output is the combined data displayed in a highly legible font and layout. Specifically, the terminal adjusts the display position and changes the font size to make the information easy for the user to understand.

[0451] Step 6:

[0452] The server securely stores the finalized textual information as the meeting minutes and uploads it to a shared folder or the cloud with appropriate access permissions. The input is the final textual data, and the output is an accessible digital record. Specifically, the server backs up the data and sends access notifications to the necessary users.

[0453] (Application Example 1)

[0454] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0455] In business environments such as data centers, efficient meeting management and accurate recording of meeting content are required. However, the diversity of specialized terminology and the inability to share information immediately are obstacles, contributing to decreased work efficiency. Therefore, a system is needed that accurately records meeting content, standardizes specialized terminology, and enables immediate information sharing between multiple terminals.

[0456] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0457] In this invention, the server includes collection means for acquiring audio and video data, generation means for converting audio data into text information, and correction means for standardizing terminology in the text information. This enables the generation of accurate meeting minutes in real time during a meeting and allows for immediate information sharing among multiple terminals.

[0458] "Collection means" refers to a device or method for acquiring audio and video data during a meeting.

[0459] "Generation means" refers to a process or system for converting collected audio data into text information.

[0460] "Correction means" refers to a function or technique for modifying technical terms or specific terminology in text information in order to standardize them.

[0461] "Display means" refers to a function or device for visually presenting corrected text information on video data.

[0462] "Storage and sharing means" refers to a procedure or device for storing corrected text information and sharing that information with those who have access rights.

[0463] "Integrated distribution means" refers to a function or method for integrating information collected from multiple terminals and distributing that information appropriately.

[0464] To implement this invention, it is necessary to establish a system that collects audio and video data in a conference environment and processes it in real time. The system's terminals use microphones and cameras to acquire audio and materials displayed on a projector or screen during the conference. This data is transmitted to a server via a network.

[0465] The server converts the collected audio data into text using the Google Cloud Speech-to-Text API. Because the converted text may not accurately represent technical terms, the server corrects the terminology within the text using the Hugging Face Transformers library. This correction process involves retrieving relevant terminology information from an internal data store and performing standardization. The corrected text is then displayed on participants' devices via the Flask server. This displayed text is visible to meeting participants and serves as an aid to facilitate understanding.

[0466] Furthermore, the server securely stores the corrected text information in cloud storage and allows users to share it by setting access permissions. This feature makes it easy to refer to the information as meeting minutes even after the meeting has ended. Because actions based on the meeting content can be taken quickly, work efficiency can be improved.

[0467] As a concrete example, consider a system operations meeting held in a data center. In this system, participants discuss new network solutions they propose, and audio and video are collected in real time, with meeting minutes generated immediately. This allows for a rapid and standardized understanding of the technical details discussed in the meeting.

[0468] An example of a prompt for a generative AI model is as follows:

[0469] Prompt: Accurately recognize the technical terms discussed during the meeting, correct the generated text information, and produce accurate meeting minutes in Japanese.

[0470] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0471] Step 1:

[0472] The device's microphone and camera capture audio and video during the meeting. Audio is received as an analog signal, and video as pixel data. The device converts this data into a digital format, performs initial processing, and then transmits it to the server via the network.

[0473] Step 2:

[0474] The server sends the received audio data to the Google Cloud Speech-to-Text API, converting the audio signal into text information. Audio data is provided as input, and a speech recognition algorithm produces text output as a string. This converted text information is stored on the server.

[0475] Step 3:

[0476] The server uses Hugging Face Transformers to correct specialized terminology in text information. Given generated text information as input, it converts relevant specialized terms into appropriately standardized terms. Specifically, it refers to an internal data store, performs a conversion process to applicable terms, and obtains corrected text information.

[0477] Step 4:

[0478] The server uses Flask to send corrected text information to the client's terminal and displays it integrated with the video data. Corrected text information is provided as input, and the output is displayed on the video in real time via the user's visual interface.

[0479] Step 5:

[0480] The user reviews the information displayed through the terminal interface and modifies or adds annotations to the meeting minutes as needed. The displayed information is provided as input, and the final meeting minutes data is generated through user interaction.

[0481] Step 6:

[0482] The server securely stores the corrected text information in cloud storage and shares it among users as needed. The final meeting minutes data is provided as input, stored under appropriate access controls, and made available to other users via links and notifications.

[0483] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0484] This invention aims to improve the accuracy and convenience of information by incorporating an emotion engine that recognizes user emotions in real time into a system for automatically generating meeting minutes with high accuracy. This system achieves effective meeting minute generation by collecting audio and screen data, converting it to text, correcting technical terms, displaying the text, saving and sharing data, and providing emotion analysis capabilities.

[0485] At the start of the meeting, the device collects audio via the microphone and video for capturing meeting materials and slides. Simultaneously, the emotion engine analyzes the audio data to determine the emotional state of participants based on their tone and tempo. The determined emotional information is used to deepen understanding of the meeting's progress through the meeting minutes.

[0486] The server uses an AI model to generate audio data and convert it into text data in real time. This text is grammatically corrected using natural language processing technology, and further standardization of technical terms is performed using the RAG module. By combining the transcribed data with sentiment data, a more comprehensive meeting transcript is generated that includes the context and emotional nuances of the conversation.

[0487] The corrected text is displayed on the device screen. The sentiment engine's analysis results are reflected here, and the style and emphasis of the text display on the screen may be dynamically adjusted. For example, statements indicating strong emotions by the user may have a larger font size to draw attention.

[0488] The final meeting minutes are stored on a server and provided to users with access rights via the cloud system or on-premises network. This allows attendees and relevant departments to quickly and smoothly share and utilize the minutes after the meeting.

[0489] As a concrete example, if a participant shows strong excitement during a new product announcement at a meeting, that emotion will be recorded. This information will be communicated to other stakeholders through the meeting minutes and used as a reference for new product development. In this way, by combining it with an emotion engine, it is possible to capture not only textual information but also the overall atmosphere of the meeting.

[0490] The following describes the processing flow.

[0491] Step 1:

[0492] The device uses a microphone at the start of the meeting to collect participants' voices in real time and simultaneously captures the meeting materials screen. This is a crucial step in enabling the recording of the entire meeting's audio and visual information in digital format.

[0493] Step 2:

[0494] The server acquires audio data transmitted from the terminal and converts it into text data in real time using a generative AI model. This process transforms speech into text information, guaranteeing grammatical accuracy.

[0495] Step 3:

[0496] The server sends the generated text to the RAG module, where it standardizes the technical terms within the text. Specifically, it compares identified terms with the company's internal database and replaces them with appropriate terms, thereby improving the consistency and clarity of the discussion.

[0497] Step 4:

[0498] The device uses an emotion engine to analyze the user's emotions from the audio data for the generated text. The analysis results are reflected in the text's display style; for example, emotionally rich statements that should be emphasized are visually highlighted with different fonts or colors.

[0499] Step 5:

[0500] The device combines corrected text and sentiment data, overlaying it onto the screen capture video. This display is designed to allow users to intuitively understand the key points and emotional nuances of the meeting.

[0501] Step 6:

[0502] The server securely stores the final text data and uploads it to cloud storage or an internal network accessible to all stakeholders. This allows users to easily access, review, and share the generated meeting minutes after the meeting.

[0503] (Example 2)

[0504] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0505] In generating meeting minutes, conventional technologies could transcribe and display speech, but they could not generate information that took into account the emotions of the participants, making it difficult to grasp the atmosphere of the meeting and the changes in emotions. Furthermore, there were challenges in maintaining consistency in technical terms and the real-time nature of the generated text, which could reduce the accuracy and usefulness of the information.

[0506] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0507] In this invention, the server includes collection means for collecting audio and video information, generation means for converting audio information into text information, correction means for standardizing specialized terminology in the text information, and analysis and integration means for analyzing emotional information and integrating it into the text information. This makes it possible not only to record the contents of a meeting as text information, but also to generate a three-dimensional meeting record that reflects the emotions of the participants. Furthermore, it is possible to provide highly accurate information in real time while maintaining consistency in specialized terminology.

[0508] "Audio information" refers to data about conversations and voices collected using microphones or other audio input devices.

[0509] "Visual information" refers to visual data such as meeting materials and slides, collected using cameras or other visual input devices.

[0510] "Collection means" refers to technical means and devices for acquiring audio and video information, thereby incorporating the data into the system.

[0511] "Generation means" refers to technical means or devices for converting collected audio information into text information, and involves using a generation AI model to perform text conversion.

[0512] "Correction means" refers to technical means or processes for standardizing and correcting specialized terminology and grammar contained in generated textual information.

[0513] "Display means" refers to technical devices or methods for visually displaying corrected text information on video information, thereby providing users with easily understandable information.

[0514] "Means of saving and sharing" refers to technical means or systems for saving corrected text information and securely sharing it with users who have access rights.

[0515] "Analysis and integration means" refers to technical means or processes for analyzing emotional information from audio and video information and integrating it into textual information.

[0516] This invention is a system for automatically generating meeting minutes, aiming to create a three-dimensional meeting record that includes emotional information by combining audio and video information. This system mainly consists of terminals and a server.

[0517] The device uses a microphone to collect audio and a camera to capture video information such as meeting materials and slides. This allows for a comprehensive recording of the meeting's content and situation. The collected audio and video information is transmitted to the server in real time.

[0518] The server uses a generative AI model to instantly convert audio information into text. The converted text is then grammatically corrected using natural language processing techniques. Furthermore, the RAG module is used as a correction tool to standardize specialized terminology. This improves the consistency and accuracy of the meeting minutes.

[0519] For emotion analysis, the server uses an emotion engine based on audio information to determine the emotional state of participants from their tone and tempo. This analysis is integrated into the text information, visually reflecting the atmosphere and activity level of the meeting. For example, it's possible to visually emphasize key statements by increasing the font size.

[0520] The final meeting minutes are securely stored on the server and shared with authorized users via the cloud system. This allows for quick and smooth information sharing after the meeting.

[0521] An example of a prompt might be, "As a prompt to input into the generation AI model, please summarize the sentiment analysis of participants regarding specific statements made during the meeting, and how those sentiments will be reflected in the meeting minutes." In this way, the entire system working together enables the generation of highly accurate meeting minutes.

[0522] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0523] Step 1:

[0524] The device uses a microphone to collect audio information. When the meeting begins, the device converts the audio into digital data in real time and sends it to the server. At this stage, the input is raw audio, and the output is digital audio data. Similarly, when collecting video information, the device uses a camera to capture visual information of the meeting and sends the digital video data to the server.

[0525] Step 2:

[0526] The server converts speech information into text information using a generative AI model. The server analyzes the digital speech data received as input and converts the utterances into text. Noise reduction and speaker identification are also performed during this process, and the output is text data. Furthermore, natural language processing techniques are applied to the converted text to perform grammatical correction.

[0527] Step 3:

[0528] The server uses the RAG module as a correction tool to standardize specialized terminology in textual information. The input is text data that has undergone natural language processing, and the RAG module compares it with a business-specific database and replaces it with standard terminology. The output here is consistent, standardized text data.

[0529] Step 4:

[0530] The server uses an emotion engine to analyze participants' emotions from audio information and integrate it into text information. The input consists of digital audio and video data, and the server determines emotional states from voice tone, tempo, and facial expressions. This analysis results in output data with emotional information added to the text.

[0531] Step 5:

[0532] The terminal receives integrated text data sent from the server and displays it visually on the screen. The input is text data containing emotional information, which is displayed through dynamic visual adjustments. For example, emphasized statements may be displayed with a larger font size. The output is visually emphasized text.

[0533] Step 6:

[0534] The server securely stores the final meeting minutes data and shares it with authorized users via the cloud. The input is the finalized meeting minutes data, which is encrypted and stored, and a shareable link is generated. The output is the meeting minutes data accessible to users.

[0535] (Application Example 2)

[0536] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0537] Manually creating meeting minutes in conferences and business meetings is time-consuming and may leave out important information. Furthermore, accurately recording and sharing participants' emotions and the atmosphere of the meeting is difficult. This can lead to a lack of information in post-meeting communication and decision-making. Moreover, accurate information retention and visualization of participants' emotions are also required in staff meetings within stores and in interactions with customers.

[0538] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0539] In this invention, the server includes a collection means for collecting information during a meeting, an analysis means for analyzing the emotions of participants, and a highlighting means for dynamically highlighting the embedded string data. This makes it possible to create three-dimensional meeting minutes that reflect the emotions of participants, even in meetings held in physical stores, while incorporating emotional information into meeting minutes that are instantly transcribed and corrected from audio.

[0540] "Collection means" refers to a device or function for collecting information such as audio data and video data during a meeting.

[0541] "Generation means" refers to a technology or device for converting collected audio data into string data.

[0542] "Correction means" refers to a method or process for standardizing and grammatically correcting technical terms in generated string data.

[0543] "Display means" refers to a method or apparatus for visually presenting corrected string data on a screen or display.

[0544] "Storage and provision means" refers to a function or mechanism that stores corrected string data in a storage device and makes it available to the user as needed.

[0545] "Analysis means" refers to technology for analyzing the emotional state of meeting participants from their voices and facial expressions and extracting that information.

[0546] "Embedding means" refers to a process or device for reflecting emotional information obtained through analysis into string data.

[0547] "Highlighting means" refers to a device or function for dynamically highlighting and displaying string data that incorporates emotional information.

[0548] The system realizing this invention first includes a means for efficiently collecting audio and video data during a meeting. A terminal collects audio data through a microphone and transmits it to a server in real time. The server uses a generative AI model to instantly convert the audio data into text data. As a result, the content spoken during the meeting is transcribed sequentially.

[0549] Next, the server uses natural language processing technology to grammatically correct the generated string data. The RAG process is used as a correction method to appropriately standardize technical terms within the text data based on internal information sources. Through this process, the string data becomes accurate and consistent.

[0550] As an analytical method, the emotional state of participants is grasped in real time from audio and video data. This utilizes voice analysis technology and facial recognition technology to analyze emotions from the tone of the user's voice and facial expressions.

[0551] The analyzed emotional information is embedded into string data using a built-in mechanism. This data is dynamically highlighted on the device, visually representing differences based on emotion. For example, statements expressing particularly heightened emotions can be made easier to distinguish from other information by increasing the font size or changing the color.

[0552] As a concrete example, in a meeting about product displays held within a physical store, the opinions expressed by staff members regarding new display designs are recorded along with their emotions. If a staff member expresses a positive opinion, that part of the text is highlighted for easier viewing.

[0553] An example of a prompt message would be, "Analyze staff reactions to the new product display and add sentiment data to the meeting minutes." This allows participants and relevant departments to review the meeting content in detail and with emotional nuances after the meeting.

[0554] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0555] Step 1:

[0556] The device collects audio data using a microphone during a meeting. The input is the voices of meeting participants, and the output is a digital audio file. This enables real-time collection of audio data.

[0557] Step 2:

[0558] The server converts the received audio data into text data using a generation AI model. The input is a digital audio file, and the output is text data. The conversion process instantly transcribes the audio content into text, and the recording of the meeting proceeds.

[0559] Step 3:

[0560] The server uses natural language processing techniques to correct the grammar of string data. The input is the converted string data, and the output is the corrected string data. Grammar correction improves the accuracy and readability of the data.

[0561] Step 4:

[0562] The server utilizes the RAG process to standardize technical terms within string data based on internal sources. The input is grammatically corrected string data, and the output is the corrected, standardized string data. This process ensures consistency in terminology.

[0563] Step 5:

[0564] The server uses audio and video data to analyze participants' emotions. The input is the collected audio and video data, and the output is the analyzed emotional information. This extracts the emotional nuances of the meeting.

[0565] Step 6:

[0566] The server incorporates the analyzed sentiment information into the corrected string data. The input consists of the sentiment information and the corrected string data, and the output is unified string data with the sentiment information incorporated. This integrates the data in a way that includes emotional nuances.

[0567] Step 7:

[0568] The device dynamically highlights embedded string data. The input is integrated string data, and the output is visually highlighted screen data. Specifically, this includes actions such as increasing the font size or changing the color of statements that express particularly strong emotions.

[0569] Step 8:

[0570] Users view highlighted text data on the screen and save it as meeting minutes. The input is the highlighted screen data, and the output is the saved meeting minutes file. Meeting minutes, reflecting sentiment data, are later used for decision-making and reference in other meetings.

[0571] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0572] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0573] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0574] [Fourth Embodiment]

[0575] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0576] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0577] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0578] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0579] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0580] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0581] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0582] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0583] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0584] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0585] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0586] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0587] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0588] This invention is embodied in a system that automatically and accurately generates meeting minutes. This system is characterized by its ability to simultaneously collect audio and screen data and to correct for technical terms.

[0589] This system consists of several main components. First, during a meeting, the terminal collects audio through a microphone and captures video of slides and documents displayed on the projector or PC screen. This allows various data from the meeting to be recorded in real time.

[0590] Next, the server receives the audio data and uses a generative AI to convert it into text data. At this stage, natural language processing techniques that take into account the characteristics of the Japanese language are used, and general grammatical corrections are made.

[0591] The generated text is further refined by the RAG module on the server. Here, the latest technical terms stored in the database are referenced, and identified words are corrected. For example, the term "AI" is corrected and standardized to "artificial intelligence."

[0592] This corrected text is returned to the device and overlaid on the captured screen data. The display format is adjusted for user readability, and the text is displayed in conjunction with the slide content, allowing for immediate review during the meeting.

[0593] The final meeting minutes data is stored on the server in a secure manner. It is then uploaded to a shared folder or appropriate platform so that those with access rights can access the data. At this stage, users receive notifications and can proactively distribute the data to relevant parties after making individual revisions or additions as needed.

[0594] As a concrete example, consider a scenario where a new proposal for product development is made at a specific meeting. A terminal collects the details of the proposal in real time, and meeting minutes are generated using standardized technical terminology, ensuring consistency and ease of understanding for future reference. This contributes to increased work efficiency and improved communication.

[0595] The following describes the processing flow.

[0596] Step 1:

[0597] The device collects audio data via the microphone at the start of the meeting and activates its screen capture function to record slides and materials displayed during the meeting. This ensures that all audio and visual information is simultaneously acquired as digital data.

[0598] Step 2:

[0599] The server receives audio data transmitted in real time and converts it into text data using a generative AI model. Here, speech recognition technology is used to transcribe the content of the audio into text, and basic Japanese grammatical corrections are applied.

[0600] Step 3:

[0601] The server sends the generated text to the RAG module, where it corrects any mistranslations of technical terms. RAG refers to the latest terminology standards in its database and replaces technical terms in the text with appropriate words as needed. This process involves natural language processing techniques.

[0602] Step 4:

[0603] The device displays the corrected text data overlaid on the screen footage captured during the meeting. The font and layout are adjusted to improve text readability, allowing the user to instantly review the corrected text while viewing the screen.

[0604] Step 5:

[0605] The server securely stores the final meeting minutes data after formatting it. Data storage is performed using encryption technology, and the files are uploaded to a shared folder accessible to all relevant parties.

[0606] Step 6:

[0607] Users receive notifications and review the meeting minutes on their devices. After making any necessary corrections or adding annotations, they can distribute the minutes to other attendees and relevant departments via email or internal communication tools.

[0608] (Example 1)

[0609] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0610] It is difficult to efficiently collect audio and visual information during meetings and automatically generate accurate and standardized meeting minutes based on that information. Furthermore, insufficient real-time information integration and consistent standardization of technical terms result in a lack of information consistency, making it inconvenient to refer to later.

[0611] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0612] In this invention, the server includes a conversion device for converting audio information into text information, an adjustment device for standardizing technical terms in the text information, and a management device for recording and distributing the adjusted text information. This enables the efficient and accurate generation and visualization of meeting minutes in real time during a meeting.

[0613] A "perceptual device" is a device for effectively collecting audio and visual information in real time during a meeting.

[0614] A "conversion device" is a device that quickly and accurately converts collected audio information into text information.

[0615] A "regulation device" is a device used to standardize and correct specialized terminology in textual information into a standardized expression.

[0616] A "representation device" is a device for interactively representing adjusted textual information on visible information.

[0617] A "management device" is a device for recording, distributing, and properly managing adjusted textual information.

[0618] This invention provides a system that effectively utilizes audio and visual information from meetings to automatically generate accurate meeting minutes in real time. Specific embodiments of the invention are described below.

[0619] During a meeting, the device collects audio information using its microphone. It also simultaneously captures visual information, such as slides and documents displayed on a projector or display device, using its connected camera and screen capture function. To record this information in high resolution, the device is equipped with optimal audio filtering and high-performance image processing software.

[0620] The server receives audio information transmitted from the terminal and converts it into text using a generative AI model. During this process, it utilizes advanced Japanese natural language processing techniques to grammatically refine the information extracted from the audio. The AI ​​model used by the server includes speech recognition technology.

[0621] The generated text information is standardized for technical terms by a server-side adjustment device. By referring to an internal database and unifying technical terms into appropriate expressions, meeting minutes with a consistent format are formed.

[0622] The adjusted text information is resent to the device and displayed overlaid on the captured visual information. The font and layout of this display are adjusted for improved readability for the user. Users can review the meeting minutes generated in real time via their device and provide feedback as needed.

[0623] The finalized meeting minutes are securely stored on the server and uploaded to cloud storage or a shared folder with appropriate access permissions. This allows users to easily share the meeting minutes later as needed.

[0624] As a concrete example, consider a meeting where new product development proposals are discussed. During this meeting, a terminal captures the proposal content, and the server standardizes the terminology, saving it as consistent information that anyone can easily refer to later.

[0625] An example of a prompt to input into a generative AI model is the instruction, "Transcribe the meeting audio, correct for technical jargon, and generate meeting minutes." This is an important prompt for reliably converting audio information into text information and automatically creating standardized meeting minutes.

[0626] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0627] Step 1:

[0628] The device uses its microphone to collect audio information from the meeting. In addition, the device uses its connected camera and screen capture function to capture visual information such as meeting slides and documents in high resolution. The input is the audio and video of the slides during the meeting, and the output is a file containing these recordings in digital format. Specifically, the device performs noise reduction and compresses the video to efficiently store the data.

[0629] Step 2:

[0630] The terminal formats the collected audio as digital data and sends it to the server via a secure communication protocol. The input is the formatted audio data, and the output is the audio file sent to the server. Specifically, the terminal encrypts the data and performs security checks to prevent unauthorized access.

[0631] Step 3:

[0632] The server inputs the received audio data into a generating AI model, which then converts it into text information using speech recognition technology. The input is audio data, and the output is grammatically corrected text data. The server performs a process of dividing the audio into segments, converting them into phonemes, and generating string data.

[0633] Step 4:

[0634] The server compares technical terms in the generated text information with an internal database and uses a reconciliation device to standardize them. The input is text data obtained from the generating AI model, and the output is text data with the technical terms reconciled. Specifically, it performs conversion processing to standardize terms such as "AI" to "artificial intelligence."

[0635] Step 5:

[0636] The terminal receives adjusted text information and displays it on the screen, overlaid with captured video data. The input is adjusted text data sent from the server, and the output is the combined data displayed in a highly legible font and layout. Specifically, the terminal adjusts the display position and changes the font size to make the information easy for the user to understand.

[0637] Step 6:

[0638] The server securely stores the finalized textual information as the meeting minutes and uploads it to a shared folder or the cloud with appropriate access permissions. The input is the final textual data, and the output is an accessible digital record. Specifically, the server backs up the data and sends access notifications to the necessary users.

[0639] (Application Example 1)

[0640] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0641] In business environments such as data centers, efficient meeting management and accurate recording of meeting content are required. However, the diversity of specialized terminology and the inability to share information immediately are obstacles, contributing to decreased work efficiency. Therefore, a system is needed that accurately records meeting content, standardizes specialized terminology, and enables immediate information sharing between multiple terminals.

[0642] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0643] In this invention, the server includes collection means for acquiring audio and video data, generation means for converting audio data into text information, and correction means for standardizing terminology in the text information. This enables the generation of accurate meeting minutes in real time during a meeting and allows for immediate information sharing among multiple terminals.

[0644] "Collection means" refers to a device or method for acquiring audio and video data during a meeting.

[0645] "Generation means" refers to a process or system for converting collected audio data into text information.

[0646] "Correction means" refers to a function or technique for modifying technical terms or specific terminology in text information in order to standardize them.

[0647] "Display means" refers to a function or device for visually presenting corrected text information on video data.

[0648] "Storage and sharing means" refers to a procedure or device for storing corrected text information and sharing that information with those who have access rights.

[0649] "Integrated distribution means" refers to a function or method for integrating information collected from multiple terminals and distributing that information appropriately.

[0650] To implement this invention, it is necessary to establish a system that collects audio and video data in a conference environment and processes it in real time. The system's terminals use microphones and cameras to acquire audio and materials displayed on a projector or screen during the conference. This data is transmitted to a server via a network.

[0651] The server converts the collected audio data into text using the Google Cloud Speech-to-Text API. Because the converted text may not accurately represent technical terms, the server corrects the terminology within the text using the Hugging Face Transformers library. This correction process involves retrieving relevant terminology information from an internal data store and performing standardization. The corrected text is then displayed on participants' devices via the Flask server. This displayed text is visible to meeting participants and serves as an aid to facilitate understanding.

[0652] Furthermore, the server securely stores the corrected text information in cloud storage and allows users to share it by setting access permissions. This feature makes it easy to refer to the information as meeting minutes even after the meeting has ended. Because actions based on the meeting content can be taken quickly, work efficiency can be improved.

[0653] As a concrete example, consider a system operations meeting held in a data center. In this system, participants discuss new network solutions they propose, and audio and video are collected in real time, with meeting minutes generated immediately. This allows for a rapid and standardized understanding of the technical details discussed in the meeting.

[0654] An example of a prompt for a generative AI model is as follows:

[0655] Prompt: Accurately recognize the technical terms discussed during the meeting, correct the generated text information, and produce accurate meeting minutes in Japanese.

[0656] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0657] Step 1:

[0658] The device's microphone and camera capture audio and video during the meeting. Audio is received as an analog signal, and video as pixel data. The device converts this data into a digital format, performs initial processing, and then transmits it to the server via the network.

[0659] Step 2:

[0660] The server sends the received audio data to the Google Cloud Speech-to-Text API, converting the audio signal into text information. Audio data is provided as input, and a speech recognition algorithm produces text output as a string. This converted text information is stored on the server.

[0661] Step 3:

[0662] The server uses Hugging Face Transformers to correct specialized terminology in text information. Given generated text information as input, it converts relevant specialized terms into appropriately standardized terms. Specifically, it refers to an internal data store, performs a conversion process to applicable terms, and obtains corrected text information.

[0663] Step 4:

[0664] The server uses Flask to send corrected text information to the client's terminal and displays it integrated with the video data. Corrected text information is provided as input, and the output is displayed on the video in real time via the user's visual interface.

[0665] Step 5:

[0666] The user reviews the information displayed through the terminal interface and modifies or adds annotations to the meeting minutes as needed. The displayed information is provided as input, and the final meeting minutes data is generated through user interaction.

[0667] Step 6:

[0668] The server securely stores the corrected text information in cloud storage and shares it among users as needed. The final meeting minutes data is provided as input, stored under appropriate access controls, and made available to other users via links and notifications.

[0669] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0670] This invention aims to improve the accuracy and convenience of information by incorporating an emotion engine that recognizes user emotions in real time into a system for automatically generating meeting minutes with high accuracy. This system achieves effective meeting minute generation by collecting audio and screen data, converting it to text, correcting technical terms, displaying the text, saving and sharing data, and providing emotion analysis capabilities.

[0671] At the start of the meeting, the device collects audio via the microphone and video for capturing meeting materials and slides. Simultaneously, the emotion engine analyzes the audio data to determine the emotional state of participants based on their tone and tempo. The determined emotional information is used to deepen understanding of the meeting's progress through the meeting minutes.

[0672] The server uses an AI model to generate audio data and convert it into text data in real time. This text is grammatically corrected using natural language processing technology, and further standardization of technical terms is performed using the RAG module. By combining the transcribed data with sentiment data, a more comprehensive meeting transcript is generated that includes the context and emotional nuances of the conversation.

[0673] The corrected text is displayed on the device screen. The sentiment engine's analysis results are reflected here, and the style and emphasis of the text display on the screen may be dynamically adjusted. For example, statements indicating strong emotions by the user may have a larger font size to draw attention.

[0674] The final meeting minutes are stored on a server and provided to users with access rights via the cloud system or on-premises network. This allows attendees and relevant departments to quickly and smoothly share and utilize the minutes after the meeting.

[0675] As a concrete example, if a participant shows strong excitement during a new product announcement at a meeting, that emotion will be recorded. This information will be communicated to other stakeholders through the meeting minutes and used as a reference for new product development. In this way, by combining it with an emotion engine, it is possible to capture not only textual information but also the overall atmosphere of the meeting.

[0676] The following describes the processing flow.

[0677] Step 1:

[0678] The device uses a microphone at the start of the meeting to collect participants' voices in real time and simultaneously captures the meeting materials screen. This is a crucial step in enabling the recording of the entire meeting's audio and visual information in digital format.

[0679] Step 2:

[0680] The server acquires audio data transmitted from the terminal and converts it into text data in real time using a generative AI model. This process transforms speech into text information, guaranteeing grammatical accuracy.

[0681] Step 3:

[0682] The server sends the generated text to the RAG module, where it standardizes the technical terms within the text. Specifically, it compares identified terms with the company's internal database and replaces them with appropriate terms, thereby improving the consistency and clarity of the discussion.

[0683] Step 4:

[0684] The device uses an emotion engine to analyze the user's emotions from the audio data for the generated text. The analysis results are reflected in the text's display style; for example, emotionally rich statements that should be emphasized are visually highlighted with different fonts or colors.

[0685] Step 5:

[0686] The device combines corrected text and sentiment data, overlaying it onto the screen capture video. This display is designed to allow users to intuitively understand the key points and emotional nuances of the meeting.

[0687] Step 6:

[0688] The server securely stores the final text data and uploads it to cloud storage or an internal network accessible to all stakeholders. This allows users to easily access, review, and share the generated meeting minutes after the meeting.

[0689] (Example 2)

[0690] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0691] In generating meeting minutes, conventional technologies could transcribe and display speech, but they could not generate information that took into account the emotions of the participants, making it difficult to grasp the atmosphere of the meeting and the changes in emotions. Furthermore, there were challenges in maintaining consistency in technical terms and the real-time nature of the generated text, which could reduce the accuracy and usefulness of the information.

[0692] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0693] In this invention, the server includes collection means for collecting audio and video information, generation means for converting audio information into text information, correction means for standardizing specialized terminology in the text information, and analysis and integration means for analyzing emotional information and integrating it into the text information. This makes it possible not only to record the contents of a meeting as text information, but also to generate a three-dimensional meeting record that reflects the emotions of the participants. Furthermore, it is possible to provide highly accurate information in real time while maintaining consistency in specialized terminology.

[0694] "Audio information" refers to data about conversations and voices collected using microphones or other audio input devices.

[0695] "Visual information" refers to visual data such as meeting materials and slides, collected using cameras or other visual input devices.

[0696] "Collection means" refers to technical means and devices for acquiring audio and video information, thereby incorporating the data into the system.

[0697] "Generation means" refers to technical means or devices for converting collected audio information into text information, and involves using a generation AI model to perform text conversion.

[0698] "Correction means" refers to technical means or processes for standardizing and correcting specialized terminology and grammar contained in generated textual information.

[0699] "Display means" refers to technical devices or methods for visually displaying corrected text information on video information, thereby providing users with easily understandable information.

[0700] "Means of saving and sharing" refers to technical means or systems for saving corrected text information and securely sharing it with users who have access rights.

[0701] "Analysis and integration means" refers to technical means or processes for analyzing emotional information from audio and video information and integrating it into textual information.

[0702] This invention is a system for automatically generating meeting minutes, aiming to create a three-dimensional meeting record that includes emotional information by combining audio and video information. This system mainly consists of terminals and a server.

[0703] The device uses a microphone to collect audio and a camera to capture video information such as meeting materials and slides. This allows for a comprehensive recording of the meeting's content and situation. The collected audio and video information is transmitted to the server in real time.

[0704] The server uses a generative AI model to instantly convert audio information into text. The converted text is then grammatically corrected using natural language processing techniques. Furthermore, the RAG module is used as a correction tool to standardize specialized terminology. This improves the consistency and accuracy of the meeting minutes.

[0705] For emotion analysis, the server uses an emotion engine based on audio information to determine the emotional state of participants from their tone and tempo. This analysis is integrated into the text information, visually reflecting the atmosphere and activity level of the meeting. For example, it's possible to visually emphasize key statements by increasing the font size.

[0706] The final meeting minutes are securely stored on the server and shared with authorized users via the cloud system. This allows for quick and smooth information sharing after the meeting.

[0707] An example of a prompt might be, "As a prompt to input into the generation AI model, please summarize the sentiment analysis of participants regarding specific statements made during the meeting, and how those sentiments will be reflected in the meeting minutes." In this way, the entire system working together enables the generation of highly accurate meeting minutes.

[0708] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0709] Step 1:

[0710] The device uses a microphone to collect audio information. When the meeting begins, the device converts the audio into digital data in real time and sends it to the server. At this stage, the input is raw audio, and the output is digital audio data. Similarly, when collecting video information, the device uses a camera to capture visual information of the meeting and sends the digital video data to the server.

[0711] Step 2:

[0712] The server converts speech information into text information using a generative AI model. The server analyzes the digital speech data received as input and converts the utterances into text. Noise reduction and speaker identification are also performed during this process, and the output is text data. Furthermore, natural language processing techniques are applied to the converted text to perform grammatical correction.

[0713] Step 3:

[0714] The server uses the RAG module as a correction tool to standardize specialized terminology in textual information. The input is text data that has undergone natural language processing, and the RAG module compares it with a business-specific database and replaces it with standard terminology. The output here is consistent, standardized text data.

[0715] Step 4:

[0716] The server uses an emotion engine to analyze participants' emotions from audio information and integrate it into text information. The input consists of digital audio and video data, and the server determines emotional states from voice tone, tempo, and facial expressions. This analysis results in output data with emotional information added to the text.

[0717] Step 5:

[0718] The terminal receives integrated text data sent from the server and displays it visually on the screen. The input is text data containing emotional information, which is displayed through dynamic visual adjustments. For example, emphasized statements may be displayed with a larger font size. The output is visually emphasized text.

[0719] Step 6:

[0720] The server securely stores the final meeting minutes data and shares it with authorized users via the cloud. The input is the finalized meeting minutes data, which is encrypted and stored, and a shareable link is generated. The output is the meeting minutes data accessible to users.

[0721] (Application Example 2)

[0722] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0723] Manually creating meeting minutes in conferences and business meetings is time-consuming and may leave out important information. Furthermore, accurately recording and sharing participants' emotions and the atmosphere of the meeting is difficult. This can lead to a lack of information in post-meeting communication and decision-making. Moreover, accurate information retention and visualization of participants' emotions are also required in staff meetings within stores and in interactions with customers.

[0724] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0725] In this invention, the server includes a collection means for collecting information during a meeting, an analysis means for analyzing the emotions of participants, and a highlighting means for dynamically highlighting the embedded string data. This makes it possible to create three-dimensional meeting minutes that reflect the emotions of participants, even in meetings held in physical stores, while incorporating emotional information into meeting minutes that are instantly transcribed and corrected from audio.

[0726] "Collection means" refers to a device or function for collecting information such as audio data and video data during a meeting.

[0727] "Generation means" refers to a technology or device for converting collected audio data into string data.

[0728] "Correction means" refers to a method or process for standardizing and grammatically correcting technical terms in generated string data.

[0729] "Display means" refers to a method or apparatus for visually presenting corrected string data on a screen or display.

[0730] "Storage and provision means" refers to a function or mechanism that stores corrected string data in a storage device and makes it available to the user as needed.

[0731] "Analysis means" refers to technology for analyzing the emotional state of meeting participants from their voices and facial expressions and extracting that information.

[0732] "Embedding means" refers to a process or device for reflecting emotional information obtained through analysis into string data.

[0733] "Highlighting means" refers to a device or function for dynamically highlighting and displaying string data that incorporates emotional information.

[0734] The system realizing this invention first includes a means for efficiently collecting audio and video data during a meeting. A terminal collects audio data through a microphone and transmits it to a server in real time. The server uses a generative AI model to instantly convert the audio data into text data. As a result, the content spoken during the meeting is transcribed sequentially.

[0735] Next, the server uses natural language processing technology to grammatically correct the generated string data. The RAG process is used as a correction method to appropriately standardize technical terms within the text data based on internal information sources. Through this process, the string data becomes accurate and consistent.

[0736] As an analytical method, the emotional state of participants is grasped in real time from audio and video data. This utilizes voice analysis technology and facial recognition technology to analyze emotions from the tone of the user's voice and facial expressions.

[0737] The analyzed emotional information is embedded into string data using a built-in mechanism. This data is dynamically highlighted on the device, visually representing differences based on emotion. For example, statements expressing particularly heightened emotions can be made easier to distinguish from other information by increasing the font size or changing the color.

[0738] As a concrete example, in a meeting about product displays held within a physical store, the opinions expressed by staff members regarding new display designs are recorded along with their emotions. If a staff member expresses a positive opinion, that part of the text is highlighted for easier viewing.

[0739] An example of a prompt message would be, "Analyze staff reactions to the new product display and add sentiment data to the meeting minutes." This allows participants and relevant departments to review the meeting content in detail and with emotional nuances after the meeting.

[0740] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0741] Step 1:

[0742] The device collects audio data using a microphone during a meeting. The input is the voices of meeting participants, and the output is a digital audio file. This enables real-time collection of audio data.

[0743] Step 2:

[0744] The server converts the received audio data into text data using a generation AI model. The input is a digital audio file, and the output is text data. The conversion process instantly transcribes the audio content into text, and the recording of the meeting proceeds.

[0745] Step 3:

[0746] The server uses natural language processing techniques to correct the grammar of string data. The input is the converted string data, and the output is the corrected string data. Grammar correction improves the accuracy and readability of the data.

[0747] Step 4:

[0748] The server utilizes the RAG process to standardize technical terms within string data based on internal sources. The input is grammatically corrected string data, and the output is the corrected, standardized string data. This process ensures consistency in terminology.

[0749] Step 5:

[0750] The server uses audio and video data to analyze participants' emotions. The input is the collected audio and video data, and the output is the analyzed emotional information. This extracts the emotional nuances of the meeting.

[0751] Step 6:

[0752] The server incorporates the analyzed sentiment information into the corrected string data. The input consists of the sentiment information and the corrected string data, and the output is unified string data with the sentiment information incorporated. This integrates the data in a way that includes emotional nuances.

[0753] Step 7:

[0754] The device dynamically highlights embedded string data. The input is integrated string data, and the output is visually highlighted screen data. Specifically, this includes actions such as increasing the font size or changing the color of statements that express particularly strong emotions.

[0755] Step 8:

[0756] Users view highlighted text data on the screen and save it as meeting minutes. The input is the highlighted screen data, and the output is the saved meeting minutes file. Meeting minutes, reflecting sentiment data, are later used for decision-making and reference in other meetings.

[0757] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0758] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0759] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0760] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0761] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0762] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0763] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0764] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0765] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0766] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0767] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0768] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0769] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0770] 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.

[0771] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0772] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0773] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0774] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0775] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0776] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0777] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0778] The following is further disclosed regarding the embodiments described above.

[0779] (Claim 1)

[0780] A means for collecting audio and screen data during a meeting,

[0781] A generation means for converting the aforementioned audio data into text data,

[0782] Correction means for standardizing technical terms in the aforementioned text data,

[0783] A display means for visually displaying the corrected text data on the screen data,

[0784] A storage and sharing means for saving and sharing the corrected text data,

[0785] A system that includes this.

[0786] (Claim 2)

[0787] The system according to claim 1, wherein the generation means has the function of converting audio data into text in real time.

[0788] (Claim 3)

[0789] The system according to claim 1, wherein the correction means corrects technical terms in text data based on an in-house database using a search extension generation process.

[0790] "Example 1"

[0791] (Claim 1)

[0792] A perceptual device for collecting audio and visual information during a meeting,

[0793] A conversion device for converting the aforementioned audio information into text information,

[0794] An adjustment device for standardizing technical terms in the aforementioned textual information,

[0795] A display device for visually representing adjusted character information on the visible information,

[0796] A management device for recording and distributing the adjusted character information,

[0797] A system that includes this.

[0798] (Claim 2)

[0799] The system according to claim 1, wherein the conversion device has the function of instantly representing audio information into text.

[0800] (Claim 3)

[0801] The system according to claim 1, wherein the adjustment device adjusts specialized terms within the text information based on an internal information repository using a wide-range search generation process.

[0802] "Application Example 1"

[0803] (Claim 1)

[0804] A means of collecting audio and video data during a meeting,

[0805] A generation means for converting the aforementioned audio data into text information,

[0806] Correction means for standardizing terms in the aforementioned text information,

[0807] A display means for displaying corrected text information on the video data,

[0808] Storage and sharing means for storing and sharing the corrected text information,

[0809] An integrated distribution means for displaying and integrating information across multiple devices and distributing information based on access permissions,

[0810] A system that includes this.

[0811] (Claim 2)

[0812] The system according to claim 1, wherein the generation means has the function of instantly converting audio data into text.

[0813] (Claim 3)

[0814] The system according to claim 1, wherein the correction means corrects terms in text information based on an internal data store using a search extension generation process.

[0815] "Example 2 of combining an emotion engine"

[0816] (Claim 1)

[0817] means for collecting audio and video information,

[0818] A generation means for converting the aforementioned audio information into text information,

[0819] Correction means for standardizing specialized terminology in the aforementioned textual information,

[0820] A display means for visually displaying corrected character information on the video information,

[0821] A storage and sharing means for saving and sharing the corrected character information,

[0822] An analysis and integration means for analyzing emotional information and integrating it with the aforementioned textual information,

[0823] A system that includes this.

[0824] (Claim 2)

[0825] The system according to claim 1, wherein the generation means has the function of converting speech information into text in real time, and further performs grammatical correction using natural language processing technology.

[0826] (Claim 3)

[0827] The system according to claim 1, wherein the correction means corrects specialized terms in text information based on an internal database using a search extension generation process and standardizes them using a RAG module.

[0828] "Application example 2 when combining with an emotional engine"

[0829] (Claim 1)

[0830] means of gathering information during a meeting,

[0831] A generation means for converting the aforementioned audio data into string data,

[0832] Correction means for standardizing technical terms in the string data,

[0833] A display means for visually displaying the corrected string data on the aforementioned information,

[0834] A storage and provision means for storing and providing the corrected string data,

[0835] Analytical methods for analyzing participants' emotions,

[0836] An embedding means for incorporating the emotional information obtained by the analysis means into string data,

[0837] A highlighting means for dynamically highlighting embedded string data,

[0838] A system that includes this.

[0839] (Claim 2)

[0840] The system according to claim 1, wherein the generation means has the function of instantly converting audio data into text.

[0841] (Claim 3)

[0842] The system according to claim 1, wherein the correction means corrects technical terms in string data based on an internal information source using a RAG process. [Explanation of symbols]

[0843] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for collecting audio and screen data during a meeting, A generation means for converting the aforementioned audio data into text data, Correction means for standardizing technical terms in the aforementioned text data, A display means for visually displaying the corrected text data on the screen data, A storage and sharing means for saving and sharing the corrected text data, A system that includes this.

2. The system according to claim 1, wherein the generation means has the function of converting audio data into text in real time.

3. The system according to claim 1, wherein the correction means corrects technical terms in text data based on an in-house database using a search extension generation process.

Images