System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting audio of a presentation; means for collecting material and question and answer; means for preprocessing the collected material and question and answer; means for generating training content using a generative AI model based on the preprocessed material and question and answer; means for converting the generated training content to a video format; and means for uploading the generated video to a learning management system.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] At briefing sessions for new services and operations, detailed explanations are given by each department, but on-site staff are required to understand the flow of customer service. This makes it difficult to provide training immediately after the explanation. As a result, employees often lack understanding, leading to inefficiencies in customer service. The problem that this invention aims to solve is to solve these problems, deepen understanding efficiently, and enable prompt customer service. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides the following means: a system including means for collecting audio data from a briefing session, means for collecting material data and question-and-answer data, means for converting the collected audio data into text data, means for preprocessing the collected material data and question-and-answer data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into a video format, and means for uploading the generated video to a learning management system. Use of this system makes it possible to provide comprehensive training content immediately after the briefing session, thereby improving employee understanding and streamlining the training process.

[0006] "Audio data of the briefing session" refers to audio file data that records what was said at the briefing session.

[0007] "Material data" refers to data on presentation materials and documents used during the briefing session.

[0008] "Question and answer data" refers to data that records questions asked by participants and answers given by the lecturer during the seminar.

[0009] "Means of converting into text data" refers to technologies such as automatic speech recognition (ASR) systems that analyze voice data and convert it into text.

[0010] "Preprocessing means" refers to methods and techniques for preparing collected data in a form that is easy to analyze, and specifically includes noise removal and format standardization.

[0011] A "generative AI model" is an artificial intelligence model that analyzes given data and automatically generates new content for a specific purpose.

[0012] "Means for generating training content" refers to methods and techniques for creating training materials using generative AI models based on pre-processed data.

[0013] "Means for converting into video format" refers to the technology or method for converting the generated training content into a video file that can be displayed visually and audibly in an easy-to-understand manner.

[0014] A "learning management system" is an online platform that manages educational and training content and allows users to access and learn from it. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0016] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0019] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0020] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0021] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0022] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0026] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0027] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0028] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0033] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0034] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0035] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0036] As a specific embodiment of this invention, we will explain a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[0037] Program processing and natural language explanation

[0038] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded at the briefing and reads the data. This audio data is stored in memory in binary format.

[0039] 2. The server collects the material data and question and answer data. The material data corresponds to the presentation materials and documents used in the briefing, and the question and answer data is a record of the questions and answers exchanged between the participants and the lecturer during the briefing. These data are also read by specifying a file path and stored in memory in text format.

[0040] 3. The server converts the voice data into text data using automatic speech recognition (ASR) technology. The converted text data makes the contents of the briefing available as text information.

[0041] 4. The server preprocesses the document data and Q&A data. The format of the document data is standardized and noise and unnecessary information is removed. The Q&A data is processed so that it can be easily organized into a Q&A format.

[0042] 5. The server generates training content using a generative AI model based on the preprocessed data. Specifically, it combines text from the audio data, material data, and Q&A data into one, and creates text content for the training. The generative AI model analyzes this input data and generates content suitable for specific purposes.

[0043] 6. The server converts the generated training content into a video format, converting the generated text content into a visually and audibly appealing format and saving it as a video file.

[0044] 7. Finally, the server uploads the generated training video to the learning management system (LMS). This allows users to access and watch the training video from their devices. Users can log in to the LMS, select the desired video from the list of video content provided, and watch it.

[0045] Specific examples

[0046] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, as well as the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also organized. This preprocessed data is then integrated into one, and training content is created using a generative AI model. Finally, the generated content is converted into video format and uploaded to the LMS. Users can access the LMS and watch the training video to quickly understand how to use the new customer management system.

[0047] The processing flow will be explained below.

[0048] Step 1:

[0049] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0050] Step 2:

[0051] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[0052] Step 3:

[0053] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[0054] Step 4:

[0055] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers into an unambiguous format to make them easier to use.

[0056] Step 5:

[0057] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0058] Step 6:

[0059] The server converts the generated training content into a video format, and generates a video file based on the generated text content that is visually and audibly easy to understand, for example by adding narration corresponding to the text and incorporating slides and animations.

[0060] Step 7:

[0061] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, allowing users to view the training content at any time.

[0062] Step 8:

[0063] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0064] Example 1

[0065] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0066] In the conventional training content creation process, it takes a lot of time and effort to manually organize and integrate audio data, materials, and Q&A data from information sessions into training content. It is also difficult to consistently provide appropriate information, resulting in inconsistent training quality. The present invention aims to automate data processing and training content generation after information sessions, thereby providing efficient, high-quality training content.

[0067] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0068] In this invention, the server includes means for acquiring audio data of the briefing session, means for acquiring material data and Q&A data, means for converting the acquired audio data into text data, means for preprocessing the acquired material data and Q&A data, means for generating training information using an artificial intelligence model based on the preprocessed data, means for converting the generated training information into visual and audio media, and means for transmitting the generated visual and audio media to a learning management system. This automates data processing after the briefing session and enables the rapid provision of unified, high-quality training content.

[0069] "Audio data from a briefing session" refers to a digital audio file that records what was said at a briefing session, lecture, etc.

[0070] "Material data" refers to presentation materials and related document files used during the briefing session.

[0071] "Question and answer data" refers to records of questions and answers exchanged between participants and instructors during the information session.

[0072] The "means for converting into text data" is an automatic speech recognition technology for extracting voice data as text information.

[0073] "Preprocessing means" refers to procedures for arranging collected document data and Q&A data into a unified format and removing noise and unnecessary information.

[0074] An "artificial intelligence model" is a machine learning algorithm that generates appropriate output based on input data.

[0075] "Training information" is educational content that organizes the contents of the information session and provides them in a format suitable for learning purposes.

[0076] "Visual and audio media" refers to video files and other multimedia formats that contain text, image, and audio data.

[0077] A "learning management system" is an information management system that centrally manages training content provided by educational institutions and companies, and controls access and progress.

[0078] The "means for obtaining" is a method or device for reading data using a specified data file path.

[0079] "Transmitting means" refers to the data transfer method or communication means for uploading the generated training content to the learning management system.

[0080] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Specific embodiments are described below.

[0081] The server first collects the audio data of the briefing. During this process, the data is read into memory in binary format using a pre-specified audio file path (e.g., " / data / meeting_audio.wav"). Next, the server collects the presentation data (e.g., " / data / presentation.pdf") and the Q&A data (e.g., " / data / Q&A.txt"). These data are read from the specified paths and stored in memory in the form of a PDF file and a text file, respectively.

[0082] The server then converts the audio data into text using automatic speech recognition (ASR) technology, such as the Google Cloud Speech-to-Text API, which converts the audio data into text information and makes the contents of the briefing available in text format.

[0083] The server then preprocesses the document data and Q&A data. The document data is converted into a unified format and unnecessary information and noise are removed. The Q&A data is then processed to make it easier to organize into a Q&A format. For example, timestamps and unnecessary background information are removed from the document data, and the Q&A data is labeled into question and answer pairs.

[0084] The server generates training content using a generative AI model based on the preprocessed data. For example, using OpenAI's GPT-3, it inputs the following prompt: "Please create training content based on the following materials, audio text, and Q&A data." Based on this prompt, appropriate text content is generated.

[0085] The generated training content is then converted into visual and audio media, using Adobe Premiere Pro, for example, to add text as subtitles and composite the video with relevant images and graphics, with the final video file saved in MP4 format or similar.

[0086] Finally, the server uploads the generated training videos to the learning management system. During this process, the video files are uploaded directly from the server using the LMS's API, and metadata (lecture title, description, etc.) is also registered. Users can then log in to the LMS from their devices and watch the generated training videos. This automates data processing after the briefing session, making it possible to quickly provide uniform, high-quality training content.

[0087] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0088] Step 1:

[0089] The server collects audio data from the briefing session.

[0090] Input: The path to the information session recording file (e.g., " / data / meeting_audio.wav").

[0091] Specific operation: The server reads the audio file in binary format from the specified path and stores it in memory.

[0092] Output: Binary format of the audio data.

[0093] Step 2:

[0094] The server collects the material data and the question and answer data.

[0095] Input: File path of presentation materials (e.g. " / data / presentation.pdf") and file path of Q&A recording (e.g. " / data / Q&A.txt").

[0096] Specific operation: The server reads the PDF file and text file from each path and stores them in memory as document data and Q&A data.

[0097] Output: Document data (PDF format) and Q&A data (text format).

[0098] Step 3:

[0099] The server converts the voice data into text data.

[0100] Input: The binary format of the audio data collected in step 1.

[0101] Specific operation: The server uses automatic speech recognition technology such as the Google Cloud Speech-to-Text API to extract the voice data as text information.

[0102] Output: Text data converted from audio data.

[0103] Step 4:

[0104] The server pre-processes the document data and the question and answer data.

[0105] Input: Document data and Q&A data collected in Step 2.

[0106] Specific operation: The server converts the document data into a unified format, removes noise and unnecessary information, and processes the Q&A data so that it can be easily organized into a Q&A format.

[0107] Output: Preprocessed material data and Q&A data.

[0108] Step 5:

[0109] The server generates training content using a generative AI model based on the preprocessed data.

[0110] Input: Text data obtained in step 3, preprocessed document data and Q&A data in step 4.

[0111] Specific operation: The server uses a generative AI model (e.g., OpenAI's GPT-3) and inputs a prompt such as: "Create training content based on the following materials, audio text, and question and answer data."

[0112] Output: Generated training text content.

[0113] Step 6:

[0114] The server converts the generated training content into a video format.

[0115] Input: The training text content generated in step 5.

[0116] What it does: The server converts the generated text content into visual and auditory media and saves it as a video file. Using software such as Adobe Premiere Pro, the text is added as subtitles and composited into a video with associated images and graphics.

[0117] Output: Training video file (e.g. MP4 format).

[0118] Step 7:

[0119] The server uploads the generated training videos to a learning management system.

[0120] Input: The training video file created in step 6.

[0121] Specific operation: The server uses the API of the LMS platform (e.g., Moodle) to upload the video file and also registers metadata (lecture title, description, etc.).

[0122] Output: Training videos uploaded to a learning management system.

[0123] (Application example 1)

[0124] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0125] As new equipment and technologies are frequently introduced in factories, it is necessary to provide quick and efficient training for workers on operation and maintenance. However, current training methods require instructors to coordinate schedules, prepare materials, and provide face-to-face explanations, which are costly and time-consuming. This has led to a demand for automated generation of training content and its effective use.

[0126] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0127] In this invention, the server includes a means for collecting audio data of the briefing session, a means for collecting material data and Q&A data, and a means for converting the collected audio data into text data, thereby enabling the server to generate training content using a generative AI model based on the preprocessed data, convert the generated training content into video format, upload the generated video to a learning management system, and provide the training content to visual and audio devices.

[0128] "Audio data from a briefing session" refers to data that digitally records the verbal information spoken during the briefing session.

[0129] "Material data" refers to the data of presentation materials and documents used at the briefing.

[0130] "Question and answer data" refers to recorded data of questions and answers exchanged between participants and the lecturer during the information session.

[0131] "Means for converting into text data" refers to the process of converting voice data into text information using natural language processing technology.

[0132] "Preprocessing" refers to operations such as standardizing the format of collected data and removing noise so that the data can be easily analyzed and integrated.

[0133] A "generative AI model" refers to an algorithm or program that uses machine learning and artificial intelligence techniques to generate new content based on input data.

[0134] "Training Content" refers collectively to information and materials provided for learning or training purposes.

[0135] "Means for converting to video format" refers to the technology and tools that convert generated content such as text and charts into video files.

[0136] A "Learning Management System (LMS)" is a system that manages and distributes electronic learning materials, and is a platform that allows users to manage their own progress while learning.

[0137] "Visual and auditory devices" refers to devices that allow users to view and listen to learning content, including smart glasses and head-mounted displays.

[0138] This invention relates to a system that automatically generates training content for learning about the operation and maintenance techniques of robots used in factories. This system collects audio data, document data, and question and answer data from briefing sessions, preprocesses them, and then automatically generates training content using a generative AI model.

[0139] Hardware and software used

[0140] The main hardware and software components for implementing this system are as follows:

[0141] 1. Smart glasses (e.g., Google Glass)

[0142] 2. Head-Mounted Display (HMD)

[0143] 3. Automatic Speech Recognition (ASR) technology (e.g., Google Speech-to-Text API)

[0144] 4. Generative AI models (e.g., OpenAI GPT-4)

[0145] 5. Video editing tools (e.g. Adobe Premiere Pro, FFmpeg)

[0146] 6. Learning Management System (LMS)

[0147] Overview of system processing

[0148] The server processes the data in the following steps and ultimately generates the training content.

[0149] 1. Data Collection:

[0150] The server collects audio data from the briefings held at the factory, along with the presentation materials and Q&A sessions. The collected audio data is recorded by microphones installed in the smart glasses or robots.

[0151] 2. Text conversion:

[0152] The audio data is converted into text using automatic speech recognition (ASR) technology, which makes the contents of the briefing available as text information.

[0153] 3. Data Preprocessing:

[0154] The document data and Q&A data are preprocessed. The format of the document data is standardized and unnecessary information is deleted. The Q&A data is also organized in Q&A format.

[0155] 4. Content Generation:

[0156] Based on the preprocessed data, a generative AI model is used to generate training content. Specifically, the generative AI model combines text from the audio data, material data, and question and answer data to create text content for training.

[0157] 5. Video conversion and uploading:

[0158] The generated text content is converted into a video format, which is then converted into a visually and audibly appealing format using a video editing tool and saved as a video file.

[0159] Finally, the generated training videos are uploaded to a learning management system (LMS), where users can access the LMS through smart glasses or HMDs and watch the videos to learn operation procedures and maintenance techniques.

[0160] Specific examples

[0161] For example, consider a training session on the introduction of a new welding robot. A worker wearing smart glasses records the session and simultaneously collects the presentation materials and Q&A. ASR converts the speech to text and organizes the material data into a unified format. The training content is then generated by inputting the following prompt sentences into a generative AI model:

[0162] Example prompt sentence:

[0163] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[0164] 1. Audio text: <Audio text data>

[0165] 2. Presentation materials: <Data in a unified format>

[0166] 3. Q&A: <Organized Q&A data>

[0167] The generated training text is then edited using a video editing tool and finally uploaded to the LMS, allowing workers to access the training video on their HMD and progress through their studies.

[0168] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0169] Step 1:

[0170] The server collects audio data from the briefing session. Audio is recorded through microphones installed in the smart glasses or robot. The collected audio data is stored in memory in binary format. The input is audio signal data, and the output is binary audio data.

[0171] Step 2:

[0172] The server collects material data and question and answer data. Material data is presentation materials and documents, and question and answer data is records of questions and answers. These data are transferred to the server via file paths and stored in memory in text format. The input is material and question and answer files, and the output is text format data.

[0173] Step 3:

[0174] The server converts the collected voice data into text data using automatic speech recognition (ASR) technology. It uses an automatic speech recognition system (e.g., Google Speech-to-Text API) to analyze the voice data and convert it into text data. In this step, the voice signal is converted into text information, and the specific content becomes available as text information. The input is binary voice data, and the output is text data.

[0175] Step 4:

[0176] The server preprocesses the collected document data and Q&A data. The document data is unified in format and unnecessary information is removed before being converted into text format. The Q&A data is similarly formatted so that questions and answers are clearly distinguished. The input is the raw document data and Q&A data, and the output is preprocessed text data.

[0177] Step 5:

[0178] The server generates training content using a generative AI model based on the preprocessed data. The following prompt sentence is input to the generative AI model (e.g., OpenAI GPT-4) to generate text content for training:

[0179] Example prompt sentence:

[0180] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[0181] 1. Audio text: <Audio text data>

[0182] 2. Presentation materials: <Data in a unified format>

[0183] 3. Q&A: <Organized Q&A data>

[0184] The input is preprocessed speech text, material data, and question and answer data, and the output is text content for training.

[0185] Step 6:

[0186] The server converts the generated training content into a video format. Using a video editing tool (e.g. Adobe Premiere Pro, FFmpeg), it converts the content into a video file with visual and auditory appeal. The input is text content, and the output is a video file.

[0187] Step 7:

[0188] The server uploads the generated training videos to a learning management system (LMS). At this time, the video files are registered in the system so that users can access the LMS and watch the videos. Users can access the LMS using smart glasses or a head-mounted display, watch the training videos, and progress with their studies. The input is the video file, and the output is the training video uploaded to the LMS.

[0189] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0190] As a specific embodiment of this invention, we will explain a system that combines an emotion engine with a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Furthermore, this system uses an emotion engine that recognizes user emotions and utilizes emotion data to generate and update training content.

[0191] Program processing and natural language explanation

[0192] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0193] 2. The server collects the material data and Q&A data. The material data is the presentation materials and documents used in the briefing, and the Q&A data is the content of the questions and answers asked during the briefing. These data are also loaded by specifying the file path. The material data and Q&A data are stored in memory in text format.

[0194] 3. The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the speech content is converted into text information. This text data makes the content of the briefing available in written form.

[0195] 4. The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use.

[0196] 5. The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are combined and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0197] 6. The server converts the generated training content into a video format, generating a visually and audibly easy-to-understand video file based on the generated text content, for example by adding narration corresponding to the text and incorporating slides and animations.

[0198] 7. The server uploads the generated videos to a learning management system (LMS), which is an online platform accessed by employees. By uploading the generated training videos to this system, users can view the training content at any time.

[0199] 8. The user uses the device to access the learning management system (LMS). The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0200] 9. The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice through a camera and microphone, and monitors changes in emotions in real time.

[0201] 10. The server evaluates the training content based on the collected emotional data. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video speed. It also uses the emotional data as feedback to improve the quality of future training content.

[0202] Specific examples

[0203] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, along with the presentation materials and Q&A sessions. The collected audio data is converted into text using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A sessions are also organized. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted accordingly based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed. This allows users to gain a deeper understanding of the new service and its operational flow, resulting in faster and more effective customer support.

[0204] The processing flow will be explained below.

[0205] Step 1:

[0206] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0207] Step 2:

[0208] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[0209] Step 3:

[0210] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[0211] Step 4:

[0212] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and removing unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use. For example, it standardizes the text to lowercase and removes unnecessary spaces and noise.

[0213] Step 5:

[0214] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0215] Step 6:

[0216] The server converts the generated training content into a video format. Based on the generated text content, it generates a video file that is visually and audibly easy to understand. Specifically, it generates narration corresponding to the text and combines it with slides and animations to create a video file.

[0217] Step 7:

[0218] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, and uploads the generated training videos to this system, allowing users to view the training content at any time.

[0219] Step 8:

[0220] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0221] Step 9:

[0222] The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice via a camera and microphone, and monitors changes in emotions in real time.

[0223] Step 10:

[0224] The server evaluates the training content based on the collected emotional data and makes necessary adjustments. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video playback speed. The emotional data can also be used as feedback to improve the quality of future training content.

[0225] Example 2

[0226] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0227] Conventional training content generation systems have difficulty efficiently collecting and processing the contents of information sessions and automatically generating training content. It is also difficult to generate and update content that reflects changes in users' understanding and emotions in real time. This results in a large amount of manual work being required to provide effective training, resulting in reduced productivity.

[0228] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0229] In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and question-and-answer data, means for converting the collected audio data into text data, means for preprocessing the collected material data and question-and-answer data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for recognizing user emotions, and means for evaluating and updating the training content based on the recognized emotion data. This makes it possible to efficiently collect the contents of the briefing session and dynamically generate and update training content according to the user's level of understanding.

[0230] "Audio Data" refers to the content of the audio files recorded during the briefing.

[0231] "Material data" refers to presentation materials and documents used at the briefing.

[0232] "Question and answer data" refers to the content of questions and answers asked during the briefing session.

[0233] "Text data" refers to collected voice data converted into text information using an automatic speech recognition system.

[0234] "Preprocessing" refers to the process of standardizing the format of document data, deleting unnecessary parts, and organizing question and answer data.

[0235] "Generative AI model" refers to artificial intelligence techniques that automatically generate new training content using collected and pre-processed data.

[0236] "Video format" refers to the conversion of generated training text content into a visually and audibly appealing video file.

[0237] "Learning Management System (LMS)" means an online platform through which employees access and learn training content via the internet.

[0238] "Emotion data" refers to data that records changes in emotions recognized by the emotion engine from the user's facial expressions and tone of voice.

[0239] "Evaluate and update" refers to the process of judging the quality of training content based on sentiment data and improving and adjusting the content as needed.

[0240] MODE FOR CARRYING OUT THE INVENTION

[0241] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). It also has the ability to recognize user emotions and reflect them in the generation and updating of training content. This system collects audio data, materials, and Q&A data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[0242] Audio data collection and conversion

[0243] The server collects the audio data of the briefing and stores it in memory in binary format. It then converts the collected audio data into text data using an automatic speech recognition (ASR) system, such as the Google Cloud Speech-to-Text API, making the audio content available as text information.

[0244] Collection and preprocessing of document data and Q&A data

[0245] The server loads the presentation materials, documents, and Q&A data used in the briefing into memory in text format. Preprocessing of the material data includes standardizing the format and deleting unnecessary parts, and for the Q&A data, it organizes the questions and answers and converts them into a more user-friendly format.

[0246] Training content generation and video conversion

[0247] The server integrates the pre-processed data and feeds it to a generative AI model to automatically generate training content. For example, the following prompt sentences can be used to input the generative AI model:

[0248] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[0249] The generated training content is converted into a video file that appeals to both the visual and auditory senses, incorporating narration, slides, animations, and other elements to create a visually easy-to-understand format.

[0250] Upload to a learning management system

[0251] The generated training content is converted into video format and uploaded to a learning management system (LMS), where users can access and watch the training videos.

[0252] User emotion recognition and content update

[0253] The server uses an emotion engine to recognize the user's emotions while watching the training video. It monitors the user's facial expressions and tone of voice through a camera and microphone to collect emotion data in real time. Based on this emotion data, the quality of the training content can be evaluated and updated as necessary. For example, if the user is confused, additional explanations can be generated or the video playback speed can be adjusted.

[0254] Specific examples

[0255] For example, when an introduction briefing for a new customer management system is held, the server collects the audio data from the briefing, as well as the presentation materials and Q&A sessions used. The audio data is converted into text data by the ASR system, the material data is organized into a unified format, and the Q&A data is also organized. Training content is then created using a generative AI model based on the preprocessed data. The generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch this training video. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. This allows users to gain a deeper understanding of the new system and enables faster and more effective customer service.

[0256] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0257] Step 1:

[0258] Audio data collection

[0259] The server receives the file path of the audio data of the briefing as input, opens the file, and reads the data into memory in binary format. This process saves the collected audio data in memory. Specifically, it opens the specified audio file and reads the contents in binary format.

[0260] Step 2:

[0261] Collection of document data and Q&A data

[0262] The server receives the file paths for the presentation data and Q&A data as input, opens each file, and reads the data into memory in text format. This process saves the presentation data and Q&A data in memory. Specifically, the presentation materials and Q&A content are read as text data.

[0263] Step 3:

[0264] Converting audio data to text

[0265] The server receives the collected voice data as input and converts it into text data using an automatic speech recognition (ASR) system. This process makes the voice content available as text information. Specifically, the voice data is input into the ASR system, and text information is obtained as output.

[0266] Step 4:

[0267] Preprocessing of document data and Q&A data

[0268] The server receives the collected document data and Q&A data as input, standardizes the data format, deletes unnecessary parts, and organizes the questions and answers. This process results in preprocessed data. Specifically, unnecessary parts of the text are deleted and the format is adjusted.

[0269] Step 5:

[0270] Training content generation

[0271] The server integrates the preprocessed speech-to-text data, the materials data, and the Q&A data, and inputs them into the generative AI model along with prompts. This process generates training content. Specifically, the following prompts are used:

[0272] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[0273] Step 6:

[0274] Content video conversion

[0275] The server receives the generated training content as input and converts it into a visually and audibly appealing video file. This process results in a training video containing narration, slides, and animations. Specifically, the server creates slides with narration and animations based on the generated text content and converts them into a video format.

[0276] Step 7:

[0277] Upload to LMS

[0278] The server receives the generated video file as input and uploads it to the learning management system (LMS). This process makes it accessible to users. Specifically, the video file is sent to the server using the LMS's upload function.

[0279] Step 8:

[0280] User LMS Access

[0281] Users access the LMS using their devices and select and watch the training videos provided. This process allows users to learn the training content. Specifically, users log in to the LMS, select the desired video from the video list, and press the play button.

[0282] Step 9:

[0283] emotion recognition

[0284] The server monitors the user's facial expressions and tone of voice while watching the training video through a camera and microphone, and collects emotional data using an emotion engine. This process yields the user's emotional data. Specifically, it analyzes changes in facial expressions and voice while watching in real time and records changes in emotions.

[0285] Step 10:

[0286] Evaluating and updating training content

[0287] The server receives the collected emotional data as input, evaluates the training content, and updates it as necessary to improve the quality of the training. Specifically, it generates additional explanations or adjusts the video playback speed depending on the user's emotional state.

[0288] (Application example 2)

[0289] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0290] Conventional training content generation systems generate training content based on recordings of information sessions and materials, but because they do not take into account the level of understanding and emotions of individual end users, the quality of the training content is uniform and may not meet the needs of end users.In addition, the generated content is fixed and difficult to adjust in real time, which creates the problem of not being able to maximize the learning effect of end users.

[0291] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and Q&A data, means for converting the collected audio data into text data, means for preprocessing the collected material data and Q&A data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for collecting end-user emotional data using sensors mounted on the robot, and means for evaluating and adjusting the training content based on the collected emotional data. This allows the training content to be adjusted in real time taking into account the end-user's level of understanding and emotions, enabling advanced training tailored to individual needs.

[0292] "Audio data from the briefing session" refers to an audio file that records what was said during the briefing session.

[0293] "Material data" refers to presentation materials and document files used in the briefing session.

[0294] "Question and answer data" is text data of questions asked during the briefing session and the answers to those questions.

[0295] "Means for converting to text data" refers to a system or algorithm that analyzes audio data and makes it usable as text information.

[0296] "Preprocessing means" refers to means for standardizing the formats of collected document data and question and answer data and deleting unnecessary parts.

[0297] A "generative AI model" is a type of artificial intelligence that automatically generates new training content based on given data.

[0298] "Training Content" means learning materials for end users generated by a generative AI model.

[0299] "Means of converting to video format" refers to the process of converting the generated training content into visual and audio materials.

[0300] A "learning management system" is an online platform where training content can be uploaded and accessed by end users for learning.

[0301] "Sensors mounted on robots" are devices such as cameras and microphones attached to robots, and are used to collect emotional data from end users.

[0302] "Emotional data" refers to psychological reaction information obtained from the end user's facial expressions and tone of voice.

[0303] The "means for evaluating and adjusting" refers to a means for appropriately correcting or providing additional explanations to pre-generated training content based on the collected emotional data.

[0304] MODE FOR CARRYING OUT THE INVENTION

[0305] As a specific embodiment for carrying out the present invention, a system will be described that automatically generates training content after an information session and adjusts the content based on emotional data of end users.

[0306] Overall system configuration

[0307] This system operates in the following manner.

[0308] Audio data collection

[0309] The server collects audio data from the briefing session. Specifically, it specifies the path to the audio file recorded during the briefing session, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0310] Collection of document data and question and answer data

[0311] The server collects presentation materials, documents, and question and answer data used in the briefing. These data are also loaded by specifying the file path. The material data and question and answer data are stored in memory in text format.

[0312] Automatic Speech Recognition (ASR) System

[0313] The server uses an automatic speech recognition (ASR) system to convert the collected voice data into text data using the Google Cloud Speech-to-Text API, which makes the content of the briefing available as a transcript.

[0314] Preprocessing of document data and Q&A data

[0315] The server standardizes the format of the document data and deletes unnecessary parts, and organizes the questions and answers in the question and answer data to make it easier to use.

[0316] Generating training content using generative AI models

[0317] Based on the preprocessed data, the server generates training content using a generative AI model such as GPT-4. The preprocessed audio text, document data, and Q&A data are combined and fed into the generative AI model as input. The generative AI model then analyzes this data and automatically generates text content for the training.

[0318] Convert to video format

[0319] Based on the generated text content, the server generates a visually and audibly comprehensible video file, including adding corresponding narration to the text and incorporating slides and animations. This process uses OpenCV and FFmpeg.

[0320] Upload to Learning Management System (LMS)

[0321] The server uploads the generated videos to a learning management system, which is an online platform accessed by employees, allowing users to view the training content at any time.

[0322] Collecting Emotional Data

[0323] Sensors (camera and microphone) installed on the robot are used to collect emotional data while the end user is watching the training video. The facial expression data is analyzed using the Microsoft Azure Face API, and emotions are recognized from the tone of voice.

[0324] Evaluate and adjust training content

[0325] The server evaluates the training content based on the collected emotional data and, if necessary, includes sections in the video for additional explanation or adjusts the playback speed.

[0326] Specific examples

[0327] For example, when a briefing on new maintenance procedures is held, the server collects the audio data of the briefing, along with the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also compiled. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed.

[0328] Prompt Sentence Examples

[0329] New maintenance procedure briefing text:

[0330] 1. Engine disassembly procedure

[0331] 2. Parts inspection items

[0332] 3. Lubrication method

[0333] 4. Vibration measurement method

[0334] Q&A:

[0335] Q: What type of lubricant?

[0336] A: We will use ABC lubricant from XYZ company.

[0337] Use this information to generate training content that includes specific operating procedures and precautions, from engine disassembly procedures to lubrication methods.

[0338] The system makes it possible to adjust training content in real time, taking into account the end user's level of understanding and emotions, improving maintenance efficiency within the factory and ensuring that new procedures are understood quickly and reliably.

[0339] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0340] Step 1:

[0341] The server collects the audio data of the briefing. Specifically, it specifies the path of the audio file recorded during the briefing, opens the file, and reads the data. The input is the recorded audio file, and the output is binary audio data. The audio data is stored in buffer memory.

[0342] Step 2:

[0343] The server collects material data and question and answer data. The material data is the presentation materials and document format files used in the briefing, and the question and answer data is text data of questions and answers asked during the briefing. The input is the path of the material file and the question and answer file, and the output is stored in memory in text format.

[0344] Step 3:

[0345] The server converts the voice data into text data. An automatic speech recognition (ASR) system, in this case the Google Cloud Speech-to-Text API, is used to analyze the voice data and convert it into text information. The input is binary voice data, and the output is text data.

[0346] Step 4:

[0347] The server preprocesses the document data and Q&A data. Specifically, it standardizes the format of the document data, deletes unnecessary parts, and organizes the questions and answers in the Q&A data. The input is text-formatted document data and Q&A data, and the output is the preprocessed unified data.

[0348] Step 5:

[0349] Based on the preprocessed data, the server generates training content using a generative AI model. Specifically, the preprocessed audio text, material data, and Q&A data are integrated into one and supplied to the generative AI model (GPT-4). The input is the integrated preprocessed data, and the output is the generated training text content.

[0350] Step 6:

[0351] The server converts the generated training content into video format. Using OpenCV and FFmpeg, it adds narration corresponding to the text and incorporates slides and animations to generate the video. The input is the generated text content, and the output is a video file that is visually and audibly easy to understand.

[0352] Step 7:

[0353] The server uploads the generated videos to a learning management system (LMS). The LMS is an online platform accessed by employees, where the generated training videos are uploaded so that users can view the training content at any time. The input is the generated video file, and the output is the training content uploaded to the LMS.

[0354] Step 8:

[0355] A user uses a device to access a learning management system (LMS) and watch a provided training video. The user logs in to the LMS, selects the target training video, and plays it. The input is the login information for the LMS and the target training video data, and the output is the training video that is played.

[0356] Step 9:

[0357] The robot collects emotional data from the user using sensors (camera and microphone) installed on the device. It uses the Microsoft Azure Face API to analyze the facial expression data and recognize emotions from the tone of voice. The input is video and audio data, and the output is analyzed emotional data.

[0358] Step 10:

[0359] The server evaluates and adjusts the training content based on the collected emotional data, including adding sections to the video for additional explanations or adjusting the playback speed as needed. The input is emotional data, and the output is the adjusted training content.

[0360] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0361] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0362] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0363] [Second embodiment]

[0364] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0365] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0366] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0367] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0368] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0369] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0370] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0371] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0372] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0373] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0374] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0375] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0376] As a specific embodiment of this invention, we will explain a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[0377] Program processing and natural language explanation

[0378] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded at the briefing and reads the data. This audio data is stored in memory in binary format.

[0379] 2. The server collects the material data and question and answer data. The material data corresponds to the presentation materials and documents used in the briefing, and the question and answer data is a record of the questions and answers exchanged between the participants and the lecturer during the briefing. These data are also read by specifying a file path and stored in memory in text format.

[0380] 3. The server converts the voice data into text data using automatic speech recognition (ASR) technology. The converted text data makes the contents of the briefing available as text information.

[0381] 4. The server preprocesses the document data and Q&A data. The format of the document data is standardized and noise and unnecessary information is removed. The Q&A data is processed so that it can be easily organized into a Q&A format.

[0382] 5. The server generates training content using a generative AI model based on the preprocessed data. Specifically, it combines text from the audio data, material data, and Q&A data into one, and creates text content for the training. The generative AI model analyzes this input data and generates content suitable for specific purposes.

[0383] 6. The server converts the generated training content into a video format, converting the generated text content into a visually and audibly appealing format and saving it as a video file.

[0384] 7. Finally, the server uploads the generated training video to the learning management system (LMS). This allows users to access and watch the training video from their devices. Users can log in to the LMS, select the desired video from the list of video content provided, and watch it.

[0385] Specific examples

[0386] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, as well as the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also organized. This preprocessed data is then integrated into one, and training content is created using a generative AI model. Finally, the generated content is converted into video format and uploaded to the LMS. Users can access the LMS and watch the training video to quickly understand how to use the new customer management system.

[0387] The processing flow will be explained below.

[0388] Step 1:

[0389] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0390] Step 2:

[0391] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[0392] Step 3:

[0393] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[0394] Step 4:

[0395] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers into an unambiguous format to make them easier to use.

[0396] Step 5:

[0397] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0398] Step 6:

[0399] The server converts the generated training content into a video format, and generates a video file based on the generated text content that is visually and audibly easy to understand, for example by adding narration corresponding to the text and incorporating slides and animations.

[0400] Step 7:

[0401] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, allowing users to view the training content at any time.

[0402] Step 8:

[0403] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0404] Example 1

[0405] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0406] In the conventional training content creation process, it takes a lot of time and effort to manually organize and integrate audio data, materials, and Q&A data from information sessions into training content. It is also difficult to consistently provide appropriate information, resulting in inconsistent training quality. The present invention aims to automate data processing and training content generation after information sessions, thereby providing efficient, high-quality training content.

[0407] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0408] In this invention, the server includes means for acquiring audio data of the briefing session, means for acquiring material data and Q&A data, means for converting the acquired audio data into text data, means for preprocessing the acquired material data and Q&A data, means for generating training information using an artificial intelligence model based on the preprocessed data, means for converting the generated training information into visual and audio media, and means for transmitting the generated visual and audio media to a learning management system. This automates data processing after the briefing session and enables the rapid provision of unified, high-quality training content.

[0409] "Audio data from a briefing session" refers to a digital audio file that records what was said at a briefing session, lecture, etc.

[0410] "Material data" refers to presentation materials and related document files used during the briefing session.

[0411] "Question and answer data" refers to records of questions and answers exchanged between participants and instructors during the information session.

[0412] The "means for converting into text data" is an automatic speech recognition technology for extracting voice data as text information.

[0413] "Preprocessing means" refers to procedures for arranging collected document data and Q&A data into a unified format and removing noise and unnecessary information.

[0414] An "artificial intelligence model" is a machine learning algorithm that generates appropriate output based on input data.

[0415] "Training information" is educational content that organizes the contents of the information session and provides them in a format suitable for learning purposes.

[0416] "Visual and audio media" refers to video files and other multimedia formats that contain text, image, and audio data.

[0417] A "learning management system" is an information management system that centrally manages training content provided by educational institutions and companies, and controls access and progress.

[0418] The "means for obtaining" is a method or device for reading data using a specified data file path.

[0419] "Transmitting means" refers to the data transfer method or communication means for uploading the generated training content to the learning management system.

[0420] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Specific embodiments are described below.

[0421] The server first collects the audio data of the briefing. During this process, the data is read into memory in binary format using a pre-specified audio file path (e.g., " / data / meeting_audio.wav"). Next, the server collects the presentation data (e.g., " / data / presentation.pdf") and the Q&A data (e.g., " / data / Q&A.txt"). These data are read from the specified paths and stored in memory in the form of a PDF file and a text file, respectively.

[0422] The server then converts the audio data into text using automatic speech recognition (ASR) technology, such as the Google Cloud Speech-to-Text API, which converts the audio data into text information and makes the contents of the briefing available in text format.

[0423] The server then preprocesses the document data and Q&A data. The document data is converted into a unified format and unnecessary information and noise are removed. The Q&A data is then processed to make it easier to organize into a Q&A format. For example, timestamps and unnecessary background information are removed from the document data, and the Q&A data is labeled into question and answer pairs.

[0424] The server generates training content using a generative AI model based on the preprocessed data. For example, using OpenAI's GPT-3, it inputs the following prompt: "Please create training content based on the following materials, audio text, and Q&A data." Based on this prompt, appropriate text content is generated.

[0425] The generated training content is then converted into visual and audio media, using Adobe Premiere Pro, for example, to add text as subtitles and composite the video with relevant images and graphics, with the final video file saved in MP4 format or similar.

[0426] Finally, the server uploads the generated training videos to the learning management system. During this process, the video files are uploaded directly from the server using the LMS's API, and metadata (lecture title, description, etc.) is also registered. Users can then log in to the LMS from their devices and watch the generated training videos. This automates data processing after the briefing session, making it possible to quickly provide uniform, high-quality training content.

[0427] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0428] Step 1:

[0429] The server collects audio data from the briefing session.

[0430] Input: The path to the information session recording file (e.g., " / data / meeting_audio.wav").

[0431] Specific operation: The server reads the audio file in binary format from the specified path and stores it in memory.

[0432] Output: Binary format of the audio data.

[0433] Step 2:

[0434] The server collects the material data and the question and answer data.

[0435] Input: File path of presentation materials (e.g. " / data / presentation.pdf") and file path of Q&A recording (e.g. " / data / Q&A.txt").

[0436] Specific operation: The server reads the PDF file and text file from each path and stores them in memory as document data and Q&A data.

[0437] Output: Document data (PDF format) and Q&A data (text format).

[0438] Step 3:

[0439] The server converts the voice data into text data.

[0440] Input: The binary format of the audio data collected in step 1.

[0441] Specific operation: The server uses automatic speech recognition technology such as the Google Cloud Speech-to-Text API to extract the voice data as text information.

[0442] Output: Text data converted from audio data.

[0443] Step 4:

[0444] The server pre-processes the document data and the question and answer data.

[0445] Input: Document data and Q&A data collected in Step 2.

[0446] Specific operation: The server converts the document data into a unified format, removes noise and unnecessary information, and processes the Q&A data so that it can be easily organized into a Q&A format.

[0447] Output: Preprocessed material data and Q&A data.

[0448] Step 5:

[0449] The server generates training content using a generative AI model based on the preprocessed data.

[0450] Input: Text data obtained in step 3, preprocessed document data and Q&A data in step 4.

[0451] Specific operation: The server uses a generative AI model (e.g., OpenAI's GPT-3) and inputs a prompt such as: "Create training content based on the following materials, audio text, and question and answer data."

[0452] Output: Generated training text content.

[0453] Step 6:

[0454] The server converts the generated training content into a video format.

[0455] Input: The training text content generated in step 5.

[0456] What it does: The server converts the generated text content into visual and auditory media and saves it as a video file. Using software such as Adobe Premiere Pro, the text is added as subtitles and composited into a video with associated images and graphics.

[0457] Output: Training video file (e.g. MP4 format).

[0458] Step 7:

[0459] The server uploads the generated training videos to a learning management system.

[0460] Input: The training video file created in step 6.

[0461] Specific operation: The server uses the API of the LMS platform (e.g., Moodle) to upload the video file and also registers metadata (lecture title, description, etc.).

[0462] Output: Training videos uploaded to a learning management system.

[0463] (Application example 1)

[0464] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0465] As new equipment and technologies are frequently introduced in factories, it is necessary to provide quick and efficient training for workers on operation and maintenance. However, current training methods require instructors to coordinate schedules, prepare materials, and provide face-to-face explanations, which are costly and time-consuming. This has led to a demand for automated generation of training content and its effective use.

[0466] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0467] In this invention, the server includes a means for collecting audio data of the briefing session, a means for collecting material data and Q&A data, and a means for converting the collected audio data into text data, thereby enabling the server to generate training content using a generative AI model based on the preprocessed data, convert the generated training content into video format, upload the generated video to a learning management system, and provide the training content to visual and audio devices.

[0468] "Audio data from a briefing session" refers to data that digitally records the verbal information spoken during the briefing session.

[0469] "Material data" refers to the data of presentation materials and documents used at the briefing.

[0470] "Question and answer data" refers to recorded data of questions and answers exchanged between participants and the lecturer during the information session.

[0471] "Means for converting into text data" refers to the process of converting voice data into text information using natural language processing technology.

[0472] "Preprocessing" refers to operations such as standardizing the format of collected data and removing noise so that the data can be easily analyzed and integrated.

[0473] A "generative AI model" refers to an algorithm or program that uses machine learning and artificial intelligence techniques to generate new content based on input data.

[0474] "Training Content" refers collectively to information and materials provided for learning or training purposes.

[0475] "Means for converting to video format" refers to the technology and tools that convert generated content such as text and charts into video files.

[0476] A "Learning Management System (LMS)" is a system that manages and distributes electronic learning materials, and is a platform that allows users to manage their own progress while learning.

[0477] "Visual and auditory devices" refers to devices that allow users to view and listen to learning content, including smart glasses and head-mounted displays.

[0478] This invention relates to a system that automatically generates training content for learning about the operation and maintenance techniques of robots used in factories. This system collects audio data, document data, and question and answer data from briefing sessions, preprocesses them, and then automatically generates training content using a generative AI model.

[0479] Hardware and software used

[0480] The main hardware and software components for implementing this system are as follows:

[0481] 1. Smart glasses (e.g., Google Glass)

[0482] 2. Head-Mounted Display (HMD)

[0483] 3. Automatic Speech Recognition (ASR) technology (e.g., Google Speech-to-Text API)

[0484] 4. Generative AI models (e.g., OpenAI GPT-4)

[0485] 5. Video editing tools (e.g. Adobe Premiere Pro, FFmpeg)

[0486] 6. Learning Management System (LMS)

[0487] Overview of system processing

[0488] The server processes the data in the following steps and ultimately generates the training content.

[0489] 1. Data Collection:

[0490] The server collects audio data from the briefings held at the factory, along with the presentation materials and Q&A sessions. The collected audio data is recorded by microphones installed in the smart glasses or robots.

[0491] 2. Text conversion:

[0492] The audio data is converted into text using automatic speech recognition (ASR) technology, which makes the contents of the briefing available as text information.

[0493] 3. Data Preprocessing:

[0494] The document data and Q&A data are preprocessed. The format of the document data is standardized and unnecessary information is deleted. The Q&A data is also organized in Q&A format.

[0495] 4. Content Generation:

[0496] Based on the preprocessed data, a generative AI model is used to generate training content. Specifically, the generative AI model combines text from the audio data, material data, and question and answer data to create text content for training.

[0497] 5. Video conversion and uploading:

[0498] The generated text content is converted into a video format, which is then converted into a visually and audibly appealing format using a video editing tool and saved as a video file.

[0499] Finally, the generated training videos are uploaded to a learning management system (LMS), where users can access the LMS through smart glasses or HMDs and watch the videos to learn operation procedures and maintenance techniques.

[0500] Specific examples

[0501] For example, consider a training session on the introduction of a new welding robot. A worker wearing smart glasses records the session and simultaneously collects the presentation materials and Q&A. ASR converts the speech to text and organizes the material data into a unified format. The training content is then generated by inputting the following prompt sentences into a generative AI model:

[0502] Example prompt sentence:

[0503] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[0504] 1. Audio text: <Audio text data>

[0505] 2. Presentation materials: <Data in a unified format>

[0506] 3. Q&A: <Organized Q&A data>

[0507] The generated training text is then edited using a video editing tool and finally uploaded to the LMS, allowing workers to access the training video on their HMD and progress through their studies.

[0508] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0509] Step 1:

[0510] The server collects audio data from the briefing session. Audio is recorded through microphones installed in the smart glasses or robot. The collected audio data is stored in memory in binary format. The input is audio signal data, and the output is binary audio data.

[0511] Step 2:

[0512] The server collects material data and question and answer data. Material data is presentation materials and documents, and question and answer data is records of questions and answers. These data are transferred to the server via file paths and stored in memory in text format. The input is material and question and answer files, and the output is text format data.

[0513] Step 3:

[0514] The server converts the collected voice data into text data using automatic speech recognition (ASR) technology. It uses an automatic speech recognition system (e.g., Google Speech-to-Text API) to analyze the voice data and convert it into text data. In this step, the voice signal is converted into text information, and the specific content becomes available as text information. The input is binary voice data, and the output is text data.

[0515] Step 4:

[0516] The server preprocesses the collected document data and Q&A data. The document data is unified in format and unnecessary information is removed before being converted into text format. The Q&A data is similarly formatted so that questions and answers are clearly distinguished. The input is the raw document data and Q&A data, and the output is preprocessed text data.

[0517] Step 5:

[0518] The server generates training content using a generative AI model based on the preprocessed data. The following prompt sentence is input to the generative AI model (e.g., OpenAI GPT-4) to generate text content for training:

[0519] Example prompt sentence:

[0520] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[0521] 1. Audio text: <Audio text data>

[0522] 2. Presentation materials: <Data in a unified format>

[0523] 3. Q&A: <Organized Q&A data>

[0524] The input is preprocessed speech text, material data, and question and answer data, and the output is text content for training.

[0525] Step 6:

[0526] The server converts the generated training content into a video format. Using a video editing tool (e.g. Adobe Premiere Pro, FFmpeg), it converts the content into a video file with visual and auditory appeal. The input is text content, and the output is a video file.

[0527] Step 7:

[0528] The server uploads the generated training videos to a learning management system (LMS). At this time, the video files are registered in the system so that users can access the LMS and watch the videos. Users can access the LMS using smart glasses or a head-mounted display, watch the training videos, and progress with their studies. The input is the video file, and the output is the training video uploaded to the LMS.

[0529] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0530] As a specific embodiment of this invention, we will explain a system that combines an emotion engine with a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Furthermore, this system uses an emotion engine that recognizes user emotions and utilizes emotion data to generate and update training content.

[0531] Program processing and natural language explanation

[0532] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0533] 2. The server collects the material data and Q&A data. The material data is the presentation materials and documents used in the briefing, and the Q&A data is the content of the questions and answers asked during the briefing. These data are also loaded by specifying the file path. The material data and Q&A data are stored in memory in text format.

[0534] 3. The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the speech content is converted into text information. This text data makes the content of the briefing available in written form.

[0535] 4. The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use.

[0536] 5. The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are combined and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0537] 6. The server converts the generated training content into a video format, generating a visually and audibly easy-to-understand video file based on the generated text content, for example by adding narration corresponding to the text and incorporating slides and animations.

[0538] 7. The server uploads the generated videos to a learning management system (LMS), which is an online platform accessed by employees. By uploading the generated training videos to this system, users can view the training content at any time.

[0539] 8. The user uses the device to access the learning management system (LMS). The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0540] 9. The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice through a camera and microphone, and monitors changes in emotions in real time.

[0541] 10. The server evaluates the training content based on the collected emotional data. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video speed. It also uses the emotional data as feedback to improve the quality of future training content.

[0542] Specific examples

[0543] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, along with the presentation materials and Q&A sessions. The collected audio data is converted into text using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A sessions are also organized. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted accordingly based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed. This allows users to gain a deeper understanding of the new service and its operational flow, resulting in faster and more effective customer support.

[0544] The processing flow will be explained below.

[0545] Step 1:

[0546] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0547] Step 2:

[0548] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[0549] Step 3:

[0550] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[0551] Step 4:

[0552] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and removing unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use. For example, it standardizes the text to lowercase and removes unnecessary spaces and noise.

[0553] Step 5:

[0554] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0555] Step 6:

[0556] The server converts the generated training content into a video format. Based on the generated text content, it generates a video file that is visually and audibly easy to understand. Specifically, it generates narration corresponding to the text and combines it with slides and animations to create a video file.

[0557] Step 7:

[0558] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, and uploads the generated training videos to this system, allowing users to view the training content at any time.

[0559] Step 8:

[0560] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0561] Step 9:

[0562] The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice via a camera and microphone, and monitors changes in emotions in real time.

[0563] Step 10:

[0564] The server evaluates the training content based on the collected emotional data and makes necessary adjustments. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video playback speed. The emotional data can also be used as feedback to improve the quality of future training content.

[0565] Example 2

[0566] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0567] Conventional training content generation systems have difficulty efficiently collecting and processing the contents of information sessions and automatically generating training content. It is also difficult to generate and update content that reflects changes in users' understanding and emotions in real time. This results in a large amount of manual work being required to provide effective training, resulting in reduced productivity.

[0568] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0569] In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and question-and-answer data, means for converting the collected audio data into text data, means for preprocessing the collected material data and question-and-answer data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for recognizing user emotions, and means for evaluating and updating the training content based on the recognized emotion data. This makes it possible to efficiently collect the contents of the briefing session and dynamically generate and update training content according to the user's level of understanding.

[0570] "Audio Data" refers to the content of the audio files recorded during the briefing.

[0571] "Material data" refers to presentation materials and documents used at the briefing.

[0572] "Question and answer data" refers to the content of questions and answers asked during the briefing session.

[0573] "Text data" refers to collected voice data converted into text information using an automatic speech recognition system.

[0574] "Preprocessing" refers to the process of standardizing the format of document data, deleting unnecessary parts, and organizing question and answer data.

[0575] "Generative AI model" refers to artificial intelligence techniques that automatically generate new training content using collected and pre-processed data.

[0576] "Video format" refers to the conversion of generated training text content into a visually and audibly appealing video file.

[0577] "Learning Management System (LMS)" means an online platform through which employees access and learn training content via the internet.

[0578] "Emotion data" refers to data that records changes in emotions recognized by the emotion engine from the user's facial expressions and tone of voice.

[0579] "Evaluate and update" refers to the process of judging the quality of training content based on sentiment data and improving and adjusting the content as needed.

[0580] MODE FOR CARRYING OUT THE INVENTION

[0581] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). It also has the ability to recognize user emotions and reflect them in the generation and updating of training content. This system collects audio data, materials, and Q&A data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[0582] Audio data collection and conversion

[0583] The server collects the audio data of the briefing and stores it in memory in binary format. It then converts the collected audio data into text data using an automatic speech recognition (ASR) system, such as the Google Cloud Speech-to-Text API, making the audio content available as text information.

[0584] Collection and preprocessing of document data and Q&A data

[0585] The server loads the presentation materials, documents, and Q&A data used in the briefing into memory in text format. Preprocessing of the material data includes standardizing the format and deleting unnecessary parts, and for the Q&A data, it organizes the questions and answers and converts them into a more user-friendly format.

[0586] Training content generation and video conversion

[0587] The server integrates the pre-processed data and feeds it to a generative AI model to automatically generate training content. For example, the following prompt sentences can be used to input the generative AI model:

[0588] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[0589] The generated training content is converted into a video file that appeals to both the visual and auditory senses, incorporating narration, slides, animations, and other elements to create a visually easy-to-understand format.

[0590] Upload to a learning management system

[0591] The generated training content is converted into video format and uploaded to a learning management system (LMS), where users can access and watch the training videos.

[0592] User emotion recognition and content update

[0593] The server uses an emotion engine to recognize the user's emotions while watching the training video. It monitors the user's facial expressions and tone of voice through a camera and microphone to collect emotion data in real time. Based on this emotion data, the quality of the training content can be evaluated and updated as necessary. For example, if the user is confused, additional explanations can be generated or the video playback speed can be adjusted.

[0594] Specific examples

[0595] For example, when an introduction briefing for a new customer management system is held, the server collects the audio data from the briefing, as well as the presentation materials and Q&A sessions used. The audio data is converted into text data by the ASR system, the material data is organized into a unified format, and the Q&A data is also organized. Training content is then created using a generative AI model based on the preprocessed data. The generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch this training video. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. This allows users to gain a deeper understanding of the new system and enables faster and more effective customer service.

[0596] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0597] Step 1:

[0598] Audio data collection

[0599] The server receives the file path of the audio data of the briefing as input, opens the file, and reads the data into memory in binary format. This process saves the collected audio data in memory. Specifically, it opens the specified audio file and reads the contents in binary format.

[0600] Step 2:

[0601] Collection of document data and Q&A data

[0602] The server receives the file paths for the presentation data and Q&A data as input, opens each file, and reads the data into memory in text format. This process saves the presentation data and Q&A data in memory. Specifically, the presentation materials and Q&A content are read as text data.

[0603] Step 3:

[0604] Converting audio data to text

[0605] The server receives the collected voice data as input and converts it into text data using an automatic speech recognition (ASR) system. This process makes the voice content available as text information. Specifically, the voice data is input into the ASR system, and text information is obtained as output.

[0606] Step 4:

[0607] Preprocessing of document data and Q&A data

[0608] The server receives the collected document data and Q&A data as input, standardizes the data format, deletes unnecessary parts, and organizes the questions and answers. This process results in preprocessed data. Specifically, unnecessary parts of the text are deleted and the format is adjusted.

[0609] Step 5:

[0610] Training content generation

[0611] The server integrates the preprocessed speech-to-text data, the materials data, and the Q&A data, and inputs them into the generative AI model along with prompts. This process generates training content. Specifically, the following prompts are used:

[0612] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[0613] Step 6:

[0614] Content video conversion

[0615] The server receives the generated training content as input and converts it into a visually and audibly appealing video file. This process results in a training video containing narration, slides, and animations. Specifically, the server creates slides with narration and animations based on the generated text content and converts them into a video format.

[0616] Step 7:

[0617] Upload to LMS

[0618] The server receives the generated video file as input and uploads it to the learning management system (LMS). This process makes it accessible to users. Specifically, the video file is sent to the server using the LMS's upload function.

[0619] Step 8:

[0620] User LMS Access

[0621] Users access the LMS using their devices and select and watch the training videos provided. This process allows users to learn the training content. Specifically, users log in to the LMS, select the desired video from the video list, and press the play button.

[0622] Step 9:

[0623] emotion recognition

[0624] The server monitors the user's facial expressions and tone of voice while watching the training video through a camera and microphone, and collects emotional data using an emotion engine. This process yields the user's emotional data. Specifically, it analyzes changes in facial expressions and voice while watching in real time and records changes in emotions.

[0625] Step 10:

[0626] Evaluating and updating training content

[0627] The server receives the collected emotional data as input, evaluates the training content, and updates it as necessary to improve the quality of the training. Specifically, it generates additional explanations or adjusts the video playback speed depending on the user's emotional state.

[0628] (Application example 2)

[0629] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0630] Conventional training content generation systems generate training content based on recordings of information sessions and materials, but because they do not take into account the level of understanding and emotions of individual end users, the quality of the training content is uniform and may not meet the needs of end users.In addition, the generated content is fixed and difficult to adjust in real time, which creates the problem of not being able to maximize the learning effect of end users.

[0631] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and Q&A data, means for converting the collected audio data into text data, means for preprocessing the collected material data and Q&A data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for collecting end-user emotional data using sensors mounted on the robot, and means for evaluating and adjusting the training content based on the collected emotional data. This allows the training content to be adjusted in real time taking into account the end-user's level of understanding and emotions, enabling advanced training tailored to individual needs.

[0632] "Audio data from the briefing session" refers to an audio file that records what was said during the briefing session.

[0633] "Material data" refers to presentation materials and document files used in the briefing session.

[0634] "Question and answer data" is text data of questions asked during the briefing session and the answers to those questions.

[0635] "Means for converting to text data" refers to a system or algorithm that analyzes audio data and makes it usable as text information.

[0636] "Preprocessing means" refers to means for standardizing the formats of collected document data and question and answer data and deleting unnecessary parts.

[0637] A "generative AI model" is a type of artificial intelligence that automatically generates new training content based on given data.

[0638] "Training Content" means learning materials for end users generated by a generative AI model.

[0639] "Means of converting to video format" refers to the process of converting the generated training content into visual and audio materials.

[0640] A "learning management system" is an online platform where training content can be uploaded and accessed by end users for learning.

[0641] "Sensors mounted on robots" are devices such as cameras and microphones attached to robots, and are used to collect emotional data from end users.

[0642] "Emotional data" refers to psychological reaction information obtained from the end user's facial expressions and tone of voice.

[0643] The "means for evaluating and adjusting" refers to a means for appropriately correcting or providing additional explanations to pre-generated training content based on the collected emotional data.

[0644] MODE FOR CARRYING OUT THE INVENTION

[0645] As a specific embodiment for carrying out the present invention, a system will be described that automatically generates training content after an information session and adjusts the content based on emotional data of end users.

[0646] Overall system configuration

[0647] This system operates in the following manner.

[0648] Audio data collection

[0649] The server collects audio data from the briefing session. Specifically, it specifies the path to the audio file recorded during the briefing session, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0650] Collection of document data and question and answer data

[0651] The server collects presentation materials, documents, and question and answer data used in the briefing. These data are also loaded by specifying the file path. The material data and question and answer data are stored in memory in text format.

[0652] Automatic Speech Recognition (ASR) System

[0653] The server uses an automatic speech recognition (ASR) system to convert the collected voice data into text data using the Google Cloud Speech-to-Text API, which makes the content of the briefing available as a transcript.

[0654] Preprocessing of document data and Q&A data

[0655] The server standardizes the format of the document data and deletes unnecessary parts, and organizes the questions and answers in the question and answer data to make it easier to use.

[0656] Generating training content using generative AI models

[0657] Based on the preprocessed data, the server generates training content using a generative AI model such as GPT-4. The preprocessed audio text, document data, and Q&A data are combined and fed into the generative AI model as input. The generative AI model then analyzes this data and automatically generates text content for the training.

[0658] Convert to video format

[0659] Based on the generated text content, the server generates a visually and audibly comprehensible video file, including adding corresponding narration to the text and incorporating slides and animations. This process uses OpenCV and FFmpeg.

[0660] Upload to Learning Management System (LMS)

[0661] The server uploads the generated videos to a learning management system, which is an online platform accessed by employees, allowing users to view the training content at any time.

[0662] Collecting Emotional Data

[0663] Sensors (camera and microphone) installed on the robot are used to collect emotional data while the end user is watching the training video. The facial expression data is analyzed using the Microsoft Azure Face API, and emotions are recognized from the tone of voice.

[0664] Evaluate and adjust training content

[0665] The server evaluates the training content based on the collected emotional data and, if necessary, includes sections in the video for additional explanation or adjusts the playback speed.

[0666] Specific examples

[0667] For example, when a briefing on new maintenance procedures is held, the server collects the audio data of the briefing, along with the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also compiled. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed.

[0668] Prompt Sentence Examples

[0669] New maintenance procedure briefing text:

[0670] 1. Engine disassembly procedure

[0671] 2. Parts inspection items

[0672] 3. Lubrication method

[0673] 4. Vibration measurement method

[0674] Q&A:

[0675] Q: What type of lubricant?

[0676] A: We will use ABC lubricant from XYZ company.

[0677] Use this information to generate training content that includes specific operating procedures and precautions, from engine disassembly procedures to lubrication methods.

[0678] The system makes it possible to adjust training content in real time, taking into account the end user's level of understanding and emotions, improving maintenance efficiency within the factory and ensuring that new procedures are understood quickly and reliably.

[0679] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0680] Step 1:

[0681] The server collects the audio data of the briefing. Specifically, it specifies the path of the audio file recorded during the briefing, opens the file, and reads the data. The input is the recorded audio file, and the output is binary audio data. The audio data is stored in buffer memory.

[0682] Step 2:

[0683] The server collects material data and question and answer data. The material data is the presentation materials and document format files used in the briefing, and the question and answer data is text data of questions and answers asked during the briefing. The input is the path of the material file and the question and answer file, and the output is stored in memory in text format.

[0684] Step 3:

[0685] The server converts the voice data into text data. An automatic speech recognition (ASR) system, in this case the Google Cloud Speech-to-Text API, is used to analyze the voice data and convert it into text information. The input is binary voice data, and the output is text data.

[0686] Step 4:

[0687] The server preprocesses the document data and Q&A data. Specifically, it standardizes the format of the document data, deletes unnecessary parts, and organizes the questions and answers in the Q&A data. The input is text-formatted document data and Q&A data, and the output is the preprocessed unified data.

[0688] Step 5:

[0689] Based on the preprocessed data, the server generates training content using a generative AI model. Specifically, the preprocessed audio text, material data, and Q&A data are integrated into one and supplied to the generative AI model (GPT-4). The input is the integrated preprocessed data, and the output is the generated training text content.

[0690] Step 6:

[0691] The server converts the generated training content into video format. Using OpenCV and FFmpeg, it adds narration corresponding to the text and incorporates slides and animations to generate the video. The input is the generated text content, and the output is a video file that is visually and audibly easy to understand.

[0692] Step 7:

[0693] The server uploads the generated videos to a learning management system (LMS). The LMS is an online platform accessed by employees, where the generated training videos are uploaded so that users can view the training content at any time. The input is the generated video file, and the output is the training content uploaded to the LMS.

[0694] Step 8:

[0695] A user uses a device to access a learning management system (LMS) and watch a provided training video. The user logs in to the LMS, selects the target training video, and plays it. The input is the login information for the LMS and the target training video data, and the output is the training video that is played.

[0696] Step 9:

[0697] The robot collects emotional data from the user using sensors (camera and microphone) installed on the device. It uses the Microsoft Azure Face API to analyze the facial expression data and recognize emotions from the tone of voice. The input is video and audio data, and the output is analyzed emotional data.

[0698] Step 10:

[0699] The server evaluates and adjusts the training content based on the collected emotional data, including adding sections to the video for additional explanations or adjusting the playback speed as needed. The input is emotional data, and the output is the adjusted training content.

[0700] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0701] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0702] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0703] [Third embodiment]

[0704] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0705] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0706] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0707] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0708] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0709] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0710] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0711] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0712] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0713] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0714] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0715] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0716] As a specific embodiment of this invention, we will explain a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[0717] Program processing and natural language explanation

[0718] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded at the briefing and reads the data. This audio data is stored in memory in binary format.

[0719] 2. The server collects the material data and question and answer data. The material data corresponds to the presentation materials and documents used in the briefing, and the question and answer data is a record of the questions and answers exchanged between the participants and the lecturer during the briefing. These data are also read by specifying a file path and stored in memory in text format.

[0720] 3. The server converts the voice data into text data using automatic speech recognition (ASR) technology. The converted text data makes the contents of the briefing available as text information.

[0721] 4. The server preprocesses the document data and Q&A data. The format of the document data is standardized and noise and unnecessary information is removed. The Q&A data is processed so that it can be easily organized into a Q&A format.

[0722] 5. The server generates training content using a generative AI model based on the preprocessed data. Specifically, it combines text from the audio data, material data, and Q&A data into one, and creates text content for the training. The generative AI model analyzes this input data and generates content suitable for specific purposes.

[0723] 6. The server converts the generated training content into a video format, converting the generated text content into a visually and audibly appealing format and saving it as a video file.

[0724] 7. Finally, the server uploads the generated training video to the learning management system (LMS). This allows users to access and watch the training video from their devices. Users can log in to the LMS, select the desired video from the list of video content provided, and watch it.

[0725] Specific examples

[0726] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, as well as the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also organized. This preprocessed data is then integrated into one, and training content is created using a generative AI model. Finally, the generated content is converted into video format and uploaded to the LMS. Users can access the LMS and watch the training video to quickly understand how to use the new customer management system.

[0727] The processing flow will be explained below.

[0728] Step 1:

[0729] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0730] Step 2:

[0731] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[0732] Step 3:

[0733] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[0734] Step 4:

[0735] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers into an unambiguous format to make them easier to use.

[0736] Step 5:

[0737] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0738] Step 6:

[0739] The server converts the generated training content into a video format, and generates a video file based on the generated text content that is visually and audibly easy to understand, for example by adding narration corresponding to the text and incorporating slides and animations.

[0740] Step 7:

[0741] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, allowing users to view the training content at any time.

[0742] Step 8:

[0743] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0744] Example 1

[0745] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0746] In the conventional training content creation process, it takes a lot of time and effort to manually organize and integrate audio data, materials, and Q&A data from information sessions into training content. It is also difficult to consistently provide appropriate information, resulting in inconsistent training quality. The present invention aims to automate data processing and training content generation after information sessions, thereby providing efficient, high-quality training content.

[0747] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0748] In this invention, the server includes means for acquiring audio data of the briefing session, means for acquiring material data and Q&A data, means for converting the acquired audio data into text data, means for preprocessing the acquired material data and Q&A data, means for generating training information using an artificial intelligence model based on the preprocessed data, means for converting the generated training information into visual and audio media, and means for transmitting the generated visual and audio media to a learning management system. This automates data processing after the briefing session and enables the rapid provision of unified, high-quality training content.

[0749] "Audio data from a briefing session" refers to a digital audio file that records what was said at a briefing session, lecture, etc.

[0750] "Material data" refers to presentation materials and related document files used during the briefing session.

[0751] "Question and answer data" refers to records of questions and answers exchanged between participants and instructors during the information session.

[0752] The "means for converting into text data" is an automatic speech recognition technology for extracting voice data as text information.

[0753] "Preprocessing means" refers to procedures for arranging collected document data and Q&A data into a unified format and removing noise and unnecessary information.

[0754] An "artificial intelligence model" is a machine learning algorithm that generates appropriate output based on input data.

[0755] "Training information" is educational content that organizes the contents of the information session and provides them in a format suitable for learning purposes.

[0756] "Visual and audio media" refers to video files and other multimedia formats that contain text, image, and audio data.

[0757] A "learning management system" is an information management system that centrally manages training content provided by educational institutions and companies, and controls access and progress.

[0758] The "means for obtaining" is a method or device for reading data using a specified data file path.

[0759] "Transmitting means" refers to the data transfer method or communication means for uploading the generated training content to the learning management system.

[0760] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Specific embodiments are described below.

[0761] The server first collects the audio data of the briefing. During this process, the data is read into memory in binary format using a pre-specified audio file path (e.g., " / data / meeting_audio.wav"). Next, the server collects the presentation data (e.g., " / data / presentation.pdf") and the Q&A data (e.g., " / data / Q&A.txt"). These data are read from the specified paths and stored in memory in the form of a PDF file and a text file, respectively.

[0762] The server then converts the audio data into text using automatic speech recognition (ASR) technology, such as the Google Cloud Speech-to-Text API, which converts the audio data into text information and makes the contents of the briefing available in text format.

[0763] The server then preprocesses the document data and Q&A data. The document data is converted into a unified format and unnecessary information and noise are removed. The Q&A data is then processed to make it easier to organize into a Q&A format. For example, timestamps and unnecessary background information are removed from the document data, and the Q&A data is labeled into question and answer pairs.

[0764] The server generates training content using a generative AI model based on the preprocessed data. For example, using OpenAI's GPT-3, it inputs the following prompt: "Please create training content based on the following materials, audio text, and Q&A data." Based on this prompt, appropriate text content is generated.

[0765] The generated training content is then converted into visual and audio media, using Adobe Premiere Pro, for example, to add text as subtitles and composite the video with relevant images and graphics, with the final video file saved in MP4 format or similar.

[0766] Finally, the server uploads the generated training videos to the learning management system. During this process, the video files are uploaded directly from the server using the LMS's API, and metadata (lecture title, description, etc.) is also registered. Users can then log in to the LMS from their devices and watch the generated training videos. This automates data processing after the briefing session, making it possible to quickly provide uniform, high-quality training content.

[0767] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0768] Step 1:

[0769] The server collects audio data from the briefing session.

[0770] Input: The path to the information session recording file (e.g., " / data / meeting_audio.wav").

[0771] Specific operation: The server reads the audio file in binary format from the specified path and stores it in memory.

[0772] Output: Binary format of the audio data.

[0773] Step 2:

[0774] The server collects the material data and the question and answer data.

[0775] Input: File path of presentation materials (e.g. " / data / presentation.pdf") and file path of Q&A recording (e.g. " / data / Q&A.txt").

[0776] Specific operation: The server reads the PDF file and text file from each path and stores them in memory as document data and Q&A data.

[0777] Output: Document data (PDF format) and Q&A data (text format).

[0778] Step 3:

[0779] The server converts the voice data into text data.

[0780] Input: The binary format of the audio data collected in step 1.

[0781] Specific operation: The server uses automatic speech recognition technology such as the Google Cloud Speech-to-Text API to extract the voice data as text information.

[0782] Output: Text data converted from audio data.

[0783] Step 4:

[0784] The server pre-processes the document data and the question and answer data.

[0785] Input: Document data and Q&A data collected in Step 2.

[0786] Specific operation: The server converts the document data into a unified format, removes noise and unnecessary information, and processes the Q&A data so that it can be easily organized into a Q&A format.

[0787] Output: Preprocessed material data and Q&A data.

[0788] Step 5:

[0789] The server generates training content using a generative AI model based on the preprocessed data.

[0790] Input: Text data obtained in step 3, preprocessed document data and Q&A data in step 4.

[0791] Specific operation: The server uses a generative AI model (e.g., OpenAI's GPT-3) and inputs a prompt such as: "Create training content based on the following materials, audio text, and question and answer data."

[0792] Output: Generated training text content.

[0793] Step 6:

[0794] The server converts the generated training content into a video format.

[0795] Input: The training text content generated in step 5.

[0796] What it does: The server converts the generated text content into visual and auditory media and saves it as a video file. Using software such as Adobe Premiere Pro, the text is added as subtitles and composited into a video with associated images and graphics.

[0797] Output: Training video file (e.g. MP4 format).

[0798] Step 7:

[0799] The server uploads the generated training videos to a learning management system.

[0800] Input: The training video file created in step 6.

[0801] Specific operation: The server uses the API of the LMS platform (e.g., Moodle) to upload the video file and also registers metadata (lecture title, description, etc.).

[0802] Output: Training videos uploaded to a learning management system.

[0803] (Application example 1)

[0804] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0805] As new equipment and technologies are frequently introduced in factories, it is necessary to provide quick and efficient training for workers on operation and maintenance. However, current training methods require instructors to coordinate schedules, prepare materials, and provide face-to-face explanations, which are costly and time-consuming. This has led to a demand for automated generation of training content and its effective use.

[0806] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0807] In this invention, the server includes a means for collecting audio data of the briefing session, a means for collecting material data and Q&A data, and a means for converting the collected audio data into text data, thereby enabling the server to generate training content using a generative AI model based on the preprocessed data, convert the generated training content into video format, upload the generated video to a learning management system, and provide the training content to visual and audio devices.

[0808] "Audio data from a briefing session" refers to data that digitally records the verbal information spoken during the briefing session.

[0809] "Material data" refers to the data of presentation materials and documents used at the briefing.

[0810] "Question and answer data" refers to recorded data of questions and answers exchanged between participants and the lecturer during the information session.

[0811] "Means for converting into text data" refers to the process of converting voice data into text information using natural language processing technology.

[0812] "Preprocessing" refers to operations such as standardizing the format of collected data and removing noise so that the data can be easily analyzed and integrated.

[0813] A "generative AI model" refers to an algorithm or program that uses machine learning and artificial intelligence techniques to generate new content based on input data.

[0814] "Training Content" refers collectively to information and materials provided for learning or training purposes.

[0815] "Means for converting to video format" refers to the technology and tools that convert generated content such as text and charts into video files.

[0816] A "Learning Management System (LMS)" is a system that manages and distributes electronic learning materials, and is a platform that allows users to manage their own progress while learning.

[0817] "Visual and auditory devices" refers to devices that allow users to view and listen to learning content, including smart glasses and head-mounted displays.

[0818] This invention relates to a system that automatically generates training content for learning about the operation and maintenance techniques of robots used in factories. This system collects audio data, document data, and question and answer data from briefing sessions, preprocesses them, and then automatically generates training content using a generative AI model.

[0819] Hardware and software used

[0820] The main hardware and software components for implementing this system are as follows:

[0821] 1. Smart glasses (e.g., Google Glass)

[0822] 2. Head-Mounted Display (HMD)

[0823] 3. Automatic Speech Recognition (ASR) technology (e.g., Google Speech-to-Text API)

[0824] 4. Generative AI models (e.g., OpenAI GPT-4)

[0825] 5. Video editing tools (e.g. Adobe Premiere Pro, FFmpeg)

[0826] 6. Learning Management System (LMS)

[0827] Overview of system processing

[0828] The server processes the data in the following steps and ultimately generates the training content.

[0829] 1. Data Collection:

[0830] The server collects audio data from the briefings held at the factory, along with the presentation materials and Q&A sessions. The collected audio data is recorded by microphones installed in the smart glasses or robots.

[0831] 2. Text conversion:

[0832] The audio data is converted into text using automatic speech recognition (ASR) technology, which makes the contents of the briefing available as text information.

[0833] 3. Data Preprocessing:

[0834] The document data and Q&A data are preprocessed. The format of the document data is standardized and unnecessary information is deleted. The Q&A data is also organized in Q&A format.

[0835] 4. Content Generation:

[0836] Based on the preprocessed data, a generative AI model is used to generate training content. Specifically, the generative AI model combines text from the audio data, material data, and question and answer data to create text content for training.

[0837] 5. Video conversion and uploading:

[0838] The generated text content is converted into a video format, which is then converted into a visually and audibly appealing format using a video editing tool and saved as a video file.

[0839] Finally, the generated training videos are uploaded to a learning management system (LMS), where users can access the LMS through smart glasses or HMDs and watch the videos to learn operation procedures and maintenance techniques.

[0840] Specific examples

[0841] For example, consider a training session on the introduction of a new welding robot. A worker wearing smart glasses records the session and simultaneously collects the presentation materials and Q&A. ASR converts the speech to text and organizes the material data into a unified format. The training content is then generated by inputting the following prompt sentences into a generative AI model:

[0842] Example prompt sentence:

[0843] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[0844] 1. Audio text: <Audio text data>

[0845] 2. Presentation materials: <Data in a unified format>

[0846] 3. Q&A: <Organized Q&A data>

[0847] The generated training text is then edited using a video editing tool and finally uploaded to the LMS, allowing workers to access the training video on their HMD and progress through their studies.

[0848] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0849] Step 1:

[0850] The server collects audio data from the briefing session. Audio is recorded through microphones installed in the smart glasses or robot. The collected audio data is stored in memory in binary format. The input is audio signal data, and the output is binary audio data.

[0851] Step 2:

[0852] The server collects material data and question and answer data. Material data is presentation materials and documents, and question and answer data is records of questions and answers. These data are transferred to the server via file paths and stored in memory in text format. The input is material and question and answer files, and the output is text format data.

[0853] Step 3:

[0854] The server converts the collected voice data into text data using automatic speech recognition (ASR) technology. It uses an automatic speech recognition system (e.g., Google Speech-to-Text API) to analyze the voice data and convert it into text data. In this step, the voice signal is converted into text information, and the specific content becomes available as text information. The input is binary voice data, and the output is text data.

[0855] Step 4:

[0856] The server preprocesses the collected document data and Q&A data. The document data is unified in format and unnecessary information is removed before being converted into text format. The Q&A data is similarly formatted so that questions and answers are clearly distinguished. The input is the raw document data and Q&A data, and the output is preprocessed text data.

[0857] Step 5:

[0858] The server generates training content using a generative AI model based on the preprocessed data. The following prompt sentence is input to the generative AI model (e.g., OpenAI GPT-4) to generate text content for training:

[0859] Example prompt sentence:

[0860] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[0861] 1. Audio text: <Audio text data>

[0862] 2. Presentation materials: <Data in a unified format>

[0863] 3. Q&A: <Organized Q&A data>

[0864] The input is preprocessed speech text, material data, and question and answer data, and the output is text content for training.

[0865] Step 6:

[0866] The server converts the generated training content into a video format. Using a video editing tool (e.g. Adobe Premiere Pro, FFmpeg), it converts the content into a video file with visual and auditory appeal. The input is text content, and the output is a video file.

[0867] Step 7:

[0868] The server uploads the generated training videos to a learning management system (LMS). At this time, the video files are registered in the system so that users can access the LMS and watch the videos. Users can access the LMS using smart glasses or a head-mounted display, watch the training videos, and progress with their studies. The input is the video file, and the output is the training video uploaded to the LMS.

[0869] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0870] As a specific embodiment of this invention, we will explain a system that combines an emotion engine with a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Furthermore, this system uses an emotion engine that recognizes user emotions and utilizes emotion data to generate and update training content.

[0871] Program processing and natural language explanation

[0872] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0873] 2. The server collects the material data and Q&A data. The material data is the presentation materials and documents used in the briefing, and the Q&A data is the content of the questions and answers asked during the briefing. These data are also loaded by specifying the file path. The material data and Q&A data are stored in memory in text format.

[0874] 3. The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the speech content is converted into text information. This text data makes the content of the briefing available in written form.

[0875] 4. The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use.

[0876] 5. The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are combined and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0877] 6. The server converts the generated training content into a video format, generating a visually and audibly easy-to-understand video file based on the generated text content, for example by adding narration corresponding to the text and incorporating slides and animations.

[0878] 7. The server uploads the generated videos to a learning management system (LMS), which is an online platform accessed by employees. By uploading the generated training videos to this system, users can view the training content at any time.

[0879] 8. The user uses the device to access the learning management system (LMS). The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0880] 9. The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice through a camera and microphone, and monitors changes in emotions in real time.

[0881] 10. The server evaluates the training content based on the collected emotional data. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video speed. It also uses the emotional data as feedback to improve the quality of future training content.

[0882] Specific examples

[0883] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, along with the presentation materials and Q&A sessions. The collected audio data is converted into text using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A sessions are also organized. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted accordingly based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed. This allows users to gain a deeper understanding of the new service and its operational flow, resulting in faster and more effective customer support.

[0884] The processing flow will be explained below.

[0885] Step 1:

[0886] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0887] Step 2:

[0888] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[0889] Step 3:

[0890] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[0891] Step 4:

[0892] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and removing unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use. For example, it standardizes the text to lowercase and removes unnecessary spaces and noise.

[0893] Step 5:

[0894] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[0895] Step 6:

[0896] The server converts the generated training content into a video format. Based on the generated text content, it generates a video file that is visually and audibly easy to understand. Specifically, it generates narration corresponding to the text and combines it with slides and animations to create a video file.

[0897] Step 7:

[0898] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, and uploads the generated training videos to this system, allowing users to view the training content at any time.

[0899] Step 8:

[0900] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[0901] Step 9:

[0902] The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice via a camera and microphone, and monitors changes in emotions in real time.

[0903] Step 10:

[0904] The server evaluates the training content based on the collected emotional data and makes necessary adjustments. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video playback speed. The emotional data can also be used as feedback to improve the quality of future training content.

[0905] Example 2

[0906] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0907] Conventional training content generation systems have difficulty efficiently collecting and processing the contents of information sessions and automatically generating training content. It is also difficult to generate and update content that reflects changes in users' understanding and emotions in real time. This results in a large amount of manual work being required to provide effective training, resulting in reduced productivity.

[0908] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0909] In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and question-and-answer data, means for converting the collected audio data into text data, means for preprocessing the collected material data and question-and-answer data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for recognizing user emotions, and means for evaluating and updating the training content based on the recognized emotion data. This makes it possible to efficiently collect the contents of the briefing session and dynamically generate and update training content according to the user's level of understanding.

[0910] "Audio Data" refers to the content of the audio files recorded during the briefing.

[0911] "Material data" refers to presentation materials and documents used at the briefing.

[0912] "Question and answer data" refers to the content of questions and answers asked during the briefing session.

[0913] "Text data" refers to collected voice data converted into text information using an automatic speech recognition system.

[0914] "Preprocessing" refers to the process of standardizing the format of document data, deleting unnecessary parts, and organizing question and answer data.

[0915] "Generative AI model" refers to artificial intelligence techniques that automatically generate new training content using collected and pre-processed data.

[0916] "Video format" refers to the conversion of generated training text content into a visually and audibly appealing video file.

[0917] "Learning Management System (LMS)" means an online platform through which employees access and learn training content via the internet.

[0918] "Emotion data" refers to data that records changes in emotions recognized by the emotion engine from the user's facial expressions and tone of voice.

[0919] "Evaluate and update" refers to the process of judging the quality of training content based on sentiment data and improving and adjusting the content as needed.

[0920] MODE FOR CARRYING OUT THE INVENTION

[0921] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). It also has the ability to recognize user emotions and reflect them in the generation and updating of training content. This system collects audio data, materials, and Q&A data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[0922] Audio data collection and conversion

[0923] The server collects the audio data of the briefing and stores it in memory in binary format. It then converts the collected audio data into text data using an automatic speech recognition (ASR) system, such as the Google Cloud Speech-to-Text API, making the audio content available as text information.

[0924] Collection and preprocessing of document data and Q&A data

[0925] The server loads the presentation materials, documents, and Q&A data used in the briefing into memory in text format. Preprocessing of the material data includes standardizing the format and deleting unnecessary parts, and for the Q&A data, it organizes the questions and answers and converts them into a more user-friendly format.

[0926] Training content generation and video conversion

[0927] The server integrates the pre-processed data and feeds it to a generative AI model to automatically generate training content. For example, the following prompt sentences can be used to input the generative AI model:

[0928] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[0929] The generated training content is converted into a video file that appeals to both the visual and auditory senses, incorporating narration, slides, animations, and other elements to create a visually easy-to-understand format.

[0930] Upload to a learning management system

[0931] The generated training content is converted into video format and uploaded to a learning management system (LMS), where users can access and watch the training videos.

[0932] User emotion recognition and content update

[0933] The server uses an emotion engine to recognize the user's emotions while watching the training video. It monitors the user's facial expressions and tone of voice through a camera and microphone to collect emotion data in real time. Based on this emotion data, the quality of the training content can be evaluated and updated as necessary. For example, if the user is confused, additional explanations can be generated or the video playback speed can be adjusted.

[0934] Specific examples

[0935] For example, when an introduction briefing for a new customer management system is held, the server collects the audio data from the briefing, as well as the presentation materials and Q&A sessions used. The audio data is converted into text data by the ASR system, the material data is organized into a unified format, and the Q&A data is also organized. Training content is then created using a generative AI model based on the preprocessed data. The generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch this training video. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. This allows users to gain a deeper understanding of the new system and enables faster and more effective customer service.

[0936] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0937] Step 1:

[0938] Audio data collection

[0939] The server receives the file path of the audio data of the briefing as input, opens the file, and reads the data into memory in binary format. This process saves the collected audio data in memory. Specifically, it opens the specified audio file and reads the contents in binary format.

[0940] Step 2:

[0941] Collection of document data and Q&A data

[0942] The server receives the file paths for the presentation data and Q&A data as input, opens each file, and reads the data into memory in text format. This process saves the presentation data and Q&A data in memory. Specifically, the presentation materials and Q&A content are read as text data.

[0943] Step 3:

[0944] Converting audio data to text

[0945] The server receives the collected voice data as input and converts it into text data using an automatic speech recognition (ASR) system. This process makes the voice content available as text information. Specifically, the voice data is input into the ASR system, and text information is obtained as output.

[0946] Step 4:

[0947] Preprocessing of document data and Q&A data

[0948] The server receives the collected document data and Q&A data as input, standardizes the data format, deletes unnecessary parts, and organizes the questions and answers. This process results in preprocessed data. Specifically, unnecessary parts of the text are deleted and the format is adjusted.

[0949] Step 5:

[0950] Training content generation

[0951] The server integrates the preprocessed speech-to-text data, the materials data, and the Q&A data, and inputs them into the generative AI model along with prompts. This process generates training content. Specifically, the following prompts are used:

[0952] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[0953] Step 6:

[0954] Content video conversion

[0955] The server receives the generated training content as input and converts it into a visually and audibly appealing video file. This process results in a training video containing narration, slides, and animations. Specifically, the server creates slides with narration and animations based on the generated text content and converts them into a video format.

[0956] Step 7:

[0957] Upload to LMS

[0958] The server receives the generated video file as input and uploads it to the learning management system (LMS). This process makes it accessible to users. Specifically, the video file is sent to the server using the LMS's upload function.

[0959] Step 8:

[0960] User LMS Access

[0961] Users access the LMS using their devices and select and watch the training videos provided. This process allows users to learn the training content. Specifically, users log in to the LMS, select the desired video from the video list, and press the play button.

[0962] Step 9:

[0963] emotion recognition

[0964] The server monitors the user's facial expressions and tone of voice while watching the training video through a camera and microphone, and collects emotional data using an emotion engine. This process yields the user's emotional data. Specifically, it analyzes changes in facial expressions and voice while watching in real time and records changes in emotions.

[0965] Step 10:

[0966] Evaluating and updating training content

[0967] The server receives the collected emotional data as input, evaluates the training content, and updates it as necessary to improve the quality of the training. Specifically, it generates additional explanations or adjusts the video playback speed depending on the user's emotional state.

[0968] (Application example 2)

[0969] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0970] Conventional training content generation systems generate training content based on recordings of information sessions and materials, but because they do not take into account the level of understanding and emotions of individual end users, the quality of the training content is uniform and may not meet the needs of end users.In addition, the generated content is fixed and difficult to adjust in real time, which creates the problem of not being able to maximize the learning effect of end users.

[0971] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and Q&A data, means for converting the collected audio data into text data, means for preprocessing the collected material data and Q&A data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for collecting end-user emotional data using sensors mounted on the robot, and means for evaluating and adjusting the training content based on the collected emotional data. This allows the training content to be adjusted in real time taking into account the end-user's level of understanding and emotions, enabling advanced training tailored to individual needs.

[0972] "Audio data from the briefing session" refers to an audio file that records what was said during the briefing session.

[0973] "Material data" refers to presentation materials and document files used in the briefing session.

[0974] "Question and answer data" is text data of questions asked during the briefing session and the answers to those questions.

[0975] "Means for converting to text data" refers to a system or algorithm that analyzes audio data and makes it usable as text information.

[0976] "Preprocessing means" refers to means for standardizing the formats of collected document data and question and answer data and deleting unnecessary parts.

[0977] A "generative AI model" is a type of artificial intelligence that automatically generates new training content based on given data.

[0978] "Training Content" means learning materials for end users generated by a generative AI model.

[0979] "Means of converting to video format" refers to the process of converting the generated training content into visual and audio materials.

[0980] A "learning management system" is an online platform where training content can be uploaded and accessed by end users for learning.

[0981] "Sensors mounted on robots" are devices such as cameras and microphones attached to robots, and are used to collect emotional data from end users.

[0982] "Emotional data" refers to psychological reaction information obtained from the end user's facial expressions and tone of voice.

[0983] The "means for evaluating and adjusting" refers to a means for appropriately correcting or providing additional explanations to pre-generated training content based on the collected emotional data.

[0984] MODE FOR CARRYING OUT THE INVENTION

[0985] As a specific embodiment for carrying out the present invention, a system will be described that automatically generates training content after an information session and adjusts the content based on emotional data of end users.

[0986] Overall system configuration

[0987] This system operates in the following manner.

[0988] Audio data collection

[0989] The server collects audio data from the briefing session. Specifically, it specifies the path to the audio file recorded during the briefing session, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[0990] Collection of document data and question and answer data

[0991] The server collects presentation materials, documents, and question and answer data used in the briefing. These data are also loaded by specifying the file path. The material data and question and answer data are stored in memory in text format.

[0992] Automatic Speech Recognition (ASR) System

[0993] The server uses an automatic speech recognition (ASR) system to convert the collected voice data into text data using the Google Cloud Speech-to-Text API, which makes the content of the briefing available as a transcript.

[0994] Preprocessing of document data and Q&A data

[0995] The server standardizes the format of the document data and deletes unnecessary parts, and organizes the questions and answers in the question and answer data to make it easier to use.

[0996] Generating training content using generative AI models

[0997] Based on the preprocessed data, the server generates training content using a generative AI model such as GPT-4. The preprocessed audio text, document data, and Q&A data are combined and fed into the generative AI model as input. The generative AI model then analyzes this data and automatically generates text content for the training.

[0998] Convert to video format

[0999] Based on the generated text content, the server generates a visually and audibly comprehensible video file, including adding corresponding narration to the text and incorporating slides and animations. This process uses OpenCV and FFmpeg.

[1000] Upload to Learning Management System (LMS)

[1001] The server uploads the generated videos to a learning management system, which is an online platform accessed by employees, allowing users to view the training content at any time.

[1002] Collecting Emotional Data

[1003] Sensors (camera and microphone) installed on the robot are used to collect emotional data while the end user is watching the training video. The facial expression data is analyzed using the Microsoft Azure Face API, and emotions are recognized from the tone of voice.

[1004] Evaluate and adjust training content

[1005] The server evaluates the training content based on the collected emotional data and, if necessary, includes sections in the video for additional explanation or adjusts the playback speed.

[1006] Specific examples

[1007] For example, when a briefing on new maintenance procedures is held, the server collects the audio data of the briefing, along with the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also compiled. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed.

[1008] Prompt Sentence Examples

[1009] New maintenance procedure briefing text:

[1010] 1. Engine disassembly procedure

[1011] 2. Parts inspection items

[1012] 3. Lubrication method

[1013] 4. Vibration measurement method

[1014] Q&A:

[1015] Q: What type of lubricant?

[1016] A: We will use ABC lubricant from XYZ company.

[1017] Use this information to generate training content that includes specific operating procedures and precautions, from engine disassembly procedures to lubrication methods.

[1018] The system makes it possible to adjust training content in real time, taking into account the end user's level of understanding and emotions, improving maintenance efficiency within the factory and ensuring that new procedures are understood quickly and reliably.

[1019] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1020] Step 1:

[1021] The server collects the audio data of the briefing. Specifically, it specifies the path of the audio file recorded during the briefing, opens the file, and reads the data. The input is the recorded audio file, and the output is binary audio data. The audio data is stored in buffer memory.

[1022] Step 2:

[1023] The server collects material data and question and answer data. The material data is the presentation materials and document format files used in the briefing, and the question and answer data is text data of questions and answers asked during the briefing. The input is the path of the material file and the question and answer file, and the output is stored in memory in text format.

[1024] Step 3:

[1025] The server converts the voice data into text data. An automatic speech recognition (ASR) system, in this case the Google Cloud Speech-to-Text API, is used to analyze the voice data and convert it into text information. The input is binary voice data, and the output is text data.

[1026] Step 4:

[1027] The server preprocesses the document data and Q&A data. Specifically, it standardizes the format of the document data, deletes unnecessary parts, and organizes the questions and answers in the Q&A data. The input is text-formatted document data and Q&A data, and the output is the preprocessed unified data.

[1028] Step 5:

[1029] Based on the preprocessed data, the server generates training content using a generative AI model. Specifically, the preprocessed audio text, material data, and Q&A data are integrated into one and supplied to the generative AI model (GPT-4). The input is the integrated preprocessed data, and the output is the generated training text content.

[1030] Step 6:

[1031] The server converts the generated training content into video format. Using OpenCV and FFmpeg, it adds narration corresponding to the text and incorporates slides and animations to generate the video. The input is the generated text content, and the output is a video file that is visually and audibly easy to understand.

[1032] Step 7:

[1033] The server uploads the generated videos to a learning management system (LMS). The LMS is an online platform accessed by employees, where the generated training videos are uploaded so that users can view the training content at any time. The input is the generated video file, and the output is the training content uploaded to the LMS.

[1034] Step 8:

[1035] A user uses a device to access a learning management system (LMS) and watch a provided training video. The user logs in to the LMS, selects the target training video, and plays it. The input is the login information for the LMS and the target training video data, and the output is the training video that is played.

[1036] Step 9:

[1037] The robot collects emotional data from the user using sensors (camera and microphone) installed on the device. It uses the Microsoft Azure Face API to analyze the facial expression data and recognize emotions from the tone of voice. The input is video and audio data, and the output is analyzed emotional data.

[1038] Step 10:

[1039] The server evaluates and adjusts the training content based on the collected emotional data, including adding sections to the video for additional explanations or adjusting the playback speed as needed. The input is emotional data, and the output is the adjusted training content.

[1040] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1041] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1042] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1043] [Fourth embodiment]

[1044] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1045] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1046] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1047] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1048] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1049] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1050] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1051] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1052] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1053] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1054] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1055] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1056] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1057] As a specific embodiment of this invention, we will explain a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[1058] Program processing and natural language explanation

[1059] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded at the briefing and reads the data. This audio data is stored in memory in binary format.

[1060] 2. The server collects the material data and question and answer data. The material data corresponds to the presentation materials and documents used in the briefing, and the question and answer data is a record of the questions and answers exchanged between the participants and the lecturer during the briefing. These data are also read by specifying a file path and stored in memory in text format.

[1061] 3. The server converts the voice data into text data using automatic speech recognition (ASR) technology. The converted text data makes the contents of the briefing available as text information.

[1062] 4. The server preprocesses the document data and Q&A data. The format of the document data is standardized and noise and unnecessary information is removed. The Q&A data is processed so that it can be easily organized into a Q&A format.

[1063] 5. The server generates training content using a generative AI model based on the preprocessed data. Specifically, it combines text from the audio data, material data, and Q&A data into one, and creates text content for the training. The generative AI model analyzes this input data and generates content suitable for specific purposes.

[1064] 6. The server converts the generated training content into a video format, converting the generated text content into a visually and audibly appealing format and saving it as a video file.

[1065] 7. Finally, the server uploads the generated training video to the learning management system (LMS). This allows users to access and watch the training video from their devices. Users can log in to the LMS, select the desired video from the list of video content provided, and watch it.

[1066] Specific examples

[1067] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, as well as the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also organized. This preprocessed data is then integrated into one, and training content is created using a generative AI model. Finally, the generated content is converted into video format and uploaded to the LMS. Users can access the LMS and watch the training video to quickly understand how to use the new customer management system.

[1068] The processing flow will be explained below.

[1069] Step 1:

[1070] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[1071] Step 2:

[1072] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[1073] Step 3:

[1074] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[1075] Step 4:

[1076] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers into an unambiguous format to make them easier to use.

[1077] Step 5:

[1078] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[1079] Step 6:

[1080] The server converts the generated training content into a video format, and generates a video file based on the generated text content that is visually and audibly easy to understand, for example by adding narration corresponding to the text and incorporating slides and animations.

[1081] Step 7:

[1082] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, allowing users to view the training content at any time.

[1083] Step 8:

[1084] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[1085] Example 1

[1086] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1087] In the conventional training content creation process, it takes a lot of time and effort to manually organize and integrate audio data, materials, and Q&A data from information sessions into training content. It is also difficult to consistently provide appropriate information, resulting in inconsistent training quality. The present invention aims to automate data processing and training content generation after information sessions, thereby providing efficient, high-quality training content.

[1088] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1089] In this invention, the server includes means for acquiring audio data of the briefing session, means for acquiring material data and Q&A data, means for converting the acquired audio data into text data, means for preprocessing the acquired material data and Q&A data, means for generating training information using an artificial intelligence model based on the preprocessed data, means for converting the generated training information into visual and audio media, and means for transmitting the generated visual and audio media to a learning management system. This automates data processing after the briefing session and enables the rapid provision of unified, high-quality training content.

[1090] "Audio data from a briefing session" refers to a digital audio file that records what was said at a briefing session, lecture, etc.

[1091] "Material data" refers to presentation materials and related document files used during the briefing session.

[1092] "Question and answer data" refers to records of questions and answers exchanged between participants and instructors during the information session.

[1093] The "means for converting into text data" is an automatic speech recognition technology for extracting voice data as text information.

[1094] "Preprocessing means" refers to procedures for arranging collected document data and Q&A data into a unified format and removing noise and unnecessary information.

[1095] An "artificial intelligence model" is a machine learning algorithm that generates appropriate output based on input data.

[1096] "Training information" is educational content that organizes the contents of the information session and provides them in a format suitable for learning purposes.

[1097] "Visual and audio media" refers to video files and other multimedia formats that contain text, image, and audio data.

[1098] A "learning management system" is an information management system that centrally manages training content provided by educational institutions and companies, and controls access and progress.

[1099] The "means for obtaining" is a method or device for reading data using a specified data file path.

[1100] "Transmitting means" refers to the data transfer method or communication means for uploading the generated training content to the learning management system.

[1101] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Specific embodiments are described below.

[1102] The server first collects the audio data of the briefing. During this process, the data is read into memory in binary format using a pre-specified audio file path (e.g., " / data / meeting_audio.wav"). Next, the server collects the presentation data (e.g., " / data / presentation.pdf") and the Q&A data (e.g., " / data / Q&A.txt"). These data are read from the specified paths and stored in memory in the form of a PDF file and a text file, respectively.

[1103] The server then converts the audio data into text using automatic speech recognition (ASR) technology, such as the Google Cloud Speech-to-Text API, which converts the audio data into text information and makes the contents of the briefing available in text format.

[1104] The server then preprocesses the document data and Q&A data. The document data is converted into a unified format and unnecessary information and noise are removed. The Q&A data is then processed to make it easier to organize into a Q&A format. For example, timestamps and unnecessary background information are removed from the document data, and the Q&A data is labeled into question and answer pairs.

[1105] The server generates training content using a generative AI model based on the preprocessed data. For example, using OpenAI's GPT-3, it inputs the following prompt: "Please create training content based on the following materials, audio text, and Q&A data." Based on this prompt, appropriate text content is generated.

[1106] The generated training content is then converted into visual and audio media, using Adobe Premiere Pro, for example, to add text as subtitles and composite the video with relevant images and graphics, with the final video file saved in MP4 format or similar.

[1107] Finally, the server uploads the generated training videos to the learning management system. During this process, the video files are uploaded directly from the server using the LMS's API, and metadata (lecture title, description, etc.) is also registered. Users can then log in to the LMS from their devices and watch the generated training videos. This automates data processing after the briefing session, making it possible to quickly provide uniform, high-quality training content.

[1108] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1109] Step 1:

[1110] The server collects audio data from the briefing session.

[1111] Input: The path to the information session recording file (e.g., " / data / meeting_audio.wav").

[1112] Specific operation: The server reads the audio file in binary format from the specified path and stores it in memory.

[1113] Output: Binary format of the audio data.

[1114] Step 2:

[1115] The server collects the material data and the question and answer data.

[1116] Input: File path of presentation materials (e.g. " / data / presentation.pdf") and file path of Q&A recording (e.g. " / data / Q&A.txt").

[1117] Specific operation: The server reads the PDF file and text file from each path and stores them in memory as document data and Q&A data.

[1118] Output: Document data (PDF format) and Q&A data (text format).

[1119] Step 3:

[1120] The server converts the voice data into text data.

[1121] Input: The binary format of the audio data collected in step 1.

[1122] Specific operation: The server uses automatic speech recognition technology such as the Google Cloud Speech-to-Text API to extract the voice data as text information.

[1123] Output: Text data converted from audio data.

[1124] Step 4:

[1125] The server pre-processes the document data and the question and answer data.

[1126] Input: Document data and Q&A data collected in Step 2.

[1127] Specific operation: The server converts the document data into a unified format, removes noise and unnecessary information, and processes the Q&A data so that it can be easily organized into a Q&A format.

[1128] Output: Preprocessed material data and Q&A data.

[1129] Step 5:

[1130] The server generates training content using a generative AI model based on the preprocessed data.

[1131] Input: Text data obtained in step 3, preprocessed document data and Q&A data in step 4.

[1132] Specific operation: The server uses a generative AI model (e.g., OpenAI's GPT-3) and inputs a prompt such as: "Create training content based on the following materials, audio text, and question and answer data."

[1133] Output: Generated training text content.

[1134] Step 6:

[1135] The server converts the generated training content into a video format.

[1136] Input: The training text content generated in step 5.

[1137] What it does: The server converts the generated text content into visual and auditory media and saves it as a video file. Using software such as Adobe Premiere Pro, the text is added as subtitles and composited into a video with associated images and graphics.

[1138] Output: Training video file (e.g. MP4 format).

[1139] Step 7:

[1140] The server uploads the generated training videos to a learning management system.

[1141] Input: The training video file created in step 6.

[1142] Specific operation: The server uses the API of the LMS platform (e.g., Moodle) to upload the video file and also registers metadata (lecture title, description, etc.).

[1143] Output: Training videos uploaded to a learning management system.

[1144] (Application example 1)

[1145] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1146] As new equipment and technologies are frequently introduced in factories, it is necessary to provide quick and efficient training for workers on operation and maintenance. However, current training methods require instructors to coordinate schedules, prepare materials, and provide face-to-face explanations, which are costly and time-consuming. This has led to a demand for automated generation of training content and its effective use.

[1147] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1148] In this invention, the server includes a means for collecting audio data of the briefing session, a means for collecting material data and Q&A data, and a means for converting the collected audio data into text data, thereby enabling the server to generate training content using a generative AI model based on the preprocessed data, convert the generated training content into video format, upload the generated video to a learning management system, and provide the training content to visual and audio devices.

[1149] "Audio data from a briefing session" refers to data that digitally records the verbal information spoken during the briefing session.

[1150] "Material data" refers to the data of presentation materials and documents used at the briefing.

[1151] "Question and answer data" refers to recorded data of questions and answers exchanged between participants and the lecturer during the information session.

[1152] "Means for converting into text data" refers to the process of converting voice data into text information using natural language processing technology.

[1153] "Preprocessing" refers to operations such as standardizing the format of collected data and removing noise so that the data can be easily analyzed and integrated.

[1154] A "generative AI model" refers to an algorithm or program that uses machine learning and artificial intelligence techniques to generate new content based on input data.

[1155] "Training Content" refers collectively to information and materials provided for learning or training purposes.

[1156] "Means for converting to video format" refers to the technology and tools that convert generated content such as text and charts into video files.

[1157] A "Learning Management System (LMS)" is a system that manages and distributes electronic learning materials, and is a platform that allows users to manage their own progress while learning.

[1158] "Visual and auditory devices" refers to devices that allow users to view and listen to learning content, including smart glasses and head-mounted displays.

[1159] This invention relates to a system that automatically generates training content for learning about the operation and maintenance techniques of robots used in factories. This system collects audio data, document data, and question and answer data from briefing sessions, preprocesses them, and then automatically generates training content using a generative AI model.

[1160] Hardware and software used

[1161] The main hardware and software components for implementing this system are as follows:

[1162] 1. Smart glasses (e.g., Google Glass)

[1163] 2. Head-Mounted Display (HMD)

[1164] 3. Automatic Speech Recognition (ASR) technology (e.g., Google Speech-to-Text API)

[1165] 4. Generative AI models (e.g., OpenAI GPT-4)

[1166] 5. Video editing tools (e.g. Adobe Premiere Pro, FFmpeg)

[1167] 6. Learning Management System (LMS)

[1168] Overview of system processing

[1169] The server processes the data in the following steps and ultimately generates the training content.

[1170] 1. Data Collection:

[1171] The server collects audio data from the briefings held at the factory, along with the presentation materials and Q&A sessions. The collected audio data is recorded by microphones installed in the smart glasses or robots.

[1172] 2. Text conversion:

[1173] The audio data is converted into text using automatic speech recognition (ASR) technology, which makes the contents of the briefing available as text information.

[1174] 3. Data Preprocessing:

[1175] The document data and Q&A data are preprocessed. The format of the document data is standardized and unnecessary information is deleted. The Q&A data is also organized in Q&A format.

[1176] 4. Content Generation:

[1177] Based on the preprocessed data, a generative AI model is used to generate training content. Specifically, the generative AI model combines text from the audio data, material data, and question and answer data to create text content for training.

[1178] 5. Video conversion and uploading:

[1179] The generated text content is converted into a video format, which is then converted into a visually and audibly appealing format using a video editing tool and saved as a video file.

[1180] Finally, the generated training videos are uploaded to a learning management system (LMS), where users can access the LMS through smart glasses or HMDs and watch the videos to learn operation procedures and maintenance techniques.

[1181] Specific examples

[1182] For example, consider a training session on the introduction of a new welding robot. A worker wearing smart glasses records the session and simultaneously collects the presentation materials and Q&A. ASR converts the speech to text and organizes the material data into a unified format. The training content is then generated by inputting the following prompt sentences into a generative AI model:

[1183] Example prompt sentence:

[1184] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[1185] 1. Audio text: <Audio text data>

[1186] 2. Presentation materials: <Data in a unified format>

[1187] 3. Q&A: <Organized Q&A data>

[1188] The generated training text is then edited using a video editing tool and finally uploaded to the LMS, allowing workers to access the training video on their HMD and progress through their studies.

[1189] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1190] Step 1:

[1191] The server collects audio data from the briefing session. Audio is recorded through microphones installed in the smart glasses or robot. The collected audio data is stored in memory in binary format. The input is audio signal data, and the output is binary audio data.

[1192] Step 2:

[1193] The server collects material data and question and answer data. Material data is presentation materials and documents, and question and answer data is records of questions and answers. These data are transferred to the server via file paths and stored in memory in text format. The input is material and question and answer files, and the output is text format data.

[1194] Step 3:

[1195] The server converts the collected voice data into text data using automatic speech recognition (ASR) technology. It uses an automatic speech recognition system (e.g., Google Speech-to-Text API) to analyze the voice data and convert it into text data. In this step, the voice signal is converted into text information, and the specific content becomes available as text information. The input is binary voice data, and the output is text data.

[1196] Step 4:

[1197] The server preprocesses the collected document data and Q&A data. The document data is unified in format and unnecessary information is removed before being converted into text format. The Q&A data is similarly formatted so that questions and answers are clearly distinguished. The input is the raw document data and Q&A data, and the output is preprocessed text data.

[1198] Step 5:

[1199] The server generates training content using a generative AI model based on the preprocessed data. The following prompt sentence is input to the generative AI model (e.g., OpenAI GPT-4) to generate text content for training:

[1200] Example prompt sentence:

[1201] "This is training content on operating procedures and precautions for a new welding robot. Please use the information below to generate specific training content to be used on the factory floor.

[1202] 1. Audio text: <Audio text data>

[1203] 2. Presentation materials: <Data in a unified format>

[1204] 3. Q&A: <Organized Q&A data>

[1205] The input is preprocessed speech text, material data, and question and answer data, and the output is text content for training.

[1206] Step 6:

[1207] The server converts the generated training content into a video format. Using a video editing tool (e.g. Adobe Premiere Pro, FFmpeg), it converts the content into a video file with visual and auditory appeal. The input is text content, and the output is a video file.

[1208] Step 7:

[1209] The server uploads the generated training videos to a learning management system (LMS). At this time, the video files are registered in the system so that users can access the LMS and watch the videos. Users can access the LMS using smart glasses or a head-mounted display, watch the training videos, and progress with their studies. The input is the video file, and the output is the training video uploaded to the LMS.

[1210] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1211] As a specific embodiment of this invention, we will explain a system that combines an emotion engine with a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). This system collects audio data, material data, and question and answer data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model. Furthermore, this system uses an emotion engine that recognizes user emotions and utilizes emotion data to generate and update training content.

[1212] Program processing and natural language explanation

[1213] 1. The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[1214] 2. The server collects the material data and Q&A data. The material data is the presentation materials and documents used in the briefing, and the Q&A data is the content of the questions and answers asked during the briefing. These data are also loaded by specifying the file path. The material data and Q&A data are stored in memory in text format.

[1215] 3. The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the speech content is converted into text information. This text data makes the content of the briefing available in written form.

[1216] 4. The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and deleting unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use.

[1217] 5. The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are combined and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[1218] 6. The server converts the generated training content into a video format, generating a visually and audibly easy-to-understand video file based on the generated text content, for example by adding narration corresponding to the text and incorporating slides and animations.

[1219] 7. The server uploads the generated videos to a learning management system (LMS), which is an online platform accessed by employees. By uploading the generated training videos to this system, users can view the training content at any time.

[1220] 8. The user uses the device to access the learning management system (LMS). The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[1221] 9. The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice through a camera and microphone, and monitors changes in emotions in real time.

[1222] 10. The server evaluates the training content based on the collected emotional data. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video speed. It also uses the emotional data as feedback to improve the quality of future training content.

[1223] Specific examples

[1224] For example, suppose a briefing session is held to introduce a new customer management system. The server collects the audio data from the session, along with the presentation materials and Q&A sessions. The collected audio data is converted into text using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A sessions are also organized. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted accordingly based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed. This allows users to gain a deeper understanding of the new service and its operational flow, resulting in faster and more effective customer support.

[1225] The processing flow will be explained below.

[1226] Step 1:

[1227] The server collects the audio data from the briefing. Specifically, it specifies the path to the audio file recorded during the briefing, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[1228] Step 2:

[1229] The server collects material data and question and answer data. Material data is the presentation materials and documents used in the briefing, and question and answer data is the content of questions and answers asked during the briefing. These data are also loaded by specifying the file path. Material data and question and answer data are stored in memory in text format.

[1230] Step 3:

[1231] The server converts the voice data into text data. Using an automatic speech recognition (ASR) system, the collected voice data is analyzed and the voice content is converted into text information. This text data makes the content of the briefing available as text.

[1232] Step 4:

[1233] The server preprocesses the document data and Q&A data. Preprocessing of the document data includes standardizing the format and removing unnecessary parts. For the Q&A data, the server organizes the questions and answers to make them easier to use. For example, it standardizes the text to lowercase and removes unnecessary spaces and noise.

[1234] Step 5:

[1235] The server generates training content using a generative AI model based on the preprocessed data. The preprocessed audio text, document data, and Q&A data are integrated into one and supplied as input to the generative AI model. The generative AI model analyzes this data and automatically generates text content for the training.

[1236] Step 6:

[1237] The server converts the generated training content into a video format. Based on the generated text content, it generates a video file that is visually and audibly easy to understand. Specifically, it generates narration corresponding to the text and combines it with slides and animations to create a video file.

[1238] Step 7:

[1239] The server uploads the generated videos to a learning management system (LMS), an online platform accessed by employees, and uploads the generated training videos to this system, allowing users to view the training content at any time.

[1240] Step 8:

[1241] The user accesses the learning management system (LMS) using a device. The user logs in to the LMS, selects the desired video from the list of training videos provided, and presses the play button to watch the training video. This allows the user to quickly understand the new service and its operational flow.

[1242] Step 9:

[1243] The server uses an emotion engine that recognizes the user's emotions to collect emotional data of the user while watching the training video. The emotion engine monitors the user's facial expressions and tone of voice via a camera and microphone, and monitors changes in emotions in real time.

[1244] Step 10:

[1245] The server evaluates the training content based on the collected emotional data and makes necessary adjustments. For example, if the user is confused, it generates additional explanations to make the content easier or adjusts the video playback speed. The emotional data can also be used as feedback to improve the quality of future training content.

[1246] Example 2

[1247] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1248] Conventional training content generation systems have difficulty efficiently collecting and processing the contents of information sessions and automatically generating training content. It is also difficult to generate and update content that reflects changes in users' understanding and emotions in real time. This results in a large amount of manual work being required to provide effective training, resulting in reduced productivity.

[1249] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1250] In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and question-and-answer data, means for converting the collected audio data into text data, means for preprocessing the collected material data and question-and-answer data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for recognizing user emotions, and means for evaluating and updating the training content based on the recognized emotion data. This makes it possible to efficiently collect the contents of the briefing session and dynamically generate and update training content according to the user's level of understanding.

[1251] "Audio Data" refers to the content of the audio files recorded during the briefing.

[1252] "Material data" refers to presentation materials and documents used at the briefing.

[1253] "Question and answer data" refers to the content of questions and answers asked during the briefing session.

[1254] "Text data" refers to collected voice data converted into text information using an automatic speech recognition system.

[1255] "Preprocessing" refers to the process of standardizing the format of document data, deleting unnecessary parts, and organizing question and answer data.

[1256] "Generative AI model" refers to artificial intelligence techniques that automatically generate new training content using collected and pre-processed data.

[1257] "Video format" refers to the conversion of generated training text content into a visually and audibly appealing video file.

[1258] "Learning Management System (LMS)" means an online platform through which employees access and learn training content via the internet.

[1259] "Emotion data" refers to data that records changes in emotions recognized by the emotion engine from the user's facial expressions and tone of voice.

[1260] "Evaluate and update" refers to the process of judging the quality of training content based on sentiment data and improving and adjusting the content as needed.

[1261] MODE FOR CARRYING OUT THE INVENTION

[1262] This invention is a system that automatically generates training content after a briefing session and uploads it to a learning management system (LMS). It also has the ability to recognize user emotions and reflect them in the generation and updating of training content. This system collects audio data, materials, and Q&A data from the briefing session, processes them appropriately, and then automatically generates training content using a generative AI model.

[1263] Audio data collection and conversion

[1264] The server collects the audio data of the briefing and stores it in memory in binary format. It then converts the collected audio data into text data using an automatic speech recognition (ASR) system, such as the Google Cloud Speech-to-Text API, making the audio content available as text information.

[1265] Collection and preprocessing of document data and Q&A data

[1266] The server loads the presentation materials, documents, and Q&A data used in the briefing into memory in text format. Preprocessing of the material data includes standardizing the format and deleting unnecessary parts, and for the Q&A data, it organizes the questions and answers and converts them into a more user-friendly format.

[1267] Training content generation and video conversion

[1268] The server integrates the pre-processed data and feeds it to a generative AI model to automatically generate training content. For example, the following prompt sentences can be used to input the generative AI model:

[1269] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[1270] The generated training content is converted into a video file that appeals to both the visual and auditory senses, incorporating narration, slides, animations, and other elements to create a visually easy-to-understand format.

[1271] Upload to a learning management system

[1272] The generated training content is converted into video format and uploaded to a learning management system (LMS), where users can access and watch the training videos.

[1273] User emotion recognition and content update

[1274] The server uses an emotion engine to recognize the user's emotions while watching the training video. It monitors the user's facial expressions and tone of voice through a camera and microphone to collect emotion data in real time. Based on this emotion data, the quality of the training content can be evaluated and updated as necessary. For example, if the user is confused, additional explanations can be generated or the video playback speed can be adjusted.

[1275] Specific examples

[1276] For example, when an introduction briefing for a new customer management system is held, the server collects the audio data from the briefing, as well as the presentation materials and Q&A sessions used. The audio data is converted into text data by the ASR system, the material data is organized into a unified format, and the Q&A data is also organized. Training content is then created using a generative AI model based on the preprocessed data. The generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch this training video. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. This allows users to gain a deeper understanding of the new system and enables faster and more effective customer service.

[1277] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1278] Step 1:

[1279] Audio data collection

[1280] The server receives the file path of the audio data of the briefing as input, opens the file, and reads the data into memory in binary format. This process saves the collected audio data in memory. Specifically, it opens the specified audio file and reads the contents in binary format.

[1281] Step 2:

[1282] Collection of document data and Q&A data

[1283] The server receives the file paths for the presentation data and Q&A data as input, opens each file, and reads the data into memory in text format. This process saves the presentation data and Q&A data in memory. Specifically, the presentation materials and Q&A content are read as text data.

[1284] Step 3:

[1285] Converting audio data to text

[1286] The server receives the collected voice data as input and converts it into text data using an automatic speech recognition (ASR) system. This process makes the voice content available as text information. Specifically, the voice data is input into the ASR system, and text information is obtained as output.

[1287] Step 4:

[1288] Preprocessing of document data and Q&A data

[1289] The server receives the collected document data and Q&A data as input, standardizes the data format, deletes unnecessary parts, and organizes the questions and answers. This process results in preprocessed data. Specifically, unnecessary parts of the text are deleted and the format is adjusted.

[1290] Step 5:

[1291] Training content generation

[1292] The server integrates the preprocessed speech-to-text data, the materials data, and the Q&A data, and inputs them into the generative AI model along with prompts. This process generates training content. Specifically, the following prompts are used:

[1293] "Generate training content based on the following data:\nAudio-text data, materials data, Q&A data"

[1294] Step 6:

[1295] Content video conversion

[1296] The server receives the generated training content as input and converts it into a visually and audibly appealing video file. This process results in a training video containing narration, slides, and animations. Specifically, the server creates slides with narration and animations based on the generated text content and converts them into a video format.

[1297] Step 7:

[1298] Upload to LMS

[1299] The server receives the generated video file as input and uploads it to the learning management system (LMS). This process makes it accessible to users. Specifically, the video file is sent to the server using the LMS's upload function.

[1300] Step 8:

[1301] User LMS Access

[1302] Users access the LMS using their devices and select and watch the training videos provided. This process allows users to learn the training content. Specifically, users log in to the LMS, select the desired video from the video list, and press the play button.

[1303] Step 9:

[1304] emotion recognition

[1305] The server monitors the user's facial expressions and tone of voice while watching the training video through a camera and microphone, and collects emotional data using an emotion engine. This process yields the user's emotional data. Specifically, it analyzes changes in facial expressions and voice while watching in real time and records changes in emotions.

[1306] Step 10:

[1307] Evaluating and updating training content

[1308] The server receives the collected emotional data as input, evaluates the training content, and updates it as necessary to improve the quality of the training. Specifically, it generates additional explanations or adjusts the video playback speed depending on the user's emotional state.

[1309] (Application example 2)

[1310] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1311] Conventional training content generation systems generate training content based on recordings of information sessions and materials, but because they do not take into account the level of understanding and emotions of individual end users, the quality of the training content is uniform and may not meet the needs of end users.In addition, the generated content is fixed and difficult to adjust in real time, which creates the problem of not being able to maximize the learning effect of end users.

[1312] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting audio data of the briefing session, means for collecting material data and Q&A data, means for converting the collected audio data into text data, means for preprocessing the collected material data and Q&A data, means for generating training content using a generative AI model based on the preprocessed data, means for converting the generated training content into video format, means for uploading the generated video to a learning management system, means for collecting end-user emotional data using sensors mounted on the robot, and means for evaluating and adjusting the training content based on the collected emotional data. This allows the training content to be adjusted in real time taking into account the end-user's level of understanding and emotions, enabling advanced training tailored to individual needs.

[1313] "Audio data from the briefing session" refers to an audio file that records what was said during the briefing session.

[1314] "Material data" refers to presentation materials and document files used in the briefing session.

[1315] "Question and answer data" is text data of questions asked during the briefing session and the answers to those questions.

[1316] "Means for converting to text data" refers to a system or algorithm that analyzes audio data and makes it usable as text information.

[1317] "Preprocessing means" refers to means for standardizing the formats of collected document data and question and answer data and deleting unnecessary parts.

[1318] A "generative AI model" is a type of artificial intelligence that automatically generates new training content based on given data.

[1319] "Training Content" means learning materials for end users generated by a generative AI model.

[1320] "Means of converting to video format" refers to the process of converting the generated training content into visual and audio materials.

[1321] A "learning management system" is an online platform where training content can be uploaded and accessed by end users for learning.

[1322] "Sensors mounted on robots" are devices such as cameras and microphones attached to robots, and are used to collect emotional data from end users.

[1323] "Emotional data" refers to psychological reaction information obtained from the end user's facial expressions and tone of voice.

[1324] The "means for evaluating and adjusting" refers to a means for appropriately correcting or providing additional explanations to pre-generated training content based on the collected emotional data.

[1325] MODE FOR CARRYING OUT THE INVENTION

[1326] As a specific embodiment for carrying out the present invention, a system will be described that automatically generates training content after an information session and adjusts the content based on emotional data of end users.

[1327] Overall system configuration

[1328] This system operates in the following manner.

[1329] Audio data collection

[1330] The server collects audio data from the briefing session. Specifically, it specifies the path to the audio file recorded during the briefing session, opens the file, and reads the data. The read audio data is stored in memory in binary format.

[1331] Collection of document data and question and answer data

[1332] The server collects presentation materials, documents, and question and answer data used in the briefing. These data are also loaded by specifying the file path. The material data and question and answer data are stored in memory in text format.

[1333] Automatic Speech Recognition (ASR) System

[1334] The server uses an automatic speech recognition (ASR) system to convert the collected voice data into text data using the Google Cloud Speech-to-Text API, which makes the content of the briefing available as a transcript.

[1335] Preprocessing of document data and Q&A data

[1336] The server standardizes the format of the document data and deletes unnecessary parts, and organizes the questions and answers in the question and answer data to make it easier to use.

[1337] Generating training content using generative AI models

[1338] Based on the preprocessed data, the server generates training content using a generative AI model such as GPT-4. The preprocessed audio text, document data, and Q&A data are combined and fed into the generative AI model as input. The generative AI model then analyzes this data and automatically generates text content for the training.

[1339] Convert to video format

[1340] Based on the generated text content, the server generates a visually and audibly comprehensible video file, including adding corresponding narration to the text and incorporating slides and animations. This process uses OpenCV and FFmpeg.

[1341] Upload to Learning Management System (LMS)

[1342] The server uploads the generated videos to a learning management system, which is an online platform accessed by employees, allowing users to view the training content at any time.

[1343] Collecting Emotional Data

[1344] Sensors (camera and microphone) installed on the robot are used to collect emotional data while the end user is watching the training video. The facial expression data is analyzed using the Microsoft Azure Face API, and emotions are recognized from the tone of voice.

[1345] Evaluate and adjust training content

[1346] The server evaluates the training content based on the collected emotional data and, if necessary, includes sections in the video for additional explanation or adjusts the playback speed.

[1347] Specific examples

[1348] For example, when a briefing on new maintenance procedures is held, the server collects the audio data of the briefing, along with the presentation materials and Q&A sessions used. The collected audio data is converted into text data using an automatic speech recognition system, the material data is organized into a unified format, and the Q&A data is also compiled. Training content is then created using a generative AI model based on the preprocessed data. This generated content is converted into video format and uploaded to the LMS. Users access the LMS to watch the training videos. Furthermore, an emotion engine monitors the user's emotions while watching, and the training content is adjusted appropriately based on that data. For example, if the user is confused, the system can include a section in the video that provides additional explanations or adjust the playback speed.

[1349] Prompt Sentence Examples

[1350] New maintenance procedure briefing text:

[1351] 1. Engine disassembly procedure

[1352] 2. Parts inspection items

[1353] 3. Lubrication method

[1354] 4. Vibration measurement method

[1355] Q&A:

[1356] Q: What type of lubricant?

[1357] A: We will use ABC lubricant from XYZ company.

[1358] Use this information to generate training content that includes specific operating procedures and precautions, from engine disassembly procedures to lubrication methods.

[1359] The system makes it possible to adjust training content in real time, taking into account the end user's level of understanding and emotions, improving maintenance efficiency within the factory and ensuring that new procedures are understood quickly and reliably.

[1360] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1361] Step 1:

[1362] The server collects the audio data of the briefing. Specifically, it specifies the path of the audio file recorded during the briefing, opens the file, and reads the data. The input is the recorded audio file, and the output is binary audio data. The audio data is stored in buffer memory.

[1363] Step 2:

[1364] The server collects material data and question and answer data. The material data is the presentation materials and document format files used in the briefing, and the question and answer data is text data of questions and answers asked during the briefing. The input is the path of the material file and the question and answer file, and the output is stored in memory in text format.

[1365] Step 3:

[1366] The server converts the voice data into text data. An automatic speech recognition (ASR) system, in this case the Google Cloud Speech-to-Text API, is used to analyze the voice data and convert it into text information. The input is binary voice data, and the output is text data.

[1367] Step 4:

[1368] The server preprocesses the document data and Q&A data. Specifically, it standardizes the format of the document data, deletes unnecessary parts, and organizes the questions and answers in the Q&A data. The input is text-formatted document data and Q&A data, and the output is the preprocessed unified data.

[1369] Step 5:

[1370] Based on the preprocessed data, the server generates training content using a generative AI model. Specifically, the preprocessed audio text, material data, and Q&A data are integrated into one and supplied to the generative AI model (GPT-4). The input is the integrated preprocessed data, and the output is the generated training text content.

[1371] Step 6:

[1372] The server converts the generated training content into video format. Using OpenCV and FFmpeg, it adds narration corresponding to the text and incorporates slides and animations to generate the video. The input is the generated text content, and the output is a video file that is visually and audibly easy to understand.

[1373] Step 7:

[1374] The server uploads the generated videos to a learning management system (LMS). The LMS is an online platform accessed by employees, where the generated training videos are uploaded so that users can view the training content at any time. The input is the generated video file, and the output is the training content uploaded to the LMS.

[1375] Step 8:

[1376] A user uses a device to access a learning management system (LMS) and watch a provided training video. The user logs in to the LMS, selects the target training video, and plays it. The input is the login information for the LMS and the target training video data, and the output is the training video that is played.

[1377] Step 9:

[1378] The robot collects emotional data from the user using sensors (camera and microphone) installed on the device. It uses the Microsoft Azure Face API to analyze the facial expression data and recognize emotions from the tone of voice. The input is video and audio data, and the output is analyzed emotional data.

[1379] Step 10:

[1380] The server evaluates and adjusts the training content based on the collected emotional data, including adding sections to the video for additional explanations or adjusting the playback speed as needed. The input is emotional data, and the output is the adjusted training content.

[1381] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1382] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1383] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1384] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1385] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1386] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1387] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1388] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1389] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1390] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1391] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1392] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1393] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1394] 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.

[1395] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1396] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1397] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1398] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1399] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1400] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1401] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1402] The following is further disclosed regarding the above embodiment.

[1403] (Claim 1)

[1404] A means for collecting audio data from the briefing session;

[1405] a means of collecting documentary data and question and answer data;

[1406] A means for converting the collected voice data into text data;

[1407] A means for preprocessing the collected document data and question and answer data;

[1408] A means for generating training content using a generative AI model based on preprocessed data;

[1409] a means for converting the generated training content into a video format;

[1410] a means for uploading the generated videos to a learning management system;

[1411] A system including:

[1412] (Claim 2)

[1413] 2. The system according to claim 1, further comprising means for integrating the preprocessed speech text, the material data, and the question and answer data to generate a single training content.

[1414] (Claim 3)

[1415] 2. The system according to claim 1, further comprising means for updating the generated training content based on the content of questions and answers at the information session.

[1416] "Example 1"

[1417] (Claim 1)

[1418] A means for acquiring audio data of the briefing session;

[1419] a means for acquiring document data and question and answer data;

[1420] A means for converting the acquired voice data into text data;

[1421] A means for preprocessing the acquired document data and question and answer data;

[1422] A means for generating training information using an artificial intelligence model based on the preprocessed data;

[1423] a means for converting the generated training information into visual and audio media;

[1424] means for transmitting the generated visual and audio media to a learning management system;

[1425] A system including:

[1426] (Claim 2)

[1427] 2. The system according to claim 1, further comprising means for integrating the preprocessed speech text, the material data, and the question and answer data to generate a single training information.

[1428] (Claim 3)

[1429] 2. The system according to claim 1, further comprising means for updating the generated training information based on the content of questions and answers at the information session.

[1430] "Application Example 1"

[1431] (Claim 1)

[1432] A means for collecting audio data from the briefing session;

[1433] a means of collecting documentary data and question and answer data;

[1434] A means for converting the collected voice data into text data;

[1435] A means for preprocessing the collected document data and question and answer data;

[1436] A means for generating training content using a generative AI model based on preprocessed data;

[1437] a means for converting the generated training content into a video format;

[1438] a means for uploading the generated videos to a learning management system;

[1439] a means for providing training content to visual and audio devices;

[1440] A system including:

[1441] (Claim 2)

[1442] 2. The system according to claim 1, further comprising means for integrating the preprocessed speech text, the material data, and the question and answer data to generate a single training content.

[1443] (Claim 3)

[1444] 2. The system according to claim 1, further comprising means for updating the generated training content based on the content of questions and answers at the information session.

[1445] "Example 2: Combining Emotion Engines"

[1446] (Claim 1)

[1447] A means for collecting audio data from the briefing session;

[1448] a means of collecting documentary data and question and answer data;

[1449] A means for converting the collected voice data into text data;

[1450] A means for preprocessing the collected document data and question and answer data;

[1451] A means for generating training content using a generative AI model based on preprocessed data;

[1452] a means for converting the generated training content into a video format;

[1453] a means for uploading the generated videos to a learning management system;

[1454] means for recognizing a user's emotion;

[1455] a means for evaluating and updating training content based on the recognized emotion data;

[1456] A system including:

[1457] (Claim 2)

[1458] 2. The system according to claim 1, further comprising means for integrating the preprocessed speech text, the material data, and the question and answer data to generate a single training content.

[1459] (Claim 3)

[1460] 2. The system according to claim 1, further comprising means for updating the generated training content based on the content of questions and answers at the information session and user emotion data.

[1461] "Application example 2 when combining emotion engines"

[1462] (Claim 1)

[1463] A means for collecting audio data from the briefing session;

[1464] a means of collecting documentary data and question and answer data;

[1465] A means for converting the collected voice data into text data;

[1466] A means for preprocessing the collected document data and question and answer data;

[1467] A means for generating training content using a generative AI model based on preprocessed data;

[1468] a means for converting the generated training content into a video format;

[1469] a means for uploading the generated videos to a learning management system;

[1470] means for collecting end-user emotion data using sensors mounted on the robot;

[1471] a means for evaluating and adjusting training content based on the collected emotional data;

[1472] A system including:

[1473] (Claim 2)

[1474] 10. The system of claim 1, further comprising means for generating a training content using a generative AI model based on the preprocessed data.

[1475] (Claim 3)

[1476] 10. The system of claim 1, further comprising means for updating the generated training content based on the content of questions and answers in the information session and emotion data of the end user. [Explanation of symbols]

[1477] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for collecting audio data from the briefing session; a means of collecting documentary data and question and answer data; A means for converting the collected voice data into text data; A means for preprocessing the collected document data and question and answer data; A means for generating training content using a generative AI model based on preprocessed data; a means for converting the generated training content into a video format; a means for uploading the generated videos to a learning management system; A system including:

2. The system according to claim 1, further comprising means for integrating the preprocessed speech text, the material data and the question and answer data to generate a single training content.

3. The system according to claim 1 , further comprising means for updating the generated training content based on the content of questions and answers at the information session.

Images