system

Abstract

Provide a system. 【Solution means】Image recognition means, Text extraction means, Summary generation means, Output format selection means, Summary presentation means, Detailed information request means, Detailed information providing means, Detailed information presentation means, A system including the above.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] When a person with a disability such as blindness reads a long document, the task is a very heavy burden. In particular, it is difficult to handle a huge amount of information even in the case of Braille reading, voice synthesis, or sign language video. Therefore, means for efficiently grasping the content of the document are required.

Means for Solving the Problems

[0005] To solve this problem, the present invention provides the following means. First, an image recognition means captures an image of the document, and a text extraction means extracts the text from the image. Next, a summary generation means on the server analyzes the text and generates a summary. This summary is presented in one of the following formats by an output format selection means: braille display, speech synthesis, or sign language video. Furthermore, when the user specifies a section for which they want to know more details, a detailed information provision means retrieves detailed information from the server, and a detailed information presentation means presents it again in the desired format. This allows the user to efficiently grasp the entire document and easily check the details of the necessary parts.

[0006] "Image recognition means" refers to a device or program that captures images of documents or other materials and analyzes them as digital data.

[0007] "Text extraction means" refers to a device or program that extracts character information from an image using optical character recognition technology.

[0008] A "summary generation means" is a device or program that analyzes extracted text, extracts important information, and generates a summary.

[0009] "Output format selection means" refers to a device or program that outputs information in one of the following formats based on user settings: braille display, speech synthesis, or sign language video.

[0010] "Summary presentation means" refers to a device or program that presents summary information to the user in a selected output format.

[0011] "Detailed information request means" refers to a device or program that requests detailed information about a summary portion specified by the user.

[0012] "Means for providing detailed information" refers to a device or program that identifies and provides the requested detailed information.

[0013] "Detailed information presentation means" refers to a device or program that presents the provided detailed information to the user in a selected output format. [Brief explanation of the drawing]

[0014] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, when an emotion engine is combined. [Figure 14]It is a sequence diagram showing the processing flow of a data processing system in Application Example 2 when a sentiment engine is combined.

Mode for Carrying Out the Invention

[0015] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0018] In the following embodiments, a numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0019] In the following embodiments, a numbered storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, etc.

[0020] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0022] [First Embodiment]

[0023] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0024] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0025] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0027] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0030] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0034] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0035] This invention is a system for users with disabilities, such as blindness, to efficiently understand the contents of documents. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, and detailed information presentation means.

[0036] Program processing and explanation in natural language

[0037] 1. Image recognition and text extraction

[0038] The user takes a picture of a document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology.

[0039] 2. Summary generation

[0040] The terminal sends the extracted text to the server.

[0041] The server analyzes the received text, extracts important information using a summarization algorithm, and generates a summary.

[0042] 3. Selection of output format and presentation of summary

[0043] The terminal checks the user's settings and outputs a summary in formats such as Braille display, speech synthesis, and sign language video.

[0044] 4. Request for detailed information

[0045] The user can specify the summary section they are interested in using gestures or device operations.

[0046] The terminal requests detailed information about the specified part from the server.

[0047] 5. Providing detailed information

[0048] The server searches for the requested details and sends them to the terminal.

[0049] 6. Output of detailed information

[0050] The device then presents detailed information to the user again in a selected format, such as braille display, speech synthesis, or sign language video.

[0051] Specific example

[0052] 1. Specific examples of image recognition and text extraction

[0053] Users take pictures of documents using their device's camera in order to read reports they need for work.

[0054] The device sends the captured image to an OCR module, which then extracts the text from the report.

[0055] 2. Specific Examples of Summary Generation

[0056] The terminal sends the extracted report text to the server.

[0057] The server analyzes the text, extracts only the important points, and generates a summary.

[0058] The server sends the generated summary to the terminal.

[0059] 3. Specific examples of output format selection and summary presentation

[0060] The device will read a summary aloud in text-to-speech format based on the user's settings.

[0061] Users can listen to audio to confirm important parts of the report.

[0062] 4. Specific examples of requests for detailed information

[0063] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[0064] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[0065] 5. Specific examples of providing detailed information

[0066] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[0067] The server identifies the details in the second paragraph and sends them to the terminal.

[0068] 6. Specific examples of outputting detailed information

[0069] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[0070] Users can view more detailed information and gain a deeper understanding of the report's contents.

[0071] In this way, users can efficiently grasp the entire document and easily check the details of the parts they need. This system is extremely convenient for users with visual impairments.

[0072] The following describes the processing flow.

[0073] Step 1:

[0074] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[0075] Step 2:

[0076] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[0077] Step 3:

[0078] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[0079] Step 4:

[0080] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[0081] Step 5:

[0082] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[0083] Step 6:

[0084] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in one of the following formats: Braille display, speech synthesis, or sign language video.

[0085] Step 7:

[0086] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[0087] Step 8:

[0088] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[0089] Step 9:

[0090] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[0091] Step 10:

[0092] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[0093] Step 11:

[0094] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[0095] Step 12:

[0096] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[0097] This series of processing steps allows users to easily review a document summary and, if more detailed information is needed, efficiently retrieve specific details from particular sections.

[0098] (Example 1)

[0099] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0100] There is a need for a system that allows visually impaired users to efficiently understand the content of documents and easily access detailed information on the parts they need. However, conventional systems require a great deal of effort and time to fully understand the content of documents, which is a significant burden on users. In particular, there were technical challenges in providing summaries and detailed information in the required format.

[0101] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0102] In this invention, the server includes an image recognition means, an optical character recognition means, and a summary generation means. This makes it possible for a visually impaired user to take an image of a document, extract text from the image, efficiently summarize the extracted text, and present it to the user in an appropriate format.

[0103] "Image recognition means" refers to technology that identifies images of documents taken by the user using the camera on their device and extracts the necessary information.

[0104] "Optical character recognition means" refers to a technology that analyzes characters within a captured image and extracts them as digital text.

[0105] A "summary generation method" is a technology that analyzes extracted text, extracts important information, and generates a concise summary.

[0106] The "output format selection means" is a technology that outputs the generated summary in an appropriate format (e.g., audio, braille display, sign language video) based on the user's settings.

[0107] A "summary presentation method" is a technology that presents a summary to the user in a selected format.

[0108] A "means for requesting detailed information" is a technology that allows a user to specify detailed information about a summary section that interests them and communicate that request to the system.

[0109] "Means of providing detailed information" refers to a technology in which a server searches for and provides detailed information on a specified part based on a user's request.

[0110] A "means for presenting detailed information" refers to a technology that presents detailed information to the user in a selected output format.

[0111] Modes for carrying out the invention

[0112] This invention is a system for visually impaired users to efficiently understand the contents of documents. This system utilizes multiple hardware and software components.

[0113] 1. Image recognition and text extraction

[0114] The user takes a picture of the document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology. Software used includes "Google® Cloud Vision API" and "Adobe OCR".

[0115] Specific example: A user takes a picture of a document using their device's camera to read a report they need for work. The device sends the captured image to an OCR module, which extracts the text from the report.

[0116] 2. Summary generation

[0117] The terminal sends the extracted text to the server.

[0118] The server analyzes the received text, extracts key information using a summarization algorithm, and generates a summary. Suitable generative AI models to use include "BERT" and "GPT-4(registered trademark)".

[0119] Specific example: The terminal sends the extracted report text to the server. The server analyzes the text, extracts only the important points, and generates a summary. The server then sends the generated summary to the terminal.

[0120] 3. Selection of output format and presentation of summary

[0121] The device checks the user's settings and outputs a summary in formats such as Braille display, text-to-speech, and sign language video. Text-to-speech uses "Amazon Polly" or "Google Text-to-Speech."

[0122] Specific example: The device reads a summary aloud in text-to-speech format based on the user's settings. The user then reviews important parts of the report by voice.

[0123] 4. Request for detailed information

[0124] Users can specify the summary sections they are interested in using gestures or device operations.

[0125] The device requests detailed information about the specified area from the server. Gesture recognition uses technologies such as "Microsoft® Kinect" or "Leap Motion".

[0126] Specific example: A user listens to the audio, feels particularly interested in the second paragraph, and wants more detailed information. They then use a gesture to indicate their interest in the second paragraph to the device.

[0127] 5. Providing detailed information

[0128] The server searches for the requested details and sends them to the terminal. When the server generates additional information based on the text of the specified paragraph, it uses the "GPT-4" AI model.

[0129] Specific example: The terminal recognizes the user's action and requests detailed information about the specified second paragraph from the server. The server identifies the detailed information about the second paragraph and sends it to the terminal.

[0130] 6. Output of detailed information

[0131] The terminal then presents detailed information to the user again in a selected format, such as Braille display, speech synthesis, or sign language video.

[0132] Review the detailed information provided by the user.

[0133] Specific example: The device reads aloud the detailed information from the second paragraph using speech synthesis. The user can then review the more detailed information and gain a deeper understanding of the report's content.

[0134] Examples of prompt statements

[0135] "Extract text from documents photographed using Google Cloud Vision, and then summarize that text using GPT-4."

[0136] "Summarize the extracted text and use Amazon Polly to have it read aloud."

[0137] "Provide details about the paragraphs that the user is interested in, and then use Amazon Polly to convert those details into speech and present them."

[0138] In this way, the system provides a way for visually impaired users to efficiently grasp the entire document and easily check the details of the parts they need.

[0139] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0140] Program processing flow

[0141] Step 1:

[0142] The user takes a picture of a document using the device's camera. The input is an image of the document, and the output is image data saved on the device.

[0143] Specific action: The user presses the "Capture" button on the device.

[0144] Step 2:

[0145] The device takes a picture and passes it to an OCR module for processing. The input is image data, and the output is text data extracted by OCR.

[0146] Specific operation: The device sends image data to the "Optical Character Recognition API".

[0147] Step 3:

[0148] The terminal receives text from the OCR module and sends it to the server. The input is text data, and the output is a transfer to the server.

[0149] Specific action: The device uses an "HTTPS request" to send text data to the server.

[0150] Step 4:

[0151] The server parses the received text and executes a summary generation algorithm. The input is text data, and the output is summarized text.

[0152] Specific operation: The server generates a summary using "natural language processing technology".

[0153] Step 5:

[0154] The server generates a summary and sends it to the terminal. The input is the summary text, and the output is the transfer to the terminal.

[0155] Specific operation: The server sends summary data to the terminal via an "HTTPS response".

[0156] Step 6:

[0157] The terminal checks the user's settings and presents the summary in the selected format. Input is the summary text and user settings information, and output is in the form of audio, Braille display, sign language video, etc.

[0158] Specific action: The device executes the "Speech Synthesis API" or another selected output method.

[0159] Step 7:

[0160] The user reviews the provided summary and specifies the areas for which they want more detailed information. The input is an audio summary, and the output is a request for more detailed information.

[0161] Specific action: The user requests more information using gestures or input devices.

[0162] Step 8:

[0163] The terminal recognizes user input and requests detailed information about the specified portion from the server. The input is a request for detailed information, and the output is the transfer of the request for detailed information to the server.

[0164] Specific action: The terminal sends an "HTTPS request" and forwards a prompt message to the server requesting more information.

[0165] Step 9:

[0166] The server prepares the requested details and sends them to the terminal. The input is the request for details, and the output is the details.

[0167] Specific operation: The server generates the specified content based on "natural language processing technology" and sends it to the terminal.

[0168] Step 10:

[0169] The terminal then presents the detailed information to the user again in the selected format. The input is detailed information, and the output is in the form of audio, braille display, sign language video, etc.

[0170] Specific action: The device executes the "Speech Synthesis API" or another selected output method and presents detailed information.

[0171] keyword

[0172] Generative AI model, prompt sentence

[0173] (Application Example 1)

[0174] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0175] This solution addresses the challenge faced by visually impaired users who lack the efficient means to quickly and accurately grasp the contents of documents such as monitoring reports in real time, and to easily access detailed information as needed.

[0176] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0177] In this invention, the server includes image recognition means, text extraction means, summary generation means, and speech synthesis means. This makes it possible for visually impaired users to obtain summarized information from documents such as monitoring reports in audio format, and to be provided with even more detailed information in audio format as well.

[0178] "Image recognition means" refers to a technology that uses an image input device such as a camera to analyze captured image data and recognize necessary information.

[0179] "Text extraction means" refers to a technology that extracts character information from image data using optical character recognition (OCR) technology.

[0180] A "summary generation method" is a technology that analyzes extracted text data, extracts important information, and generates a concise summary.

[0181] "Output format selection means" refers to a technology that outputs the generated summary information in formats such as Braille display, speech synthesis, and sign language video, according to the user's settings and preferences.

[0182] A "summary presentation method" is a technology that presents summary information to the user based on the selected output format.

[0183] A "means for requesting detailed information" is a technology that allows a user to specify a part of a summary that interests them and request detailed information about that part.

[0184] "Means of providing detailed information" refers to technology that searches for requested detailed information and provides it to the user.

[0185] "Means for presenting detailed information" refers to technologies that present the provided detailed information to the user in the form of Braille displays, speech synthesis, sign language videos, etc.

[0186] "Speech synthesis means" refers to a technology that converts text data into speech and provides information in audio format.

[0187] "Voice recognition means" refers to technology that analyzes the user's voice input and recognizes its content.

[0188] This invention is a system for visually impaired users to efficiently understand the contents of documents such as papers and surveillance reports. Next, a specific embodiment of the system will be described.

[0189] This system consists of a user-facing terminal and multiple processing modules running on a server. The terminal is equipped with a camera, microphone, and speaker. The main processing includes the following:

[0190] 1. Image recognition and text extraction

[0191] The user takes a picture of a surveillance report or document using the device's camera. The device's camera app captures the image, and the image is converted into text using optical character recognition (OCR) technology. For example, software such as Tesseract OCR can be used.

[0192] 2. Sending text to the server and generating a summary

[0193] The extracted text data is sent to the server in real time. On the server, a generative AI model runs, using text analysis and natural language processing techniques to generate a summary of the key information. Natural language processing algorithms such as BERT are used.

[0194] 3. Selection and presentation of summary output format

[0195] The generated summary is provided as audio using speech synthesis technology, depending on the user's settings. Speech synthesis engines such as Google TTS are used. Users can listen to the summary through their speakers.

[0196] 4. Requesting and providing detailed information

[0197] If a user is interested in a particular summary, they can request more information using speech recognition technology. Speech recognition engines such as Sphinx or Google Speech Recognition are used. The requested information is sent back to the server, searched on the server, and then provided via speech synthesis.

[0198] The following are specific usage scenarios.

[0199] Specific usage scenarios:

[0200] Consider a scenario where a visually impaired security staff member is reviewing a surveillance report. The staff member uses their smartphone camera to photograph the report, and OCR technology extracts the text on the device. This text is sent to a server, where a generative AI model generates a summary. The summary is then provided to the staff member in audio format via a speech synthesis engine. If the staff member is interested in a particular paragraph, they can request further details by communicating this using speech recognition technology, and these details are also provided in audio format.

[0201] Examples of input prompts for a generative AI model:

[0202] Extract the key information from the following text and generate a summary. The theme is a monitoring report:

[0203] (Text from the monitoring report)

[0204] This system allows visually impaired users to quickly understand important information and obtain detailed information as needed.

[0205] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0206] Step 1:

[0207] The user takes a picture of the monitoring report with their camera. The device's camera app captures the image, and that image is converted into text using optical character recognition technology (such as Tesseract OCR). Specifically, the captured image data is input into an OCR module, and the text data is output.

[0208] Step 2:

[0209] The terminal sends the extracted text data to the server. Here, the text data is packetized in formats such as JSON or XML and sent to the server over the network. The server stores the received text data for analysis.

[0210] Step 3:

[0211] The server analyzes the received text data. First, it uses a natural language processing algorithm (such as a generative AI model like BERT) to extract important information from the text and generate a summary. This summary generation algorithm produces a compact text that extracts only the essential points while eliminating redundant parts.

[0212] Step 4:

[0213] The server generates a summary and sends it to the terminal. The summary text is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received summary data and continues processing based on the output format set by the user.

[0214] Step 5:

[0215] The device processes the received summary text into speech using a text-to-speech engine (such as Google TTS). When summary text is input, the text-to-speech engine converts the text into natural-sounding speech, which is then presented to the user through the speaker.

[0216] Step 6:

[0217] The user listens to an audio summary and requests more detailed information on the parts that interest them through a speech recognition system (such as Google Speech Recognition). Specifically, the user verbally indicates the parts they want to know more about, the microphone captures this, and it is input into the speech recognition engine. The speech recognition engine converts the audio into text, and that text is sent to the server.

[0218] Step 7:

[0219] The server receives a request for user details. The server searches the database for and extracts detailed text information for the requested portion. This extraction process generates detailed text data.

[0220] Step 8:

[0221] The server sends the extracted detailed information to the terminal. The detailed information is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received detailed information and prepares to present it in the user-configured output format.

[0222] Step 9:

[0223] The device receives detailed information and converts it into speech using a speech synthesis engine. Detailed information text is input, the speech synthesis engine converts that text into natural-sounding speech, and it is presented to the user through the speaker.

[0224] This series of processing steps allows visually impaired users to quickly and efficiently understand the contents of monitoring reports.

[0225] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0226] This invention is a system for users with disabilities, such as blindness, to efficiently grasp the contents of documents and to dynamically select and present an appropriate output format based on the user's emotional state. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, detailed information presentation means, and emotion recognition means and emotion engine.

[0227] Program processing and explanation in natural language

[0228] 1. Image recognition and text extraction

[0229] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[0230] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[0231] 2. Summary generation

[0232] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[0233] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[0234] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[0235] 3. Selection of output format and presentation of summary

[0236] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[0237] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[0238] 4. Emotion Recognition and Dynamic Adjustment

[0239] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[0240] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results. For example, if the user is feeling stressed, the system will make the summary more concise and soften the tone of the voice output.

[0241] 5. Requesting and providing detailed information

[0242] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[0243] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[0244] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[0245] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[0246] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[0247] Specific example

[0248] 1. Specific examples of image recognition and text extraction

[0249] Users take pictures of documents using their device's camera in order to read reports they need for work.

[0250] The device sends the captured image to an OCR module, which then extracts the text from the report.

[0251] 2. Specific Examples of Summary Generation

[0252] The terminal sends the extracted report text to the server.

[0253] The server analyzes the text, extracts only the important points, and generates a summary.

[0254] The server sends the generated summary to the terminal.

[0255] 3. Specific examples of output format selection and summary presentation

[0256] The device will read a summary aloud in text-to-speech format based on the user's settings.

[0257] Users can listen to audio to confirm important parts of the report.

[0258] 4. Specific examples of emotion recognition and dynamic regulation

[0259] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[0260] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[0261] 5. Specific examples of requesting and providing detailed information

[0262] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[0263] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[0264] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[0265] The server identifies the details in the second paragraph and sends them to the terminal.

[0266] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[0267] Users can view more detailed information and gain a deeper understanding of the report's contents.

[0268] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[0269] The following describes the processing flow.

[0270] Step 1:

[0271] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[0272] Step 2:

[0273] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[0274] Step 3:

[0275] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[0276] Step 4:

[0277] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[0278] Step 5:

[0279] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[0280] Step 6:

[0281] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[0282] Step 7:

[0283] The terminal presents the summary to the user in the selected format. For example, if the speech synthesis format is selected, the terminal reads out the summary text in voice.

[0284] Step 8:

[0285] The terminal activates the emotion recognition means to detect the user's emotion from the user's expression, voice, and gesture. The terminal monitors the user's emotion in real time using a camera or microphone.

[0286] Step 9:

[0287] The emotion engine analyzes the user's emotion and dynamically adjusts the output format based on the result. For example, if it is determined that the user is confused, change the output format of the summary and add a detailed explanation if necessary.

[0288] Step 10:

[0289] The user listens to or reads the summary and specifies the parts for which they want to know the details. The user uses gestures or device operations to indicate the summary parts of interest.

[0290] Step 11:

[0291] The terminal recognizes the user's operation and identifies the specified summary part. The terminal prepares to request detailed information based on the user's instructions.

[0292] Step 12:

[0293] The terminal requests the detailed information of the specified summary part from the server. The terminal sends the detailed information request to the server via the network.

[0294] Step 13:

[0295] The server receives the detailed information request and searches for the detailed information of the specified summary part. The server identifies the relevant detailed information and returns it to the terminal.

[0296] Step 14:

[0297] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[0298] This series of processing steps allows users to easily review document summaries and efficiently obtain details on specific parts they need. Dynamic adjustments based on emotion recognition further enhance user understanding and enable support tailored to individual needs.

[0299] (Example 2)

[0300] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0301] Visually impaired users have difficulty efficiently understanding the content of documents and obtaining detailed information as needed. Furthermore, while presenting document content requires adjustments based on the user's emotional state, no existing systems dynamically perform these adjustments.

[0302] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an image recognition means, a text extraction means, a summary generation means, an output format selection means, a summary presentation means, a detailed information request means, a detailed information provision means, a detailed information presentation means, an emotion recognition means, and an emotion engine. This enables the user to efficiently grasp the contents of a document and to be provided with information in an appropriate format based on their emotional state.

[0303] "Image recognition means" refers to a means for recognizing the contents of a document from an image taken by a user.

[0304] The "text extraction means" is a means for extracting text data from the image data recognized by the image recognition means.

[0305] The "summary generation means" is a means for analyzing the extracted text data, extracting important information therefrom, and generating a summary.

[0306] The "output format selection means" is a means for selecting the generated summary from output formats such as a braille display, speech synthesis, and sign language video based on the user's settings.

[0307] The "summary presentation means" is a means for presenting the summary to the user according to the selected output format.

[0308] The "detailed information request means" is a means for receiving and processing the request when the user wants to know detailed information about a specific summary part.

[0309] The "detailed information providing means" is a means for obtaining detailed information of the summary part specified by the detailed information request means from the server.

[0310] The "detailed information presentation means" is a means for presenting the obtained detailed information to the user based on the selected output format.

[0311] The "emotion recognition means" is a means for detecting the user's emotional state from the user's expression, voice, gesture, etc.

[0312] The "emotion engine" is a means for analyzing the detected user's emotional state and dynamically adjusting the output format based on the result.

[0313] Mode for implementing the invention

[0314] This invention is a system for enabling visually impaired users to efficiently understand the contents of documents and to provide information in an appropriate format according to their emotional state. This system utilizes the following hardware and software.

[0315] Hardware and software to be used

[0316] Device: This refers to the user's smartphone or tablet. Peripherals such as cameras, microphones, speakers, and braille displays are connected to it.

[0317] Server: Utilizes cloud-based computing resources to perform text analysis, summary generation, and sentiment analysis.

[0318] OCR module: Software that provides optical character recognition technology, such as the Google Cloud Vision API.

[0319] Summarization generation algorithm: Summarization is generated using natural language processing (NLP) techniques such as BERT and GPT-3 (registered trademark).

[0320] Text-to-speech software: Software that converts text into speech, such as Amazon Polly.

[0321] Emotion recognition software: Software for analyzing a user's emotional state, such as Microsoft Azure® Emotion API.

[0322] Specific examples of program processing

[0323] Specific examples of image recognition and text extraction

[0324] Users take pictures of documents using their device's camera in order to read reports they need for work.

[0325] The user launches the camera app, frames the document, and taps the "Shoot" button.

[0326] The device sends the captured image to the Google Cloud Vision API, and the text data is extracted.

[0327] The extracted text data is returned to the terminal.

[0328] Example of a prompt:

[0329] Extract text from the captured images and summarize the contents of the report.

[0330] Concrete examples of summary generation

[0331] The terminal sends the extracted report text to the server.

[0332] Text data is sent to the server via the network.

[0333] The server uses BERT or GPT-3 to extract only the important points and generate a summary.

[0334] The generated summary is sent to the terminal.

[0335] Example of a prompt:

[0336] Please summarize this text data, focusing only on the key points.

[0337] Examples of selecting output formats and presenting summaries

[0338] The device will read a summary aloud in text-to-speech format based on the user's settings.

[0339] The device calls Amazon Polly and converts the summary into speech.

[0340] The synthesized speech summary is played back to the user.

[0341] Example of a prompt:

[0342] Please read this summary aloud using text-to-speech.

[0343] Examples of emotion recognition and dynamic regulation

[0344] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[0345] The device uses the Microsoft Azure Emotion API to analyze the user's emotions.

[0346] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[0347] Example of a prompt:

[0348] If the user is confused, adjust the summary accordingly.

[0349] Examples of requesting and providing detailed information

[0350] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[0351] The user taps the screen to indicate interest in the second paragraph.

[0352] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[0353] The server identifies the details in the second paragraph and sends them to the terminal.

[0354] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[0355] Example of a prompt:

[0356] Please retrieve detailed information about the paragraph specified by the user and provide it in audio format.

[0357] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[0358] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0359] Step 1:

[0360] The user launches the camera app on their device and takes a picture of a physical document.

[0361] Action: The user opens the camera app, frames the document, and taps the "Shoot" button.

[0362] Input: Image of a document.

[0363] Output: Captured image data.

[0364] Step 2:

[0365] The device sends the captured image to an OCR (Optical Character Recognition) module, which then extracts the text data.

[0366] Operation: The device sends image data to the Google Cloud Vision API, and the OCR module analyzes and extracts text data from the image.

[0367] Input: Captured image data.

[0368] Output: Extracted text data.

[0369] Step 3:

[0370] The terminal sends the extracted text data to the server.

[0371] Operation: The device sends text data to the server over the network.

[0372] Input: Extracted text data.

[0373] Output: Text data sent to the server.

[0374] Step 4:

[0375] The server processes the received text data through a summarization algorithm.

[0376] Operation: The server launches an NLP model (e.g., BERT or GPT-3), analyzes the text data, extracts important information, and generates a summary.

[0377] Input: Text data.

[0378] Output: Generated summary data.

[0379] Step 5:

[0380] The server sends the generated summary back to the terminal.

[0381] Operation: The server generates summary data and sends it to the terminal via the network.

[0382] Input: Generated summary data.

[0383] Output: Summary data sent to the terminal.

[0384] Step 6:

[0385] The terminal checks the user's settings and selects the output format.

[0386] Operation: The terminal checks the settings menu and confirms the output format selected by the user (synthesized speech, braille display, sign language video).

[0387] Input: User settings information.

[0388] Output: Selected output format.

[0389] Step 7:

[0390] The device presents the summary to the user in the format selected by the terminal.

[0391] Operation: For example, if the speech synthesis format is selected, the device will call Amazon Polly to convert the summarized text into speech and play it back.

[0392] Input: Summary data.

[0393] Output: Summary presented via speech synthesis.

[0394] Step 8:

[0395] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[0396] Operation: The device uses its camera and microphone to monitor the user's facial expressions and voice, and uses the Microsoft Azure Emotion API to analyze emotions.

[0397] Input: User's facial expressions and voice.

[0398] Output: Detected emotion data.

[0399] Step 9:

[0400] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results.

[0401] Operation: The emotion engine analyzes the detected emotion data and adjusts the output format of the summary as needed (e.g., making the summary more concise and softening the voice tone).

[0402] Input: Sentiment data.

[0403] Output: A summary of the adjusted output format.

[0404] Step 10:

[0405] The user specifies the area they want to know more about.

[0406] Action: The user can use gestures or device operations to highlight the summary section they are interested in.

[0407] Input: User gestures or actions.

[0408] Output: The specified summary portion.

[0409] Step 11:

[0410] The terminal recognizes the user's action and requests detailed information about the specified summary from the server.

[0411] Action: The terminal requests detailed information about the specified summary from the server.

[0412] Input: The specified summary portion.

[0413] Output: Request for more information.

[0414] Step 12:

[0415] The server receives a request for detailed information and searches for detailed information about the specified summary section.

[0416] Operation: The server searches the database and identifies relevant details.

[0417] Input: Request for detailed information.

[0418] Output: Relevant details.

[0419] Step 13:

[0420] The server sends detailed information to the terminal.

[0421] Operation: The server sends detailed information to the terminal over the network.

[0422] Input: Detailed information.

[0423] Output: Detailed information sent to the terminal.

[0424] Step 14:

[0425] The terminal then presents the detailed information received from the server to the user again, based on the output format.

[0426] Operation: For example, if the Braille display is selected, the terminal will display detailed information on the Braille display.

[0427] Input: Detailed information.

[0428] Output: Detailed information presented in output format.

[0429] (Application Example 2)

[0430] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0431] It is difficult for visually impaired users to efficiently grasp product information in physical stores, and there is a lack of systems that provide information in an appropriate format according to the user's emotional state. Therefore, it is necessary to provide an environment in physical stores where users can enjoy shopping with peace of mind, and to improve the efficiency of information access and understanding.

[0432] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0433] In this invention, the server includes image recognition means, text extraction means, and summary generation means. This allows for efficient and effective understanding of product information by extracting product information from images taken by visually impaired users and providing that information as a summary. Furthermore, by using emotion recognition means and emotion engine means, dynamic information presentation according to the user's emotional state becomes possible, allowing users to enjoy shopping with peace of mind.

[0434] An "image recognition system" is a system that analyzes image data acquired by a camera and extracts specific information or features from it.

[0435] "Text extraction means" refers to a function that extracts character information contained within an image as text data using optical character recognition (OCR) technology.

[0436] A "summary generation method" is a system that analyzes extracted text data, extracts only the important information, and summarizes it concisely.

[0437] An "output format selection mechanism" is a system for selecting the format of the information to be output according to the user's settings and circumstances.

[0438] "Means of presenting the summary" refers to means of notifying the user of the generated summary information, and includes various methods such as audio, braille display, and sign language video.

[0439] A "detailed information request system" is a system that accepts requests from users who want to know more detailed information than what is presented in the summary.

[0440] A "detailed information provision means" is a system that obtains and provides additional detailed information based on requests received through a detailed information request means.

[0441] A "means for presenting detailed information" refers to a means of presenting the provided detailed information to the user.

[0442] An "emotion recognition system" is a system that identifies a user's emotional state from their facial expressions, tone of voice, gestures, etc.

[0443] An "emotion engine means" is a system that analyzes emotional information detected by an emotion recognition means and dynamically adjusts the output format and information content based on that analysis.

[0444] To implement this invention, it is first necessary to prepare a terminal equipped with an image recognition means. When a user takes a picture of a product label using the camera of this terminal, the terminal analyzes the image data using the image recognition means and extracts the text information within the label using the text extraction means. By using optical character recognition (OCR) technology, it is possible to accurately obtain the text information within the image as text data. For this technology, open-source OCR libraries such as "Tesseract" can be used.

[0445] Next, the extracted text data is sent to a server, where a summary generation mechanism uses text analysis and natural language processing (NLP) algorithms to summarize the important information. The server returns the generated summary to the terminal, which uses an output format selection mechanism to select an output format according to the user's settings and circumstances. These output formats include audio, braille display, and sign language video. For audio output, "gTTS (Google Text-to-Speech)" and other formats can be used.

[0446] When a summary is presented to the user, an emotion recognition system identifies the user's emotional state from their facial expressions, tone of voice, gestures, etc. An emotion engine analyzes this emotional information and dynamically adjusts the output format and content. For example, if the system determines that the user is confused, it makes the summary more concise and softens the tone of the voice output to aid the user's understanding.

[0447] Furthermore, if a user listens to or reads a summary and requires specific details, they can specify that information using the details request mechanism. The terminal requests detailed information for the specified portion from the server, and the server retrieves the relevant details and sends them to the terminal. The terminal then presents the information back to the user via the details provision mechanism. This process allows the user to efficiently and flexibly grasp product information.

[0448] For example, consider a scenario where a user wants to scan a product label in a supermarket. The user takes a picture of the label with their smartphone camera. The device uses OCR technology to extract the text from the label, and a summary is generated on the server. The summary is then returned to the device and presented to the user as audio output. If the emotion engine detects the user's confusion based on their facial expression and tone of voice, the summary becomes more concise and the audio tone softens. If the user shows interest in specific information, additional details are provided.

[0449] Example of a prompt:

[0450] Please briefly summarize the following text:

[0451] Product name: The product in question

[0452] Price: 1000 yen

[0453] Ingredients: Water, sugar, salt

[0454] Manufacturer: That company

[0455] Country of origin: Japan

[0456] Summary: The product in question can be purchased for 1000 yen. Its ingredients include water, sugar, and salt. It is manufactured by the company mentioned, and its country of origin is Japan.

[0457] This allows visually impaired users to efficiently understand product information even in physical stores, enabling them to enjoy shopping with peace of mind.

[0458] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0459] Step 1:

[0460] The user takes a picture of the product label with the device's camera. The user's input is image data, and the device captures and saves this image data. This image data is then processed in the following steps.

[0461] Step 2:

[0462] The device sends the captured image to the OCR module. The OCR module uses optical character recognition technology to extract text information from the image. The input is image data, and the output is extracted text data. This text data is used for summary generation in the next step.

[0463] Step 3:

[0464] The terminal sends text data acquired from the OCR module to the server. The server processes the received text data using a natural language processing algorithm. The input is text data, and the output is generated summary data. The server generates the summary data and sends it back to the terminal.

[0465] Step 4:

[0466] The terminal receives summary data from the server and converts it to the appropriate output format based on user settings. For example, if audio output is selected, the summary text is converted to audio. The input is the summary data and user settings, and the output is audio data.

[0467] Step 5:

[0468] The device presents the user with summary information in an output format. In the case of audio, the device reads the summary aloud. In this step, the input is audio data or other output format data, and the output is the information presented to the user.

[0469] Step 6:

[0470] The device activates an emotion recognition mechanism to detect emotions from the user's facial expressions, tone of voice, and gestures. The input is data of the user's facial expressions and voice, and the output is the analyzed emotion information. The emotion recognition mechanism transmits this to the emotion engine.

[0471] Step 7:

[0472] The emotion engine analyzes the user's emotions and dynamically adjusts the output format and summary content based on the results. The input is emotional information, and the output is the adjusted summary content and output format. For example, if the user is confused, the summary content will be made simpler and the voice tone will be softened.

[0473] Step 8:

[0474] After the user has heard the summary, if they require specific details, they can use the details request mechanism to provide instructions. The input is the user's request for details, and the output is the specific details requested.

[0475] Step 9:

[0476] The terminal sends a specified request for detailed information to the server. The server searches for and retrieves the necessary detailed information. The input is the request for detailed information, and the output is the detailed information data.

[0477] Step 10:

[0478] The server sends detailed information to the terminal, and the terminal converts the detailed information into an output format and presents it to the user. The input consists of detailed information data and output format settings, and the output is the final presentation of the detailed information.

[0479] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0480] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0481] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0482] [Second Embodiment]

[0483] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0484] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0485] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0486] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0487] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0488] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0489] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0490] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0491] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0492] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0493] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0494] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0495] This invention is a system for users with disabilities, such as blindness, to efficiently understand the contents of documents. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, and detailed information presentation means.

[0496] Program processing and explanation in natural language

[0497] 1. Image recognition and text extraction

[0498] The user takes a picture of a document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology.

[0499] 2. Summary generation

[0500] The terminal sends the extracted text to the server.

[0501] The server analyzes the received text, extracts important information using a summarization algorithm, and generates a summary.

[0502] 3. Selection of output format and presentation of summary

[0503] The terminal checks the user's settings and outputs a summary in formats such as Braille display, speech synthesis, and sign language video.

[0504] 4. Request for detailed information

[0505] The user can specify the summary section they are interested in using gestures or device operations.

[0506] The terminal requests detailed information about the specified part from the server.

[0507] 5. Providing detailed information

[0508] The server searches for the requested details and sends them to the terminal.

[0509] 6. Output of detailed information

[0510] The device then presents detailed information to the user again in a selected format, such as braille display, speech synthesis, or sign language video.

[0511] Specific example

[0512] 1. Specific examples of image recognition and text extraction

[0513] Users take pictures of documents using their device's camera in order to read reports they need for work.

[0514] The device sends the captured image to an OCR module, which then extracts the text from the report.

[0515] 2. Specific Examples of Summary Generation

[0516] The terminal sends the extracted report text to the server.

[0517] The server analyzes the text, extracts only the important points, and generates a summary.

[0518] The server sends the generated summary to the terminal.

[0519] 3. Specific examples of output format selection and summary presentation

[0520] The device will read a summary aloud in text-to-speech format based on the user's settings.

[0521] Users can listen to audio to confirm important parts of the report.

[0522] 4. Specific examples of requests for detailed information

[0523] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[0524] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[0525] 5. Specific examples of providing detailed information

[0526] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[0527] The server identifies the details in the second paragraph and sends them to the terminal.

[0528] 6. Specific examples of outputting detailed information

[0529] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[0530] Users can view more detailed information and gain a deeper understanding of the report's contents.

[0531] In this way, users can efficiently grasp the entire document and easily check the details of the parts they need. This system is extremely convenient for users with visual impairments.

[0532] The following describes the processing flow.

[0533] Step 1:

[0534] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[0535] Step 2:

[0536] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[0537] Step 3:

[0538] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[0539] Step 4:

[0540] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[0541] Step 5:

[0542] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[0543] Step 6:

[0544] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in one of the following formats: Braille display, speech synthesis, or sign language video.

[0545] Step 7:

[0546] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[0547] Step 8:

[0548] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[0549] Step 9:

[0550] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[0551] Step 10:

[0552] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[0553] Step 11:

[0554] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[0555] Step 12:

[0556] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[0557] This series of processing steps allows users to easily review a document summary and, if more detailed information is needed, efficiently retrieve specific details from particular sections.

[0558] (Example 1)

[0559] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0560] There is a need for a system that allows visually impaired users to efficiently understand the content of documents and easily access detailed information on the parts they need. However, conventional systems require a great deal of effort and time to fully understand the content of documents, which is a significant burden on users. In particular, there were technical challenges in providing summaries and detailed information in the required format.

[0561] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0562] In this invention, the server includes an image recognition means, an optical character recognition means, and a summary generation means. This makes it possible for a visually impaired user to take an image of a document, extract text from the image, efficiently summarize the extracted text, and present it to the user in an appropriate format.

[0563] "Image recognition means" refers to technology that identifies images of documents taken by the user using the camera on their device and extracts the necessary information.

[0564] "Optical character recognition means" refers to a technology that analyzes characters within a captured image and extracts them as digital text.

[0565] A "summary generation method" is a technology that analyzes extracted text, extracts important information, and generates a concise summary.

[0566] The "output format selection means" is a technology that outputs the generated summary in an appropriate format (e.g., audio, braille display, sign language video) based on the user's settings.

[0567] A "summary presentation method" is a technology that presents a summary to the user in a selected format.

[0568] A "means for requesting detailed information" is a technology that allows a user to specify detailed information about a summary section that interests them and communicate that request to the system.

[0569] "Means of providing detailed information" refers to a technology in which a server searches for and provides detailed information on a specified part based on a user's request.

[0570] A "means for presenting detailed information" refers to a technology that presents detailed information to the user in a selected output format.

[0571] Modes for carrying out the invention

[0572] This invention is a system for visually impaired users to efficiently understand the contents of documents. This system utilizes multiple hardware and software components.

[0573] 1. Image recognition and text extraction

[0574] The user takes a picture of the document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology. Software used includes "Google Cloud Vision API" and "Adobe OCR".

[0575] Specific example: A user takes a picture of a document using their device's camera to read a report they need for work. The device sends the captured image to an OCR module, which extracts the text from the report.

[0576] 2. Summary generation

[0577] The terminal sends the extracted text to the server.

[0578] The server analyzes the received text, extracts key information using a summarization algorithm, and generates a summary. Suitable generative AI models to use include "BERT" and "GPT-4".

[0579] Specific example: The terminal sends the extracted report text to the server. The server analyzes the text, extracts only the important points, and generates a summary. The server then sends the generated summary to the terminal.

[0580] 3. Selection of output format and presentation of summary

[0581] The device checks the user's settings and outputs a summary in formats such as Braille display, text-to-speech, and sign language video. Text-to-speech uses "Amazon Polly" or "Google Text-to-Speech."

[0582] Specific example: The device reads a summary aloud in text-to-speech format based on the user's settings. The user then reviews important parts of the report by voice.

[0583] 4. Request for detailed information

[0584] Users can specify the summary sections they are interested in using gestures or device operations.

[0585] The device requests detailed information about the specified area from the server. Gesture recognition uses technologies such as "Microsoft Kinect" or "Leap Motion".

[0586] Specific example: A user listens to the audio, feels particularly interested in the second paragraph, and wants more detailed information. They then use a gesture to indicate their interest in the second paragraph to the device.

[0587] 5. Providing detailed information

[0588] The server searches for the requested details and sends them to the terminal. When the server generates additional information based on the text of the specified paragraph, it uses the "GPT-4" AI model.

[0589] Specific example: The terminal recognizes the user's action and requests detailed information about the specified second paragraph from the server. The server identifies the detailed information about the second paragraph and sends it to the terminal.

[0590] 6. Output of detailed information

[0591] The terminal then presents detailed information to the user again in a selected format, such as Braille display, speech synthesis, or sign language video.

[0592] Review the detailed information provided by the user.

[0593] Specific example: The device reads aloud the detailed information from the second paragraph using speech synthesis. The user can then review the more detailed information and gain a deeper understanding of the report's content.

[0594] Examples of prompt statements

[0595] "Extract text from documents photographed using Google Cloud Vision, and then summarize that text using GPT-4."

[0596] "Summarize the extracted text and use Amazon Polly to have it read aloud."

[0597] "Provide details about the paragraphs that the user is interested in, and then use Amazon Polly to convert those details into speech and present them."

[0598] In this way, the system provides a way for visually impaired users to efficiently grasp the entire document and easily check the details of the parts they need.

[0599] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0600] Program processing flow

[0601] Step 1:

[0602] The user takes a picture of a document using the device's camera. The input is an image of the document, and the output is image data saved on the device.

[0603] Specific action: The user presses the "Capture" button on the device.

[0604] Step 2:

[0605] The device takes a picture and passes it to an OCR module for processing. The input is image data, and the output is text data extracted by OCR.

[0606] Specific operation: The device sends image data to the "Optical Character Recognition API".

[0607] Step 3:

[0608] The terminal receives text from the OCR module and sends it to the server. The input is text data, and the output is a transfer to the server.

[0609] Specific action: The device uses an "HTTPS request" to send text data to the server.

[0610] Step 4:

[0611] The server parses the received text and executes a summary generation algorithm. The input is text data, and the output is summarized text.

[0612] Specific operation: The server generates a summary using "natural language processing technology".

[0613] Step 5:

[0614] The server generates a summary and sends it to the terminal. The input is the summary text, and the output is the transfer to the terminal.

[0615] Specific operation: The server sends summary data to the terminal via an "HTTPS response".

[0616] Step 6:

[0617] The terminal checks the user's settings and presents the summary in the selected format. Input is the summary text and user settings information, and output is in the form of audio, Braille display, sign language video, etc.

[0618] Specific action: The device executes the "Speech Synthesis API" or another selected output method.

[0619] Step 7:

[0620] The user reviews the provided summary and specifies the areas for which they want more detailed information. The input is an audio summary, and the output is a request for more detailed information.

[0621] Specific action: The user requests more information using gestures or input devices.

[0622] Step 8:

[0623] The terminal recognizes user input and requests detailed information about the specified portion from the server. The input is a request for detailed information, and the output is the transfer of the request for detailed information to the server.

[0624] Specific action: The terminal sends an "HTTPS request" and forwards a prompt message to the server requesting more information.

[0625] Step 9:

[0626] The server prepares the requested details and sends them to the terminal. The input is the request for details, and the output is the details.

[0627] Specific operation: The server generates the specified content based on "natural language processing technology" and sends it to the terminal.

[0628] Step 10:

[0629] The terminal then presents the detailed information to the user again in the selected format. The input is detailed information, and the output is in the form of audio, braille display, sign language video, etc.

[0630] Specific action: The device executes the "Speech Synthesis API" or another selected output method and presents detailed information.

[0631] keyword

[0632] Generative AI model, prompt sentence

[0633] (Application Example 1)

[0634] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0635] This solution addresses the challenge faced by visually impaired users who lack the efficient means to quickly and accurately grasp the contents of documents such as monitoring reports in real time, and to easily access detailed information as needed.

[0636] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0637] In this invention, the server includes image recognition means, text extraction means, summary generation means, and speech synthesis means. This makes it possible for visually impaired users to obtain summarized information from documents such as monitoring reports in audio format, and to be provided with even more detailed information in audio format as well.

[0638] "Image recognition means" refers to a technology that uses an image input device such as a camera to analyze captured image data and recognize necessary information.

[0639] "Text extraction means" refers to a technology that extracts character information from image data using optical character recognition (OCR) technology.

[0640] A "summary generation method" is a technology that analyzes extracted text data, extracts important information, and generates a concise summary.

[0641] "Output format selection means" refers to a technology that outputs the generated summary information in formats such as Braille display, speech synthesis, and sign language video, according to the user's settings and preferences.

[0642] A "summary presentation method" is a technology that presents summary information to the user based on the selected output format.

[0643] A "means for requesting detailed information" is a technology that allows a user to specify a part of a summary that interests them and request detailed information about that part.

[0644] "Means of providing detailed information" refers to technology that searches for requested detailed information and provides it to the user.

[0645] "Means for presenting detailed information" refers to technologies that present the provided detailed information to the user in the form of Braille displays, speech synthesis, sign language videos, etc.

[0646] "Speech synthesis means" refers to a technology that converts text data into speech and provides information in audio format.

[0647] "Voice recognition means" refers to technology that analyzes the user's voice input and recognizes its content.

[0648] This invention is a system for visually impaired users to efficiently understand the contents of documents such as papers and surveillance reports. Next, a specific embodiment of the system will be described.

[0649] This system consists of a user-facing terminal and multiple processing modules running on a server. The terminal is equipped with a camera, microphone, and speaker. The main processing includes the following:

[0650] 1. Image recognition and text extraction

[0651] The user takes a picture of a surveillance report or document using the device's camera. The device's camera app captures the image, and that image is converted into text using optical character recognition (OCR) technology. For example, software such as Tesseract OCR can be used.

[0652] 2. Sending text to the server and generating a summary

[0653] The extracted text data is sent to the server in real time. On the server, a generative AI model runs, using text analysis and natural language processing techniques to generate a summary of the key information. Natural language processing algorithms such as BERT are used.

[0654] 3. Selection and presentation of summary output format

[0655] The generated summary is provided as audio using speech synthesis technology, depending on the user's settings. Speech synthesis engines such as Google TTS are used. Users can listen to the summary through their speakers.

[0656] 4. Requesting and providing detailed information

[0657] If a user is interested in a particular summary, they can request more information using speech recognition technology. Speech recognition engines such as Sphinx or Google Speech Recognition are used. The requested information is sent back to the server, searched on the server, and then provided via speech synthesis.

[0658] The following are specific usage scenarios.

[0659] Specific usage scenarios:

[0660] Consider a scenario where a visually impaired security staff member is reviewing a surveillance report. The staff member uses their smartphone camera to photograph the report, and OCR technology extracts the text on the device. This text is sent to a server, where a generative AI model generates a summary. The summary is then provided to the staff member in audio format via a speech synthesis engine. If the staff member is interested in a particular paragraph, they can request further details by communicating this using speech recognition technology, and these details are also provided in audio format.

[0661] Examples of input prompts for a generative AI model:

[0662] Extract the key information from the following text and generate a summary. The theme is a monitoring report:

[0663] (Text from the monitoring report)

[0664] This system allows visually impaired users to quickly understand important information and obtain detailed information as needed.

[0665] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0666] Step 1:

[0667] The user takes a picture of the monitoring report with their camera. The device's camera app captures the image, and that image is converted into text using optical character recognition technology (such as Tesseract OCR). Specifically, the captured image data is input into an OCR module, and the text data is output.

[0668] Step 2:

[0669] The terminal sends the extracted text data to the server. Here, the text data is packetized in formats such as JSON or XML and sent to the server over the network. The server stores the received text data for analysis.

[0670] Step 3:

[0671] The server analyzes the received text data. First, it uses a natural language processing algorithm (such as a generative AI model like BERT) to extract important information from the text and generate a summary. This summary generation algorithm produces a compact text that extracts only the essential points while eliminating redundant parts.

[0672] Step 4:

[0673] The server generates a summary and sends it to the terminal. The summary text is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received summary data and continues processing based on the output format set by the user.

[0674] Step 5:

[0675] The device processes the received summary text into speech using a text-to-speech engine (such as Google TTS). When summary text is input, the text-to-speech engine converts the text into natural-sounding speech, which is then presented to the user through the speaker.

[0676] Step 6:

[0677] The user listens to an audio summary and requests more detailed information on the parts that interest them through a speech recognition system (such as Google Speech Recognition). Specifically, the user verbally indicates the parts they want to know more about, the microphone captures this, and it is input into the speech recognition engine. The speech recognition engine converts the audio into text, and that text is sent to the server.

[0678] Step 7:

[0679] The server receives a request for user details. The server searches the database for and extracts detailed text information for the requested portion. This extraction process generates detailed text data.

[0680] Step 8:

[0681] The server sends the extracted detailed information to the terminal. The detailed information is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received detailed information and prepares to present it in the user-configured output format.

[0682] Step 9:

[0683] The device receives detailed information and converts it into speech using a speech synthesis engine. Detailed information text is input, the speech synthesis engine converts that text into natural-sounding speech, and it is presented to the user through the speaker.

[0684] This series of processing steps allows visually impaired users to quickly and efficiently understand the contents of monitoring reports.

[0685] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0686] This invention is a system for users with disabilities, such as blindness, to efficiently grasp the contents of documents and to dynamically select and present an appropriate output format based on the user's emotional state. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, detailed information presentation means, and emotion recognition means and emotion engine.

[0687] Program processing and explanation in natural language

[0688] 1. Image recognition and text extraction

[0689] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[0690] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[0691] 2. Summary generation

[0692] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[0693] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[0694] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[0695] 3. Selection of output format and presentation of summary

[0696] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[0697] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[0698] 4. Emotion Recognition and Dynamic Adjustment

[0699] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[0700] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results. For example, if the user is feeling stressed, the system will make the summary more concise and soften the tone of the voice output.

[0701] 5. Requesting and providing detailed information

[0702] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[0703] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[0704] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[0705] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[0706] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[0707] Specific example

[0708] 1. Specific examples of image recognition and text extraction

[0709] Users take pictures of documents using their device's camera in order to read reports they need for work.

[0710] The device sends the captured image to an OCR module, which then extracts the text from the report.

[0711] 2. Specific Examples of Summary Generation

[0712] The terminal sends the extracted report text to the server.

[0713] The server analyzes the text, extracts only the important points, and generates a summary.

[0714] The server sends the generated summary to the terminal.

[0715] 3. Specific examples of output format selection and summary presentation

[0716] The device will read a summary aloud in text-to-speech format based on the user's settings.

[0717] Users can listen to audio to confirm important parts of the report.

[0718] 4. Specific examples of emotion recognition and dynamic regulation

[0719] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[0720] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[0721] 5. Specific examples of requesting and providing detailed information

[0722] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[0723] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[0724] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[0725] The server identifies the details in the second paragraph and sends them to the terminal.

[0726] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[0727] Users can view more detailed information and gain a deeper understanding of the report's contents.

[0728] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[0729] The following describes the processing flow.

[0730] Step 1:

[0731] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[0732] Step 2:

[0733] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[0734] Step 3:

[0735] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[0736] Step 4:

[0737] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[0738] Step 5:

[0739] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[0740] Step 6:

[0741] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[0742] Step 7:

[0743] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[0744] Step 8:

[0745] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures. The device uses its camera and microphone to monitor the user's emotions in real time.

[0746] Step 9:

[0747] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results. For example, if it determines that the user is confused, it changes the output format of the summary and adds more detailed explanations as needed.

[0748] Step 10:

[0749] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[0750] Step 11:

[0751] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[0752] Step 12:

[0753] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[0754] Step 13:

[0755] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[0756] Step 14:

[0757] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[0758] This series of processing steps allows users to easily review document summaries and efficiently obtain details on specific parts they need. Dynamic adjustments based on emotion recognition further enhance user understanding and enable support tailored to individual needs.

[0759] (Example 2)

[0760] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0761] Visually impaired users have difficulty efficiently understanding the content of documents and obtaining detailed information as needed. Furthermore, while presenting document content requires adjustments based on the user's emotional state, no existing systems dynamically perform these adjustments.

[0762] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an image recognition means, a text extraction means, a summary generation means, an output format selection means, a summary presentation means, a detailed information request means, a detailed information provision means, a detailed information presentation means, an emotion recognition means, and an emotion engine. This enables the user to efficiently grasp the contents of a document and to be provided with information in an appropriate format based on their emotional state.

[0763] "Image recognition means" refers to a means for recognizing the contents of a document from an image taken by a user.

[0764] A "text extraction means" is a means for extracting text data from image data recognized by an image recognition means.

[0765] A "summary generation means" is a means for analyzing extracted text data, extracting important information from it, and generating a summary.

[0766] The "output format selection means" is a means for selecting the output format of the generated summary from among options such as braille display, speech synthesis, and sign language video, based on the user's settings.

[0767] "Means for presenting a summary" refers to means for presenting a summary to the user according to the selected output format.

[0768] A "means for requesting detailed information" refers to a means for receiving and processing requests from users who wish to know more detailed information about a particular summary.

[0769] "Detailed information provision means" refers to means for obtaining detailed information from a server about the summary portion identified by the detailed information request means.

[0770] A "means for presenting detailed information" refers to a means for presenting acquired detailed information to the user based on a selected output format.

[0771] "Emotion recognition means" refers to methods for detecting a user's emotional state from their facial expressions, voice, gestures, etc.

[0772] An "emotion engine" is a means of analyzing the detected emotional state of the user and dynamically adjusting the output format based on the results.

[0773] Modes for carrying out the invention

[0774] This invention is a system for enabling visually impaired users to efficiently understand the contents of documents and to provide information in an appropriate format according to their emotional state. This system utilizes the following hardware and software.

[0775] Hardware and software to be used

[0776] Device: This refers to the user's smartphone or tablet. Peripherals such as cameras, microphones, speakers, and braille displays are connected to it.

[0777] Server: Utilizes cloud-based computing resources to perform text analysis, summary generation, and sentiment analysis.

[0778] OCR module: Software that provides optical character recognition technology, such as the Google Cloud Vision API.

[0779] Summarization generation algorithm: Summarization is generated using natural language processing (NLP) techniques such as BERT and GPT-3.

[0780] Text-to-speech software: Software that converts text into speech, such as Amazon Polly.

[0781] Emotion recognition software: Software used to analyze a user's emotional state, such as the Microsoft Azure Emotion API.

[0782] Specific examples of program processing

[0783] Specific examples of image recognition and text extraction

[0784] Users take pictures of documents using their device's camera in order to read reports they need for work.

[0785] The user launches the camera app, frames the document, and taps the "Shoot" button.

[0786] The device sends the captured image to the Google Cloud Vision API, and the text data is extracted.

[0787] The extracted text data is returned to the terminal.

[0788] Example of a prompt:

[0789] Extract text from the captured images and summarize the contents of the report.

[0790] Concrete examples of summary generation

[0791] The terminal sends the extracted report text to the server.

[0792] Text data is sent to the server via the network.

[0793] The server uses BERT or GPT-3 to extract only the important points and generate a summary.

[0794] The generated summary is sent to the terminal.

[0795] Example of a prompt:

[0796] Please summarize this text data, focusing only on the key points.

[0797] Examples of selecting output formats and presenting summaries

[0798] The device will read a summary aloud in text-to-speech format based on the user's settings.

[0799] The device calls Amazon Polly and converts the summary into speech.

[0800] The synthesized speech summary is played back to the user.

[0801] Example of a prompt:

[0802] Please read this summary aloud using text-to-speech.

[0803] Examples of emotion recognition and dynamic regulation

[0804] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[0805] The device uses the Microsoft Azure Emotion API to analyze the user's emotions.

[0806] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[0807] Example of a prompt:

[0808] If the user is confused, adjust the summary accordingly.

[0809] Examples of requesting and providing detailed information

[0810] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[0811] The user taps the screen to indicate interest in the second paragraph.

[0812] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[0813] The server identifies the details in the second paragraph and sends them to the terminal.

[0814] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[0815] Example of a prompt:

[0816] Please retrieve detailed information about the paragraph specified by the user and provide it in audio format.

[0817] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[0818] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0819] Step 1:

[0820] The user launches the camera app on their device and takes a picture of a physical document.

[0821] Action: The user opens the camera app, frames the document, and taps the "Shoot" button.

[0822] Input: Image of a document.

[0823] Output: Captured image data.

[0824] Step 2:

[0825] The device sends the captured image to an OCR (Optical Character Recognition) module, which then extracts the text data.

[0826] Operation: The device sends image data to the Google Cloud Vision API, and the OCR module analyzes and extracts text data from the image.

[0827] Input: Captured image data.

[0828] Output: Extracted text data.

[0829] Step 3:

[0830] The terminal sends the extracted text data to the server.

[0831] Operation: The device sends text data to the server over the network.

[0832] Input: Extracted text data.

[0833] Output: Text data sent to the server.

[0834] Step 4:

[0835] The server processes the received text data through a summarization algorithm.

[0836] Operation: The server launches an NLP model (e.g., BERT or GPT-3), analyzes the text data, extracts important information, and generates a summary.

[0837] Input: Text data.

[0838] Output: Generated summary data.

[0839] Step 5:

[0840] The server sends the generated summary back to the terminal.

[0841] Operation: The server generates summary data and sends it to the terminal via the network.

[0842] Input: Generated summary data.

[0843] Output: Summary data sent to the terminal.

[0844] Step 6:

[0845] The terminal checks the user's settings and selects the output format.

[0846] Operation: The terminal checks the settings menu and confirms the output format selected by the user (synthesized speech, braille display, sign language video).

[0847] Input: User settings information.

[0848] Output: Selected output format.

[0849] Step 7:

[0850] The device presents the summary to the user in the format selected by the terminal.

[0851] Operation: For example, if the speech synthesis format is selected, the device will call Amazon Polly to convert the summarized text into speech and play it back.

[0852] Input: Summary data.

[0853] Output: Summary presented via speech synthesis.

[0854] Step 8:

[0855] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[0856] Operation: The device uses its camera and microphone to monitor the user's facial expressions and voice, and uses the Microsoft Azure Emotion API to analyze emotions.

[0857] Input: User's facial expressions and voice.

[0858] Output: Detected emotion data.

[0859] Step 9:

[0860] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results.

[0861] Operation: The emotion engine analyzes the detected emotion data and adjusts the output format of the summary as needed (e.g., making the summary more concise and softening the voice tone).

[0862] Input: Sentiment data.

[0863] Output: A summary of the adjusted output format.

[0864] Step 10:

[0865] The user specifies the area they want to know more about.

[0866] Action: The user can use gestures or device operations to highlight the summary section they are interested in.

[0867] Input: User gestures or actions.

[0868] Output: The specified summary portion.

[0869] Step 11:

[0870] The terminal recognizes the user's action and requests detailed information about the specified summary from the server.

[0871] Action: The terminal requests detailed information about the specified summary from the server.

[0872] Input: The specified summary portion.

[0873] Output: Request for more information.

[0874] Step 12:

[0875] The server receives a request for detailed information and searches for detailed information about the specified summary section.

[0876] Operation: The server searches the database and identifies relevant details.

[0877] Input: Request for detailed information.

[0878] Output: Relevant details.

[0879] Step 13:

[0880] The server sends detailed information to the terminal.

[0881] Operation: The server sends detailed information to the terminal over the network.

[0882] Input: Detailed information.

[0883] Output: Detailed information sent to the terminal.

[0884] Step 14:

[0885] The terminal then presents the detailed information received from the server to the user again, based on the output format.

[0886] Operation: For example, if the Braille display is selected, the terminal will display detailed information on the Braille display.

[0887] Input: Detailed information.

[0888] Output: Detailed information presented in output format.

[0889] (Application Example 2)

[0890] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0891] It is difficult for visually impaired users to efficiently grasp product information in physical stores, and there is a lack of systems that provide information in an appropriate format according to the user's emotional state. Therefore, it is necessary to provide an environment in physical stores where users can enjoy shopping with peace of mind, and to improve the efficiency of information access and understanding.

[0892] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0893] In this invention, the server includes image recognition means, text extraction means, and summary generation means. This allows for efficient and effective understanding of product information by extracting product information from images taken by visually impaired users and providing that information as a summary. Furthermore, by using emotion recognition means and emotion engine means, dynamic information presentation according to the user's emotional state becomes possible, allowing users to enjoy shopping with peace of mind.

[0894] An "image recognition system" is a system that analyzes image data acquired by a camera and extracts specific information or features from it.

[0895] "Text extraction means" refers to a function that extracts character information contained within an image as text data using optical character recognition (OCR) technology.

[0896] A "summary generation method" is a system that analyzes extracted text data, extracts only the important information, and summarizes it concisely.

[0897] An "output format selection mechanism" is a system for selecting the format of the information to be output according to the user's settings and circumstances.

[0898] "Means of presenting the summary" refers to means of notifying the user of the generated summary information, and includes various methods such as audio, braille display, and sign language video.

[0899] A "detailed information request system" is a system that accepts requests from users who want to know more detailed information than what is presented in the summary.

[0900] A "detailed information provision means" is a system that obtains and provides additional detailed information based on requests received through a detailed information request means.

[0901] A "means for presenting detailed information" refers to a means of presenting the provided detailed information to the user.

[0902] An "emotion recognition system" is a system that identifies a user's emotional state from their facial expressions, tone of voice, gestures, etc.

[0903] An "emotion engine means" is a system that analyzes emotional information detected by an emotion recognition means and dynamically adjusts the output format and information content based on that analysis.

[0904] To implement this invention, it is first necessary to prepare a terminal equipped with an image recognition means. When a user takes a picture of a product label using the camera of this terminal, the terminal analyzes the image data using the image recognition means and extracts the text information within the label using the text extraction means. By using optical character recognition (OCR) technology, it is possible to accurately obtain the text information within the image as text data. For this technology, open-source OCR libraries such as "Tesseract" can be used.

[0905] Next, the extracted text data is sent to a server, where a summary generation mechanism uses text analysis and natural language processing (NLP) algorithms to summarize the important information. The server returns the generated summary to the terminal, which uses an output format selection mechanism to select an output format according to the user's settings and circumstances. These output formats include audio, braille display, and sign language video. For audio output, "gTTS (Google Text-to-Speech)" and other formats can be used.

[0906] When a summary is presented to the user, an emotion recognition system identifies the user's emotional state from their facial expressions, tone of voice, gestures, etc. An emotion engine analyzes this emotional information and dynamically adjusts the output format and content. For example, if the system determines that the user is confused, it makes the summary more concise and softens the tone of the voice output to aid the user's understanding.

[0907] Furthermore, if a user listens to or reads a summary and requires specific details, they can specify that information using the details request mechanism. The terminal requests detailed information for the specified portion from the server, and the server retrieves the relevant details and sends them to the terminal. The terminal then presents the information back to the user via the details provision mechanism. This process allows the user to efficiently and flexibly grasp product information.

[0908] For example, consider a scenario where a user wants to scan a product label in a supermarket. The user takes a picture of the label with their smartphone camera. The device uses OCR technology to extract the text from the label, and a summary is generated on the server. The summary is then returned to the device and presented to the user as audio output. If the emotion engine detects the user's confusion based on their facial expression and tone of voice, the summary becomes more concise and the audio tone softens. If the user shows interest in specific information, additional details are provided.

[0909] Example of a prompt:

[0910] Please briefly summarize the following text:

[0911] Product name: The product in question

[0912] Price: 1000 yen

[0913] Ingredients: Water, sugar, salt

[0914] Manufacturer: That company

[0915] Country of origin: Japan

[0916] Summary: The product in question can be purchased for 1000 yen. Its ingredients include water, sugar, and salt. It is manufactured by the company mentioned, and its country of origin is Japan.

[0917] This allows visually impaired users to efficiently understand product information even in physical stores, enabling them to enjoy shopping with peace of mind.

[0918] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0919] Step 1:

[0920] The user takes a picture of the product label with the device's camera. The user's input is image data, and the device captures and saves this image data. This image data is then processed in the following steps.

[0921] Step 2:

[0922] The device sends the captured image to the OCR module. The OCR module uses optical character recognition technology to extract text information from the image. The input is image data, and the output is extracted text data. This text data is used for summary generation in the next step.

[0923] Step 3:

[0924] The terminal sends text data acquired from the OCR module to the server. The server processes the received text data using a natural language processing algorithm. The input is text data, and the output is generated summary data. The server generates the summary data and sends it back to the terminal.

[0925] Step 4:

[0926] The terminal receives summary data from the server and converts it to the appropriate output format based on user settings. For example, if audio output is selected, the summary text is converted to audio. The input is the summary data and user settings, and the output is audio data.

[0927] Step 5:

[0928] The device presents the user with summary information in an output format. In the case of audio, the device reads the summary aloud. In this step, the input is audio data or other output format data, and the output is the information presented to the user.

[0929] Step 6:

[0930] The device activates an emotion recognition mechanism to detect emotions from the user's facial expressions, tone of voice, and gestures. The input is data of the user's facial expressions and voice, and the output is the analyzed emotion information. The emotion recognition mechanism transmits this to the emotion engine.

[0931] Step 7:

[0932] The emotion engine analyzes the user's emotions and dynamically adjusts the output format and summary content based on the results. The input is emotional information, and the output is the adjusted summary content and output format. For example, if the user is confused, the summary content will be made simpler and the voice tone will be softened.

[0933] Step 8:

[0934] After the user has heard the summary, if they require specific details, they can use the details request mechanism to provide instructions. The input is the user's request for details, and the output is the specific details requested.

[0935] Step 9:

[0936] The terminal sends a specified request for detailed information to the server. The server searches for and retrieves the necessary detailed information. The input is the request for detailed information, and the output is the detailed information data.

[0937] Step 10:

[0938] The server sends detailed information to the terminal, and the terminal converts the detailed information into an output format and presents it to the user. The input consists of detailed information data and output format settings, and the output is the final presentation of the detailed information.

[0939] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0940] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0941] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0942] [Third Embodiment]

[0943] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0944] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0945] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0946] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0947] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0948] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0949] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0950] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0951] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0952] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0953] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0954] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0955] This invention is a system for users with disabilities, such as blindness, to efficiently understand the contents of documents. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, and detailed information presentation means.

[0956] Program processing and explanation in natural language

[0957] 1. Image recognition and text extraction

[0958] The user takes a picture of a document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology.

[0959] 2. Summary generation

[0960] The terminal sends the extracted text to the server.

[0961] The server analyzes the received text, extracts important information using a summarization algorithm, and generates a summary.

[0962] 3. Selection of output format and presentation of summary

[0963] The terminal checks the user's settings and outputs a summary in formats such as Braille display, speech synthesis, and sign language video.

[0964] 4. Request for detailed information

[0965] The user can specify the summary section they are interested in using gestures or device operations.

[0966] The terminal requests detailed information about the specified part from the server.

[0967] 5. Providing detailed information

[0968] The server searches for the requested details and sends them to the terminal.

[0969] 6. Output of detailed information

[0970] The device then presents detailed information to the user again in a selected format, such as braille display, speech synthesis, or sign language video.

[0971] Specific example

[0972] 1. Specific examples of image recognition and text extraction

[0973] Users take pictures of documents using their device's camera in order to read reports they need for work.

[0974] The device sends the captured image to an OCR module, which then extracts the text from the report.

[0975] 2. Specific Examples of Summary Generation

[0976] The terminal sends the extracted report text to the server.

[0977] The server analyzes the text, extracts only the important points, and generates a summary.

[0978] The server sends the generated summary to the terminal.

[0979] 3. Specific examples of output format selection and summary presentation

[0980] The device will read a summary aloud in text-to-speech format based on the user's settings.

[0981] Users can listen to audio to confirm important parts of the report.

[0982] 4. Specific examples of requests for detailed information

[0983] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[0984] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[0985] 5. Specific examples of providing detailed information

[0986] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[0987] The server identifies the details in the second paragraph and sends them to the terminal.

[0988] 6. Specific examples of outputting detailed information

[0989] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[0990] Users can view more detailed information and gain a deeper understanding of the report's contents.

[0991] In this way, users can efficiently grasp the entire document and easily check the details of the parts they need. This system is extremely convenient for users with visual impairments.

[0992] The following describes the processing flow.

[0993] Step 1:

[0994] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[0995] Step 2:

[0996] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[0997] Step 3:

[0998] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[0999] Step 4:

[1000] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[1001] Step 5:

[1002] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[1003] Step 6:

[1004] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in one of the following formats: Braille display, speech synthesis, or sign language video.

[1005] Step 7:

[1006] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[1007] Step 8:

[1008] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[1009] Step 9:

[1010] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[1011] Step 10:

[1012] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[1013] Step 11:

[1014] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[1015] Step 12:

[1016] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[1017] This series of processing steps allows users to easily review a document summary and, if more detailed information is needed, efficiently retrieve specific details from particular sections.

[1018] (Example 1)

[1019] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1020] There is a need for a system that allows visually impaired users to efficiently understand the content of documents and easily access detailed information on the parts they need. However, conventional systems require a great deal of effort and time to fully understand the content of documents, which is a significant burden on users. In particular, there were technical challenges in providing summaries and detailed information in the required format.

[1021] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1022] In this invention, the server includes an image recognition means, an optical character recognition means, and a summary generation means. This makes it possible for a visually impaired user to take an image of a document, extract text from the image, efficiently summarize the extracted text, and present it to the user in an appropriate format.

[1023] "Image recognition means" refers to technology that identifies images of documents taken by the user using the camera on their device and extracts the necessary information.

[1024] "Optical character recognition means" refers to a technology that analyzes characters within a captured image and extracts them as digital text.

[1025] A "summary generation method" is a technology that analyzes extracted text, extracts important information, and generates a concise summary.

[1026] The "output format selection means" is a technology that outputs the generated summary in an appropriate format (e.g., audio, braille display, sign language video) based on the user's settings.

[1027] A "summary presentation method" is a technology that presents a summary to the user in a selected format.

[1028] A "means for requesting detailed information" is a technology that allows a user to specify detailed information about a summary section that interests them and communicate that request to the system.

[1029] "Means of providing detailed information" refers to a technology in which a server searches for and provides detailed information on a specified part based on a user's request.

[1030] A "means for presenting detailed information" refers to a technology that presents detailed information to the user in a selected output format.

[1031] Modes for carrying out the invention

[1032] This invention is a system for visually impaired users to efficiently understand the contents of documents. This system utilizes multiple hardware and software components.

[1033] 1. Image recognition and text extraction

[1034] The user takes a picture of the document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology. Software used includes "Google Cloud Vision API" and "Adobe OCR".

[1035] Specific example: A user takes a picture of a document using their device's camera to read a report they need for work. The device sends the captured image to an OCR module, which extracts the text from the report.

[1036] 2. Summary generation

[1037] The terminal sends the extracted text to the server.

[1038] The server analyzes the received text, extracts key information using a summarization algorithm, and generates a summary. Suitable generative AI models to use include "BERT" and "GPT-4".

[1039] Specific example: The terminal sends the extracted report text to the server. The server analyzes the text, extracts only the important points, and generates a summary. The server then sends the generated summary to the terminal.

[1040] 3. Selection of output format and presentation of summary

[1041] The device checks the user's settings and outputs a summary in formats such as Braille display, text-to-speech, and sign language video. Text-to-speech uses "Amazon Polly" or "Google Text-to-Speech."

[1042] Specific example: The device reads a summary aloud in text-to-speech format based on the user's settings. The user then reviews important parts of the report by voice.

[1043] 4. Request for detailed information

[1044] Users can specify the summary sections they are interested in using gestures or device operations.

[1045] The device requests detailed information about the specified area from the server. Gesture recognition uses technologies such as "Microsoft Kinect" or "Leap Motion".

[1046] Specific example: A user listens to the audio, feels particularly interested in the second paragraph, and wants more detailed information. They then use a gesture to indicate their interest in the second paragraph to the device.

[1047] 5. Providing detailed information

[1048] The server searches for the requested details and sends them to the terminal. When the server generates additional information based on the text of the specified paragraph, it uses the "GPT-4" AI model.

[1049] Specific example: The terminal recognizes the user's action and requests detailed information about the specified second paragraph from the server. The server identifies the detailed information about the second paragraph and sends it to the terminal.

[1050] 6. Output of detailed information

[1051] The terminal then presents detailed information to the user again in a selected format, such as Braille display, speech synthesis, or sign language video.

[1052] Review the detailed information provided by the user.

[1053] Specific example: The device reads aloud the detailed information from the second paragraph using speech synthesis. The user can then review the more detailed information and gain a deeper understanding of the report's content.

[1054] Examples of prompt statements

[1055] "Extract text from documents photographed using Google Cloud Vision, and then summarize that text using GPT-4."

[1056] "Summarize the extracted text and use Amazon Polly to have it read aloud."

[1057] "Provide details about the paragraphs that the user is interested in, and then use Amazon Polly to convert those details into speech and present them."

[1058] In this way, the system provides a way for visually impaired users to efficiently grasp the entire document and easily check the details of the parts they need.

[1059] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1060] Program processing flow

[1061] Step 1:

[1062] The user takes a picture of a document using the device's camera. The input is an image of the document, and the output is image data saved on the device.

[1063] Specific action: The user presses the "Capture" button on the device.

[1064] Step 2:

[1065] The device takes a picture and passes it to an OCR module for processing. The input is image data, and the output is text data extracted by OCR.

[1066] Specific operation: The device sends image data to the "Optical Character Recognition API".

[1067] Step 3:

[1068] The terminal receives text from the OCR module and sends it to the server. The input is text data, and the output is a transfer to the server.

[1069] Specific action: The device uses an "HTTPS request" to send text data to the server.

[1070] Step 4:

[1071] The server parses the received text and executes a summary generation algorithm. The input is text data, and the output is summarized text.

[1072] Specific operation: The server generates a summary using "natural language processing technology".

[1073] Step 5:

[1074] The server generates a summary and sends it to the terminal. The input is the summary text, and the output is the transfer to the terminal.

[1075] Specific operation: The server sends summary data to the terminal via an "HTTPS response".

[1076] Step 6:

[1077] The terminal checks the user's settings and presents the summary in the selected format. Input is the summary text and user settings information, and output is in the form of audio, Braille display, sign language video, etc.

[1078] Specific action: The device executes the "Speech Synthesis API" or another selected output method.

[1079] Step 7:

[1080] The user reviews the provided summary and specifies the areas for which they want more detailed information. The input is an audio summary, and the output is a request for more detailed information.

[1081] Specific action: The user requests more information using gestures or input devices.

[1082] Step 8:

[1083] The terminal recognizes user input and requests detailed information about the specified portion from the server. The input is a request for detailed information, and the output is the transfer of the request for detailed information to the server.

[1084] Specific action: The terminal sends an "HTTPS request" and forwards a prompt message to the server requesting more information.

[1085] Step 9:

[1086] The server prepares the requested details and sends them to the terminal. The input is the request for details, and the output is the details.

[1087] Specific operation: The server generates the specified content based on "natural language processing technology" and sends it to the terminal.

[1088] Step 10:

[1089] The terminal then presents the detailed information to the user again in the selected format. The input is detailed information, and the output is in the form of audio, braille display, sign language video, etc.

[1090] Specific action: The device executes the "Speech Synthesis API" or another selected output method and presents detailed information.

[1091] keyword

[1092] Generative AI model, prompt sentence

[1093] (Application Example 1)

[1094] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1095] This solution addresses the challenge faced by visually impaired users who lack the efficient means to quickly and accurately grasp the contents of documents such as monitoring reports in real time, and to easily access detailed information as needed.

[1096] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1097] In this invention, the server includes image recognition means, text extraction means, summary generation means, and speech synthesis means. This makes it possible for visually impaired users to obtain summarized information from documents such as monitoring reports in audio format, and to be provided with even more detailed information in audio format as well.

[1098] "Image recognition means" refers to a technology that uses an image input device such as a camera to analyze captured image data and recognize necessary information.

[1099] "Text extraction means" refers to a technology that extracts character information from image data using optical character recognition (OCR) technology.

[1100] A "summary generation method" is a technology that analyzes extracted text data, extracts important information, and generates a concise summary.

[1101] "Output format selection means" refers to a technology that outputs the generated summary information in formats such as Braille display, speech synthesis, and sign language video, according to the user's settings and preferences.

[1102] A "summary presentation method" is a technology that presents summary information to the user based on the selected output format.

[1103] A "means for requesting detailed information" is a technology that allows a user to specify a part of a summary that interests them and request detailed information about that part.

[1104] "Means of providing detailed information" refers to technology that searches for requested detailed information and provides it to the user.

[1105] "Means for presenting detailed information" refers to technologies that present the provided detailed information to the user in the form of Braille displays, speech synthesis, sign language videos, etc.

[1106] "Speech synthesis means" refers to a technology that converts text data into speech and provides information in audio format.

[1107] "Voice recognition means" refers to technology that analyzes the user's voice input and recognizes its content.

[1108] This invention is a system for visually impaired users to efficiently understand the contents of documents such as papers and surveillance reports. Next, a specific embodiment of the system will be described.

[1109] This system consists of a user-facing terminal and multiple processing modules running on a server. The terminal is equipped with a camera, microphone, and speaker. The main processing includes the following:

[1110] 1. Image recognition and text extraction

[1111] The user takes a picture of a surveillance report or document using the device's camera. The device's camera app captures the image, and that image is converted into text using optical character recognition (OCR) technology. For example, software such as Tesseract OCR can be used.

[1112] 2. Sending text to the server and generating a summary

[1113] The extracted text data is sent to the server in real time. On the server, a generative AI model runs, using text analysis and natural language processing techniques to generate a summary of the key information. Natural language processing algorithms such as BERT are used.

[1114] 3. Selection and presentation of summary output format

[1115] The generated summary is provided as audio using speech synthesis technology, depending on the user's settings. Speech synthesis engines such as Google TTS are used. Users can listen to the summary through their speakers.

[1116] 4. Requesting and providing detailed information

[1117] If a user is interested in a particular summary, they can request more information using speech recognition technology. Speech recognition engines such as Sphinx or Google Speech Recognition are used. The requested information is sent back to the server, searched on the server, and then provided via speech synthesis.

[1118] The following are specific usage scenarios.

[1119] Specific usage scenarios:

[1120] Consider a scenario where a visually impaired security staff member is reviewing a surveillance report. The staff member uses their smartphone camera to photograph the report, and OCR technology extracts the text on the device. This text is sent to a server, where a generative AI model generates a summary. The summary is then provided to the staff member in audio format via a speech synthesis engine. If the staff member is interested in a particular paragraph, they can request further details by communicating this using speech recognition technology, and these details are also provided in audio format.

[1121] Examples of input prompts for a generative AI model:

[1122] Extract the key information from the following text and generate a summary. The theme is a monitoring report:

[1123] (Text from the monitoring report)

[1124] This system allows visually impaired users to quickly understand important information and obtain detailed information as needed.

[1125] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1126] Step 1:

[1127] The user takes a picture of the monitoring report with their camera. The device's camera app captures the image, and that image is converted into text using optical character recognition technology (such as Tesseract OCR). Specifically, the captured image data is input into an OCR module, and the text data is output.

[1128] Step 2:

[1129] The terminal sends the extracted text data to the server. Here, the text data is packetized in formats such as JSON or XML and sent to the server over the network. The server stores the received text data for analysis.

[1130] Step 3:

[1131] The server analyzes the received text data. First, it uses a natural language processing algorithm (such as a generative AI model like BERT) to extract important information from the text and generate a summary. This summary generation algorithm produces a compact text that extracts only the essential points while eliminating redundant parts.

[1132] Step 4:

[1133] The server generates a summary and sends it to the terminal. The summary text is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received summary data and continues processing based on the output format set by the user.

[1134] Step 5:

[1135] The device processes the received summary text into speech using a text-to-speech engine (such as Google TTS). When summary text is input, the text-to-speech engine converts the text into natural-sounding speech, which is then presented to the user through the speaker.

[1136] Step 6:

[1137] The user listens to an audio summary and requests more detailed information on the parts that interest them through a speech recognition system (such as Google Speech Recognition). Specifically, the user verbally indicates the parts they want to know more about, the microphone captures this, and it is input into the speech recognition engine. The speech recognition engine converts the audio into text, and that text is sent to the server.

[1138] Step 7:

[1139] The server receives a request for user details. The server searches the database for and extracts detailed text information for the requested portion. This extraction process generates detailed text data.

[1140] Step 8:

[1141] The server sends the extracted detailed information to the terminal. The detailed information is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received detailed information and prepares to present it in the user-configured output format.

[1142] Step 9:

[1143] The device receives detailed information and converts it into speech using a speech synthesis engine. Detailed information text is input, the speech synthesis engine converts that text into natural-sounding speech, and it is presented to the user through the speaker.

[1144] This series of processing steps allows visually impaired users to quickly and efficiently understand the contents of monitoring reports.

[1145] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1146] This invention is a system for users with disabilities, such as blindness, to efficiently grasp the contents of documents and to dynamically select and present an appropriate output format based on the user's emotional state. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, detailed information presentation means, and emotion recognition means and emotion engine.

[1147] Program processing and explanation in natural language

[1148] 1. Image recognition and text extraction

[1149] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[1150] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[1151] 2. Summary generation

[1152] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[1153] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[1154] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[1155] 3. Selection of output format and presentation of summary

[1156] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[1157] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[1158] 4. Emotion Recognition and Dynamic Adjustment

[1159] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[1160] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results. For example, if the user is feeling stressed, the system will make the summary more concise and soften the tone of the voice output.

[1161] 5. Requesting and providing detailed information

[1162] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[1163] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[1164] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[1165] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[1166] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[1167] Specific example

[1168] 1. Specific examples of image recognition and text extraction

[1169] Users take pictures of documents using their device's camera in order to read reports they need for work.

[1170] The device sends the captured image to an OCR module, which then extracts the text from the report.

[1171] 2. Specific Examples of Summary Generation

[1172] The terminal sends the extracted report text to the server.

[1173] The server analyzes the text, extracts only the important points, and generates a summary.

[1174] The server sends the generated summary to the terminal.

[1175] 3. Specific examples of output format selection and summary presentation

[1176] The device will read a summary aloud in text-to-speech format based on the user's settings.

[1177] Users can listen to audio to confirm important parts of the report.

[1178] 4. Specific examples of emotion recognition and dynamic regulation

[1179] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[1180] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[1181] 5. Specific examples of requesting and providing detailed information

[1182] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[1183] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[1184] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[1185] The server identifies the details in the second paragraph and sends them to the terminal.

[1186] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[1187] Users can view more detailed information and gain a deeper understanding of the report's contents.

[1188] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[1189] The following describes the processing flow.

[1190] Step 1:

[1191] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[1192] Step 2:

[1193] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[1194] Step 3:

[1195] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[1196] Step 4:

[1197] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[1198] Step 5:

[1199] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[1200] Step 6:

[1201] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[1202] Step 7:

[1203] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[1204] Step 8:

[1205] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures. The device uses its camera and microphone to monitor the user's emotions in real time.

[1206] Step 9:

[1207] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results. For example, if it determines that the user is confused, it changes the output format of the summary and adds more detailed explanations as needed.

[1208] Step 10:

[1209] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[1210] Step 11:

[1211] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[1212] Step 12:

[1213] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[1214] Step 13:

[1215] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[1216] Step 14:

[1217] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[1218] This series of processing steps allows users to easily review document summaries and efficiently obtain details on specific parts they need. Dynamic adjustments based on emotion recognition further enhance user understanding and enable support tailored to individual needs.

[1219] (Example 2)

[1220] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1221] Visually impaired users have difficulty efficiently understanding the content of documents and obtaining detailed information as needed. Furthermore, while presenting document content requires adjustments based on the user's emotional state, no existing systems dynamically perform these adjustments.

[1222] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an image recognition means, a text extraction means, a summary generation means, an output format selection means, a summary presentation means, a detailed information request means, a detailed information provision means, a detailed information presentation means, an emotion recognition means, and an emotion engine. This enables the user to efficiently grasp the contents of a document and to be provided with information in an appropriate format based on their emotional state.

[1223] "Image recognition means" refers to a means for recognizing the contents of a document from an image taken by a user.

[1224] A "text extraction means" is a means for extracting text data from image data recognized by an image recognition means.

[1225] A "summary generation means" is a means for analyzing extracted text data, extracting important information from it, and generating a summary.

[1226] The "output format selection means" is a means for selecting the output format of the generated summary from among options such as braille display, speech synthesis, and sign language video, based on the user's settings.

[1227] "Means for presenting a summary" refers to means for presenting a summary to the user according to the selected output format.

[1228] A "means for requesting detailed information" refers to a means for receiving and processing requests from users who wish to know more detailed information about a particular summary.

[1229] "Detailed information provision means" refers to means for obtaining detailed information from a server about the summary portion identified by the detailed information request means.

[1230] A "means for presenting detailed information" refers to a means for presenting acquired detailed information to the user based on a selected output format.

[1231] "Emotion recognition means" refers to methods for detecting a user's emotional state from their facial expressions, voice, gestures, etc.

[1232] An "emotion engine" is a means of analyzing the detected emotional state of the user and dynamically adjusting the output format based on the results.

[1233] Modes for carrying out the invention

[1234] This invention is a system for enabling visually impaired users to efficiently understand the contents of documents and to provide information in an appropriate format according to their emotional state. This system utilizes the following hardware and software.

[1235] Hardware and software to be used

[1236] Device: This refers to the user's smartphone or tablet. Peripherals such as cameras, microphones, speakers, and braille displays are connected to it.

[1237] Server: Utilizes cloud-based computing resources to perform text analysis, summary generation, and sentiment analysis.

[1238] OCR module: Software that provides optical character recognition technology, such as the Google Cloud Vision API.

[1239] Summarization generation algorithm: Summarization is generated using natural language processing (NLP) techniques such as BERT and GPT-3.

[1240] Text-to-speech software: Software that converts text into speech, such as Amazon Polly.

[1241] Emotion recognition software: Software used to analyze a user's emotional state, such as the Microsoft Azure Emotion API.

[1242] Specific examples of program processing

[1243] Specific examples of image recognition and text extraction

[1244] Users take pictures of documents using their device's camera in order to read reports they need for work.

[1245] The user launches the camera app, frames the document, and taps the "Shoot" button.

[1246] The device sends the captured image to the Google Cloud Vision API, and the text data is extracted.

[1247] The extracted text data is returned to the terminal.

[1248] Example of a prompt:

[1249] Extract text from the captured images and summarize the contents of the report.

[1250] Concrete examples of summary generation

[1251] The terminal sends the extracted report text to the server.

[1252] Text data is sent to the server via the network.

[1253] The server uses BERT or GPT-3 to extract only the important points and generate a summary.

[1254] The generated summary is sent to the terminal.

[1255] Example of a prompt:

[1256] Please summarize this text data, focusing only on the key points.

[1257] Examples of selecting output formats and presenting summaries

[1258] The device will read a summary aloud in text-to-speech format based on the user's settings.

[1259] The device calls Amazon Polly and converts the summary into speech.

[1260] The synthesized speech summary is played back to the user.

[1261] Example of a prompt:

[1262] Please read this summary aloud using text-to-speech.

[1263] Examples of emotion recognition and dynamic regulation

[1264] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[1265] The device uses the Microsoft Azure Emotion API to analyze the user's emotions.

[1266] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[1267] Example of a prompt:

[1268] If the user is confused, adjust the summary accordingly.

[1269] Examples of requesting and providing detailed information

[1270] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[1271] The user taps the screen to indicate interest in the second paragraph.

[1272] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[1273] The server identifies the details in the second paragraph and sends them to the terminal.

[1274] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[1275] Example of a prompt:

[1276] Please retrieve detailed information about the paragraph specified by the user and provide it in audio format.

[1277] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[1278] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1279] Step 1:

[1280] The user launches the camera app on their device and takes a picture of a physical document.

[1281] Action: The user opens the camera app, frames the document, and taps the "Shoot" button.

[1282] Input: Image of a document.

[1283] Output: Captured image data.

[1284] Step 2:

[1285] The device sends the captured image to an OCR (Optical Character Recognition) module, which then extracts the text data.

[1286] Operation: The device sends image data to the Google Cloud Vision API, and the OCR module analyzes and extracts text data from the image.

[1287] Input: Captured image data.

[1288] Output: Extracted text data.

[1289] Step 3:

[1290] The terminal sends the extracted text data to the server.

[1291] Operation: The device sends text data to the server over the network.

[1292] Input: Extracted text data.

[1293] Output: Text data sent to the server.

[1294] Step 4:

[1295] The server processes the received text data through a summarization algorithm.

[1296] Operation: The server launches an NLP model (e.g., BERT or GPT-3), analyzes the text data, extracts important information, and generates a summary.

[1297] Input: Text data.

[1298] Output: Generated summary data.

[1299] Step 5:

[1300] The server sends the generated summary back to the terminal.

[1301] Operation: The server generates summary data and sends it to the terminal via the network.

[1302] Input: Generated summary data.

[1303] Output: Summary data sent to the terminal.

[1304] Step 6:

[1305] The terminal checks the user's settings and selects the output format.

[1306] Operation: The terminal checks the settings menu and confirms the output format selected by the user (synthesized speech, braille display, sign language video).

[1307] Input: User settings information.

[1308] Output: Selected output format.

[1309] Step 7:

[1310] The device presents the summary to the user in the format selected by the terminal.

[1311] Operation: For example, if the speech synthesis format is selected, the device will call Amazon Polly to convert the summarized text into speech and play it back.

[1312] Input: Summary data.

[1313] Output: Summary presented via speech synthesis.

[1314] Step 8:

[1315] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[1316] Operation: The device uses its camera and microphone to monitor the user's facial expressions and voice, and uses the Microsoft Azure Emotion API to analyze emotions.

[1317] Input: User's facial expressions and voice.

[1318] Output: Detected emotion data.

[1319] Step 9:

[1320] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results.

[1321] Operation: The emotion engine analyzes the detected emotion data and adjusts the output format of the summary as needed (e.g., making the summary more concise and softening the voice tone).

[1322] Input: Sentiment data.

[1323] Output: A summary of the adjusted output format.

[1324] Step 10:

[1325] The user specifies the area they want to know more about.

[1326] Action: The user can use gestures or device operations to highlight the summary section they are interested in.

[1327] Input: User gestures or actions.

[1328] Output: The specified summary portion.

[1329] Step 11:

[1330] The terminal recognizes the user's action and requests detailed information about the specified summary from the server.

[1331] Action: The terminal requests detailed information about the specified summary from the server.

[1332] Input: The specified summary portion.

[1333] Output: Request for more information.

[1334] Step 12:

[1335] The server receives a request for detailed information and searches for detailed information about the specified summary section.

[1336] Operation: The server searches the database and identifies relevant details.

[1337] Input: Request for detailed information.

[1338] Output: Relevant details.

[1339] Step 13:

[1340] The server sends detailed information to the terminal.

[1341] Operation: The server sends detailed information to the terminal over the network.

[1342] Input: Detailed information.

[1343] Output: Detailed information sent to the terminal.

[1344] Step 14:

[1345] The terminal then presents the detailed information received from the server to the user again, based on the output format.

[1346] Operation: For example, if the Braille display is selected, the terminal will display detailed information on the Braille display.

[1347] Input: Detailed information.

[1348] Output: Detailed information presented in output format.

[1349] (Application Example 2)

[1350] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1351] It is difficult for visually impaired users to efficiently grasp product information in physical stores, and there is a lack of systems that provide information in an appropriate format according to the user's emotional state. Therefore, it is necessary to provide an environment in physical stores where users can enjoy shopping with peace of mind, and to improve the efficiency of information access and understanding.

[1352] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1353] In this invention, the server includes image recognition means, text extraction means, and summary generation means. This allows for efficient and effective understanding of product information by extracting product information from images taken by visually impaired users and providing that information as a summary. Furthermore, by using emotion recognition means and emotion engine means, dynamic information presentation according to the user's emotional state becomes possible, allowing users to enjoy shopping with peace of mind.

[1354] An "image recognition system" is a system that analyzes image data acquired by a camera and extracts specific information or features from it.

[1355] "Text extraction means" refers to a function that extracts character information contained within an image as text data using optical character recognition (OCR) technology.

[1356] A "summary generation method" is a system that analyzes extracted text data, extracts only the important information, and summarizes it concisely.

[1357] An "output format selection mechanism" is a system for selecting the format of the information to be output according to the user's settings and circumstances.

[1358] "Means of presenting the summary" refers to means of notifying the user of the generated summary information, and includes various methods such as audio, braille display, and sign language video.

[1359] A "detailed information request system" is a system that accepts requests from users who want to know more detailed information than what is presented in the summary.

[1360] A "detailed information provision means" is a system that obtains and provides additional detailed information based on requests received through a detailed information request means.

[1361] A "means for presenting detailed information" refers to a means of presenting the provided detailed information to the user.

[1362] An "emotion recognition system" is a system that identifies a user's emotional state from their facial expressions, tone of voice, gestures, etc.

[1363] An "emotion engine means" is a system that analyzes emotional information detected by an emotion recognition means and dynamically adjusts the output format and information content based on that analysis.

[1364] To implement this invention, it is first necessary to prepare a terminal equipped with an image recognition means. When a user takes a picture of a product label using the camera of this terminal, the terminal analyzes the image data using the image recognition means and extracts the text information within the label using the text extraction means. By using optical character recognition (OCR) technology, it is possible to accurately obtain the text information within the image as text data. For this technology, open-source OCR libraries such as "Tesseract" can be used.

[1365] Next, the extracted text data is sent to a server, where a summary generation mechanism uses text analysis and natural language processing (NLP) algorithms to summarize the important information. The server returns the generated summary to the terminal, which uses an output format selection mechanism to select an output format according to the user's settings and circumstances. These output formats include audio, braille display, and sign language video. For audio output, "gTTS (Google Text-to-Speech)" and other formats can be used.

[1366] When a summary is presented to the user, an emotion recognition system identifies the user's emotional state from their facial expressions, tone of voice, gestures, etc. An emotion engine analyzes this emotional information and dynamically adjusts the output format and content. For example, if the system determines that the user is confused, it makes the summary more concise and softens the tone of the voice output to aid the user's understanding.

[1367] Furthermore, if a user listens to or reads a summary and requires specific details, they can specify that information using the details request mechanism. The terminal requests detailed information for the specified portion from the server, and the server retrieves the relevant details and sends them to the terminal. The terminal then presents the information back to the user via the details provision mechanism. This process allows the user to efficiently and flexibly grasp product information.

[1368] For example, consider a scenario where a user wants to scan a product label in a supermarket. The user takes a picture of the label with their smartphone camera. The device uses OCR technology to extract the text from the label, and a summary is generated on the server. The summary is then returned to the device and presented to the user as audio output. If the emotion engine detects the user's confusion based on their facial expression and tone of voice, the summary becomes more concise and the audio tone softens. If the user shows interest in specific information, additional details are provided.

[1369] Example of a prompt:

[1370] Please briefly summarize the following text:

[1371] Product name: The product in question

[1372] Price: 1000 yen

[1373] Ingredients: Water, sugar, salt

[1374] Manufacturer: That company

[1375] Country of origin: Japan

[1376] Summary: The product in question can be purchased for 1000 yen. Its ingredients include water, sugar, and salt. It is manufactured by the company mentioned, and its country of origin is Japan.

[1377] This allows visually impaired users to efficiently understand product information even in physical stores, enabling them to enjoy shopping with peace of mind.

[1378] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1379] Step 1:

[1380] The user takes a picture of the product label with the device's camera. The user's input is image data, and the device captures and saves this image data. This image data is then processed in the following steps.

[1381] Step 2:

[1382] The device sends the captured image to the OCR module. The OCR module uses optical character recognition technology to extract text information from the image. The input is image data, and the output is extracted text data. This text data is used for summary generation in the next step.

[1383] Step 3:

[1384] The terminal sends text data acquired from the OCR module to the server. The server processes the received text data using a natural language processing algorithm. The input is text data, and the output is generated summary data. The server generates the summary data and sends it back to the terminal.

[1385] Step 4:

[1386] The terminal receives summary data from the server and converts it to the appropriate output format based on user settings. For example, if audio output is selected, the summary text is converted to audio. The input is the summary data and user settings, and the output is audio data.

[1387] Step 5:

[1388] The device presents the user with summary information in an output format. In the case of audio, the device reads the summary aloud. In this step, the input is audio data or other output format data, and the output is the information presented to the user.

[1389] Step 6:

[1390] The device activates an emotion recognition mechanism to detect emotions from the user's facial expressions, tone of voice, and gestures. The input is data of the user's facial expressions and voice, and the output is the analyzed emotion information. The emotion recognition mechanism transmits this to the emotion engine.

[1391] Step 7:

[1392] The emotion engine analyzes the user's emotions and dynamically adjusts the output format and summary content based on the results. The input is emotional information, and the output is the adjusted summary content and output format. For example, if the user is confused, the summary content will be made simpler and the voice tone will be softened.

[1393] Step 8:

[1394] After the user has heard the summary, if they require specific details, they can use the details request mechanism to provide instructions. The input is the user's request for details, and the output is the specific details requested.

[1395] Step 9:

[1396] The terminal sends a specified request for detailed information to the server. The server searches for and retrieves the necessary detailed information. The input is the request for detailed information, and the output is the detailed information data.

[1397] Step 10:

[1398] The server sends detailed information to the terminal, and the terminal converts the detailed information into an output format and presents it to the user. The input consists of detailed information data and output format settings, and the output is the final presentation of the detailed information.

[1399] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1400] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1401] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1402] [Fourth Embodiment]

[1403] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1404] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1405] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1406] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1407] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1408] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1409] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1410] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1411] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1412] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1413] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1414] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1415] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1416] This invention is a system for users with disabilities, such as blindness, to efficiently understand the contents of documents. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, and detailed information presentation means.

[1417] Program processing and explanation in natural language

[1418] 1. Image recognition and text extraction

[1419] The user takes a picture of a document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology.

[1420] 2. Summary generation

[1421] The terminal sends the extracted text to the server.

[1422] The server analyzes the received text, extracts important information using a summarization algorithm, and generates a summary.

[1423] 3. Selection of output format and presentation of summary

[1424] The terminal checks the user's settings and outputs a summary in formats such as Braille display, speech synthesis, and sign language video.

[1425] 4. Request for detailed information

[1426] The user can specify the summary section they are interested in using gestures or device operations.

[1427] The terminal requests detailed information about the specified part from the server.

[1428] 5. Providing detailed information

[1429] The server searches for the requested details and sends them to the terminal.

[1430] 6. Output of detailed information

[1431] The device then presents detailed information to the user again in a selected format, such as braille display, speech synthesis, or sign language video.

[1432] Specific example

[1433] 1. Specific examples of image recognition and text extraction

[1434] Users take pictures of documents using their device's camera in order to read reports they need for work.

[1435] The device sends the captured image to an OCR module, which then extracts the text from the report.

[1436] 2. Specific Examples of Summary Generation

[1437] The terminal sends the extracted report text to the server.

[1438] The server analyzes the text, extracts only the important points, and generates a summary.

[1439] The server sends the generated summary to the terminal.

[1440] 3. Specific examples of output format selection and summary presentation

[1441] The device will read a summary aloud in text-to-speech format based on the user's settings.

[1442] Users can listen to audio to confirm important parts of the report.

[1443] 4. Specific examples of requests for detailed information

[1444] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[1445] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[1446] 5. Specific examples of providing detailed information

[1447] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[1448] The server identifies the details in the second paragraph and sends them to the terminal.

[1449] 6. Specific examples of outputting detailed information

[1450] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[1451] Users can view more detailed information and gain a deeper understanding of the report's contents.

[1452] In this way, users can efficiently grasp the entire document and easily check the details of the parts they need. This system is extremely convenient for users with visual impairments.

[1453] The following describes the processing flow.

[1454] Step 1:

[1455] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[1456] Step 2:

[1457] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[1458] Step 3:

[1459] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[1460] Step 4:

[1461] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[1462] Step 5:

[1463] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[1464] Step 6:

[1465] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in one of the following formats: Braille display, speech synthesis, or sign language video.

[1466] Step 7:

[1467] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[1468] Step 8:

[1469] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[1470] Step 9:

[1471] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[1472] Step 10:

[1473] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[1474] Step 11:

[1475] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[1476] Step 12:

[1477] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[1478] This series of processing steps allows users to easily review a document summary and, if more detailed information is needed, efficiently retrieve specific details from particular sections.

[1479] (Example 1)

[1480] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1481] There is a need for a system that allows visually impaired users to efficiently understand the content of documents and easily access detailed information on the parts they need. However, conventional systems require a great deal of effort and time to fully understand the content of documents, which is a significant burden on users. In particular, there were technical challenges in providing summaries and detailed information in the required format.

[1482] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1483] In this invention, the server includes an image recognition means, an optical character recognition means, and a summary generation means. This makes it possible for a visually impaired user to take an image of a document, extract text from the image, efficiently summarize the extracted text, and present it to the user in an appropriate format.

[1484] "Image recognition means" refers to technology that identifies images of documents taken by the user using the camera on their device and extracts the necessary information.

[1485] "Optical character recognition means" refers to a technology that analyzes characters within a captured image and extracts them as digital text.

[1486] A "summary generation method" is a technology that analyzes extracted text, extracts important information, and generates a concise summary.

[1487] The "output format selection means" is a technology that outputs the generated summary in an appropriate format (e.g., audio, braille display, sign language video) based on the user's settings.

[1488] A "summary presentation method" is a technology that presents a summary to the user in a selected format.

[1489] A "means for requesting detailed information" is a technology that allows a user to specify detailed information about a summary section that interests them and communicate that request to the system.

[1490] "Means of providing detailed information" refers to a technology in which a server searches for and provides detailed information on a specified part based on a user's request.

[1491] A "means for presenting detailed information" refers to a technology that presents detailed information to the user in a selected output format.

[1492] Modes for carrying out the invention

[1493] This invention is a system for visually impaired users to efficiently understand the contents of documents. This system utilizes multiple hardware and software components.

[1494] 1. Image recognition and text extraction

[1495] The user takes a picture of the document using the device's camera. This image is processed on the device, and text is extracted using OCR (Optical Character Recognition) technology. Software used includes "Google Cloud Vision API" and "Adobe OCR".

[1496] Specific example: A user takes a picture of a document using their device's camera to read a report they need for work. The device sends the captured image to an OCR module, which extracts the text from the report.

[1497] 2. Summary generation

[1498] The terminal sends the extracted text to the server.

[1499] The server analyzes the received text, extracts key information using a summarization algorithm, and generates a summary. Suitable generative AI models to use include "BERT" and "GPT-4".

[1500] Specific example: The terminal sends the extracted report text to the server. The server analyzes the text, extracts only the important points, and generates a summary. The server then sends the generated summary to the terminal.

[1501] 3. Selection of output format and presentation of summary

[1502] The device checks the user's settings and outputs a summary in formats such as Braille display, text-to-speech, and sign language video. Text-to-speech uses "Amazon Polly" or "Google Text-to-Speech."

[1503] Specific example: The device reads a summary aloud in text-to-speech format based on the user's settings. The user then reviews important parts of the report by voice.

[1504] 4. Request for detailed information

[1505] Users can specify the summary sections they are interested in using gestures or device operations.

[1506] The device requests detailed information about the specified area from the server. Gesture recognition uses technologies such as "Microsoft Kinect" or "Leap Motion".

[1507] Specific example: A user listens to the audio, feels particularly interested in the second paragraph, and wants more detailed information. They then use a gesture to indicate their interest in the second paragraph to the device.

[1508] 5. Providing detailed information

[1509] The server searches for the requested details and sends them to the terminal. When the server generates additional information based on the text of the specified paragraph, it uses the "GPT-4" AI model.

[1510] Specific example: The terminal recognizes the user's action and requests detailed information about the specified second paragraph from the server. The server identifies the detailed information about the second paragraph and sends it to the terminal.

[1511] 6. Output of detailed information

[1512] The terminal then presents detailed information to the user again in a selected format, such as Braille display, speech synthesis, or sign language video.

[1513] Review the detailed information provided by the user.

[1514] Specific example: The device reads aloud the detailed information from the second paragraph using speech synthesis. The user can then review the more detailed information and gain a deeper understanding of the report's content.

[1515] Examples of prompt statements

[1516] "Extract text from documents photographed using Google Cloud Vision, and then summarize that text using GPT-4."

[1517] "Summarize the extracted text and use Amazon Polly to have it read aloud."

[1518] "Provide details about the paragraphs that the user is interested in, and then use Amazon Polly to convert those details into speech and present them."

[1519] In this way, the system provides a way for visually impaired users to efficiently grasp the entire document and easily check the details of the parts they need.

[1520] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1521] Program processing flow

[1522] Step 1:

[1523] The user takes a picture of a document using the device's camera. The input is an image of the document, and the output is image data saved on the device.

[1524] Specific action: The user presses the "Capture" button on the device.

[1525] Step 2:

[1526] The device takes a picture and passes it to an OCR module for processing. The input is image data, and the output is text data extracted by OCR.

[1527] Specific operation: The device sends image data to the "Optical Character Recognition API".

[1528] Step 3:

[1529] The terminal receives text from the OCR module and sends it to the server. The input is text data, and the output is a transfer to the server.

[1530] Specific action: The device uses an "HTTPS request" to send text data to the server.

[1531] Step 4:

[1532] The server parses the received text and executes a summary generation algorithm. The input is text data, and the output is summarized text.

[1533] Specific operation: The server generates a summary using "natural language processing technology".

[1534] Step 5:

[1535] The server generates a summary and sends it to the terminal. The input is the summary text, and the output is the transfer to the terminal.

[1536] Specific operation: The server sends summary data to the terminal via an "HTTPS response".

[1537] Step 6:

[1538] The terminal checks the user's settings and presents the summary in the selected format. Input is the summary text and user settings information, and output is in the form of audio, Braille display, sign language video, etc.

[1539] Specific action: The device executes the "Speech Synthesis API" or another selected output method.

[1540] Step 7:

[1541] The user reviews the provided summary and specifies the areas for which they want more detailed information. The input is an audio summary, and the output is a request for more detailed information.

[1542] Specific action: The user requests more information using gestures or input devices.

[1543] Step 8:

[1544] The terminal recognizes user input and requests detailed information about the specified portion from the server. The input is a request for detailed information, and the output is the transfer of the request for detailed information to the server.

[1545] Specific action: The terminal sends an "HTTPS request" and forwards a prompt message to the server requesting more information.

[1546] Step 9:

[1547] The server prepares the requested details and sends them to the terminal. The input is the request for details, and the output is the details.

[1548] Specific operation: The server generates the specified content based on "natural language processing technology" and sends it to the terminal.

[1549] Step 10:

[1550] The terminal then presents the detailed information to the user again in the selected format. The input is detailed information, and the output is in the form of audio, braille display, sign language video, etc.

[1551] Specific action: The device executes the "Speech Synthesis API" or another selected output method and presents detailed information.

[1552] keyword

[1553] Generative AI model, prompt sentence

[1554] (Application Example 1)

[1555] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1556] This solution addresses the challenge faced by visually impaired users who lack the efficient means to quickly and accurately grasp the contents of documents such as monitoring reports in real time, and to easily access detailed information as needed.

[1557] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1558] In this invention, the server includes image recognition means, text extraction means, summary generation means, and speech synthesis means. This makes it possible for visually impaired users to obtain summarized information from documents such as monitoring reports in audio format, and to be provided with even more detailed information in audio format as well.

[1559] "Image recognition means" refers to a technology that uses an image input device such as a camera to analyze captured image data and recognize necessary information.

[1560] "Text extraction means" refers to a technology that extracts character information from image data using optical character recognition (OCR) technology.

[1561] A "summary generation method" is a technology that analyzes extracted text data, extracts important information, and generates a concise summary.

[1562] "Output format selection means" refers to a technology that outputs the generated summary information in formats such as Braille display, speech synthesis, and sign language video, according to the user's settings and preferences.

[1563] A "summary presentation method" is a technology that presents summary information to the user based on the selected output format.

[1564] A "means for requesting detailed information" is a technology that allows a user to specify a part of a summary that interests them and request detailed information about that part.

[1565] "Means of providing detailed information" refers to technology that searches for requested detailed information and provides it to the user.

[1566] "Means for presenting detailed information" refers to technologies that present the provided detailed information to the user in the form of Braille displays, speech synthesis, sign language videos, etc.

[1567] "Speech synthesis means" refers to a technology that converts text data into speech and provides information in audio format.

[1568] "Voice recognition means" refers to technology that analyzes the user's voice input and recognizes its content.

[1569] This invention is a system for visually impaired users to efficiently understand the contents of documents such as papers and surveillance reports. Next, a specific embodiment of the system will be described.

[1570] This system consists of a user-facing terminal and multiple processing modules running on a server. The terminal is equipped with a camera, microphone, and speaker. The main processing includes the following:

[1571] 1. Image recognition and text extraction

[1572] The user takes a picture of a surveillance report or document using the device's camera. The device's camera app captures the image, and that image is converted into text using optical character recognition (OCR) technology. For example, software such as Tesseract OCR can be used.

[1573] 2. Sending text to the server and generating a summary

[1574] The extracted text data is sent to the server in real time. On the server, a generative AI model runs, using text analysis and natural language processing techniques to generate a summary of the key information. Natural language processing algorithms such as BERT are used.

[1575] 3. Selection and presentation of summary output format

[1576] The generated summary is provided as audio using speech synthesis technology, depending on the user's settings. Speech synthesis engines such as Google TTS are used. Users can listen to the summary through their speakers.

[1577] 4. Requesting and providing detailed information

[1578] If a user is interested in a particular summary, they can request more information using speech recognition technology. Speech recognition engines such as Sphinx or Google Speech Recognition are used. The requested information is sent back to the server, searched on the server, and then provided via speech synthesis.

[1579] The following are specific usage scenarios.

[1580] Specific usage scenarios:

[1581] Consider a scenario where a visually impaired security staff member is reviewing a surveillance report. The staff member uses their smartphone camera to photograph the report, and OCR technology extracts the text on the device. This text is sent to a server, where a generative AI model generates a summary. The summary is then provided to the staff member in audio format via a speech synthesis engine. If the staff member is interested in a particular paragraph, they can request further details by communicating this using speech recognition technology, and these details are also provided in audio format.

[1582] Examples of input prompts for a generative AI model:

[1583] Extract the key information from the following text and generate a summary. The theme is a monitoring report:

[1584] (Text from the monitoring report)

[1585] This system allows visually impaired users to quickly understand important information and obtain detailed information as needed.

[1586] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1587] Step 1:

[1588] The user takes a picture of the monitoring report with their camera. The device's camera app captures the image, and that image is converted into text using optical character recognition technology (such as Tesseract OCR). Specifically, the captured image data is input into an OCR module, and the text data is output.

[1589] Step 2:

[1590] The terminal sends the extracted text data to the server. Here, the text data is packetized in formats such as JSON or XML and sent to the server over the network. The server stores the received text data for analysis.

[1591] Step 3:

[1592] The server analyzes the received text data. First, it uses a natural language processing algorithm (such as a generative AI model like BERT) to extract important information from the text and generate a summary. This summary generation algorithm produces a compact text that extracts only the essential points while eliminating redundant parts.

[1593] Step 4:

[1594] The server generates a summary and sends it to the terminal. The summary text is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received summary data and continues processing based on the output format set by the user.

[1595] Step 5:

[1596] The device processes the received summary text into speech using a text-to-speech engine (such as Google TTS). When summary text is input, the text-to-speech engine converts the text into natural-sounding speech, which is then presented to the user through the speaker.

[1597] Step 6:

[1598] The user listens to an audio summary and requests more detailed information on the parts that interest them through a speech recognition system (such as Google Speech Recognition). Specifically, the user verbally indicates the parts they want to know more about, the microphone captures this, and it is input into the speech recognition engine. The speech recognition engine converts the audio into text, and that text is sent to the server.

[1599] Step 7:

[1600] The server receives a request for user details. The server searches the database for and extracts detailed text information for the requested portion. This extraction process generates detailed text data.

[1601] Step 8:

[1602] The server sends the extracted detailed information to the terminal. The detailed information is then packetized again in JSON or XML format and sent to the terminal. The terminal parses the received detailed information and prepares to present it in the user-configured output format.

[1603] Step 9:

[1604] The device receives detailed information and converts it into speech using a speech synthesis engine. Detailed information text is input, the speech synthesis engine converts that text into natural-sounding speech, and it is presented to the user through the speaker.

[1605] This series of processing steps allows visually impaired users to quickly and efficiently understand the contents of monitoring reports.

[1606] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1607] This invention is a system for users with disabilities, such as blindness, to efficiently grasp the contents of documents and to dynamically select and present an appropriate output format based on the user's emotional state. This system includes image recognition means, text extraction means, summary generation means, output format selection means, summary presentation means, detailed information request means, detailed information provision means, detailed information presentation means, and emotion recognition means and emotion engine.

[1608] Program processing and explanation in natural language

[1609] 1. Image recognition and text extraction

[1610] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[1611] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[1612] 2. Summary generation

[1613] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[1614] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[1615] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[1616] 3. Selection of output format and presentation of summary

[1617] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[1618] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[1619] 4. Emotion Recognition and Dynamic Adjustment

[1620] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[1621] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results. For example, if the user is feeling stressed, the system will make the summary more concise and soften the tone of the voice output.

[1622] 5. Requesting and providing detailed information

[1623] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[1624] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[1625] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[1626] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[1627] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[1628] Specific example

[1629] 1. Specific examples of image recognition and text extraction

[1630] Users take pictures of documents using their device's camera in order to read reports they need for work.

[1631] The device sends the captured image to an OCR module, which then extracts the text from the report.

[1632] 2. Specific Examples of Summary Generation

[1633] The terminal sends the extracted report text to the server.

[1634] The server analyzes the text, extracts only the important points, and generates a summary.

[1635] The server sends the generated summary to the terminal.

[1636] 3. Specific examples of output format selection and summary presentation

[1637] The device will read a summary aloud in text-to-speech format based on the user's settings.

[1638] Users can listen to audio to confirm important parts of the report.

[1639] 4. Specific examples of emotion recognition and dynamic regulation

[1640] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[1641] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[1642] 5. Specific examples of requesting and providing detailed information

[1643] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[1644] The user uses gestures to indicate to the device that they are interested in the second paragraph.

[1645] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[1646] The server identifies the details in the second paragraph and sends them to the terminal.

[1647] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[1648] Users can view more detailed information and gain a deeper understanding of the report's contents.

[1649] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[1650] The following describes the processing flow.

[1651] Step 1:

[1652] The user takes a picture of a document using the device's camera. The device activates the camera and captures the image in order for the user to scan the document.

[1653] Step 2:

[1654] The device sends the captured image to an OCR (Optical Character Recognition) module. The OCR module analyzes the text data within the image and returns the extracted text information to the device.

[1655] Step 3:

[1656] The terminal sends the text data acquired from the OCR module to the server. The text data is transferred to the server via the network.

[1657] Step 4:

[1658] The server processes the received text data through a summarization algorithm. The server analyzes the text using natural language processing (NLP) techniques, extracts important information, and creates a summary.

[1659] Step 5:

[1660] The server generates a summary and sends it back to the terminal. The server transmits the summary data to the terminal via the network.

[1661] Step 6:

[1662] The terminal checks the user's settings and selects the output format. Based on the user's settings, the summary is output in formats such as Braille display, speech synthesis, or sign language video.

[1663] Step 7:

[1664] The device presents the summary to the user in the selected format. For example, if text-to-speech is selected, the device will read the summary text aloud.

[1665] Step 8:

[1666] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures. The device uses its camera and microphone to monitor the user's emotions in real time.

[1667] Step 9:

[1668] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results. For example, if it determines that the user is confused, it changes the output format of the summary and adds more detailed explanations as needed.

[1669] Step 10:

[1670] Users listen to or read summaries and specify the parts they want to learn more about. Users use gestures and device controls to indicate the summaries they are interested in.

[1671] Step 11:

[1672] The device recognizes the user's actions and identifies the specified summary section. The device then prepares to request further information based on the user's instructions.

[1673] Step 12:

[1674] The terminal requests detailed information about the specified summary from the server. The terminal sends the detailed information request to the server over the network.

[1675] Step 13:

[1676] The server receives a request for detailed information and searches for details about the specified summary section. The server identifies the relevant details and sends them back to the terminal.

[1677] Step 14:

[1678] The terminal then presents the detailed information received from the server to the user again, based on the output format. For example, if a braille display is selected, the terminal will display the detailed information on the braille display.

[1679] This series of processing steps allows users to easily review document summaries and efficiently obtain details on specific parts they need. Dynamic adjustments based on emotion recognition further enhance user understanding and enable support tailored to individual needs.

[1680] (Example 2)

[1681] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1682] Visually impaired users have difficulty efficiently understanding the content of documents and obtaining detailed information as needed. Furthermore, while presenting document content requires adjustments based on the user's emotional state, no existing systems dynamically perform these adjustments.

[1683] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an image recognition means, a text extraction means, a summary generation means, an output format selection means, a summary presentation means, a detailed information request means, a detailed information provision means, a detailed information presentation means, an emotion recognition means, and an emotion engine. This enables the user to efficiently grasp the contents of a document and to be provided with information in an appropriate format based on their emotional state.

[1684] "Image recognition means" refers to a means for recognizing the contents of a document from an image taken by a user.

[1685] A "text extraction means" is a means for extracting text data from image data recognized by an image recognition means.

[1686] A "summary generation means" is a means for analyzing extracted text data, extracting important information from it, and generating a summary.

[1687] The "output format selection means" is a means for selecting the output format of the generated summary from among options such as braille display, speech synthesis, and sign language video, based on the user's settings.

[1688] "Means for presenting a summary" refers to means for presenting a summary to the user according to the selected output format.

[1689] A "means for requesting detailed information" refers to a means for receiving and processing requests from users who wish to know more detailed information about a particular summary.

[1690] "Detailed information provision means" refers to means for obtaining detailed information from a server about the summary portion identified by the detailed information request means.

[1691] A "means for presenting detailed information" refers to a means for presenting acquired detailed information to the user based on a selected output format.

[1692] "Emotion recognition means" refers to methods for detecting a user's emotional state from their facial expressions, voice, gestures, etc.

[1693] An "emotion engine" is a means of analyzing the detected emotional state of the user and dynamically adjusting the output format based on the results.

[1694] Modes for carrying out the invention

[1695] This invention is a system for enabling visually impaired users to efficiently understand the contents of documents and to provide information in an appropriate format according to their emotional state. This system utilizes the following hardware and software.

[1696] Hardware and software to be used

[1697] Device: This refers to the user's smartphone or tablet. Peripherals such as cameras, microphones, speakers, and braille displays are connected to it.

[1698] Server: Utilizes cloud-based computing resources to perform text analysis, summary generation, and sentiment analysis.

[1699] OCR module: Software that provides optical character recognition technology, such as the Google Cloud Vision API.

[1700] Summarization generation algorithm: Summarization is generated using natural language processing (NLP) techniques such as BERT and GPT-3.

[1701] Text-to-speech software: Software that converts text into speech, such as Amazon Polly.

[1702] Emotion recognition software: Software used to analyze a user's emotional state, such as the Microsoft Azure Emotion API.

[1703] Specific examples of program processing

[1704] Specific examples of image recognition and text extraction

[1705] Users take pictures of documents using their device's camera in order to read reports they need for work.

[1706] The user launches the camera app, frames the document, and taps the "Shoot" button.

[1707] The device sends the captured image to the Google Cloud Vision API, and the text data is extracted.

[1708] The extracted text data is returned to the terminal.

[1709] Example of a prompt:

[1710] Extract text from the captured images and summarize the contents of the report.

[1711] Concrete examples of summary generation

[1712] The terminal sends the extracted report text to the server.

[1713] Text data is sent to the server via the network.

[1714] The server uses BERT or GPT-3 to extract only the important points and generate a summary.

[1715] The generated summary is sent to the terminal.

[1716] Example of a prompt:

[1717] Please summarize this text data, focusing only on the key points.

[1718] Examples of selecting output formats and presenting summaries

[1719] The device will read a summary aloud in text-to-speech format based on the user's settings.

[1720] The device calls Amazon Polly and converts the summary into speech.

[1721] The synthesized speech summary is played back to the user.

[1722] Example of a prompt:

[1723] Please read this summary aloud using text-to-speech.

[1724] Examples of emotion recognition and dynamic regulation

[1725] When the device reads out the summary, it monitors the user's facial expressions and tone of voice using emotion recognition technology.

[1726] The device uses the Microsoft Azure Emotion API to analyze the user's emotions.

[1727] If the sentiment engine determines that the user is confused, it will change the output format of the summary and add more detailed explanations as needed.

[1728] Example of a prompt:

[1729] If the user is confused, adjust the summary accordingly.

[1730] Examples of requesting and providing detailed information

[1731] The user listened to the audio and felt the need for more detailed information on the second paragraph, which they found particularly interesting.

[1732] The user taps the screen to indicate interest in the second paragraph.

[1733] The terminal recognizes the user's action and requests detailed information for the specified second paragraph from the server.

[1734] The server identifies the details in the second paragraph and sends them to the terminal.

[1735] The device then reads aloud the detailed information from the second paragraph using speech synthesis.

[1736] Example of a prompt:

[1737] Please retrieve detailed information about the paragraph specified by the user and provide it in audio format.

[1738] In this way, users can efficiently grasp the entire document and receive appropriate support based on their emotional state, leading to a deeper understanding of the information. This system is highly convenient and flexible for users with visual impairments.

[1739] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1740] Step 1:

[1741] The user launches the camera app on their device and takes a picture of a physical document.

[1742] Action: The user opens the camera app, frames the document, and taps the "Shoot" button.

[1743] Input: Image of a document.

[1744] Output: Captured image data.

[1745] Step 2:

[1746] The device sends the captured image to an OCR (Optical Character Recognition) module, which then extracts the text data.

[1747] Operation: The device sends image data to the Google Cloud Vision API, and the OCR module analyzes and extracts text data from the image.

[1748] Input: Captured image data.

[1749] Output: Extracted text data.

[1750] Step 3:

[1751] The terminal sends the extracted text data to the server.

[1752] Operation: The device sends text data to the server over the network.

[1753] Input: Extracted text data.

[1754] Output: Text data sent to the server.

[1755] Step 4:

[1756] The server processes the received text data through a summarization algorithm.

[1757] Operation: The server launches an NLP model (e.g., BERT or GPT-3), analyzes the text data, extracts important information, and generates a summary.

[1758] Input: Text data.

[1759] Output: Generated summary data.

[1760] Step 5:

[1761] The server sends the generated summary back to the terminal.

[1762] Operation: The server generates summary data and sends it to the terminal via the network.

[1763] Input: Generated summary data.

[1764] Output: Summary data sent to the terminal.

[1765] Step 6:

[1766] The terminal checks the user's settings and selects the output format.

[1767] Operation: The terminal checks the settings menu and confirms the output format selected by the user (synthesized speech, braille display, sign language video).

[1768] Input: User settings information.

[1769] Output: Selected output format.

[1770] Step 7:

[1771] The device presents the summary to the user in the format selected by the terminal.

[1772] Operation: For example, if the speech synthesis format is selected, the device will call Amazon Polly to convert the summarized text into speech and play it back.

[1773] Input: Summary data.

[1774] Output: Summary presented via speech synthesis.

[1775] Step 8:

[1776] The device activates emotion recognition mechanisms to detect emotions from the user's facial expressions, voice, and gestures.

[1777] Operation: The device uses its camera and microphone to monitor the user's facial expressions and voice, and uses the Microsoft Azure Emotion API to analyze emotions.

[1778] Input: User's facial expressions and voice.

[1779] Output: Detected emotion data.

[1780] Step 9:

[1781] The emotion engine analyzes the user's emotions and dynamically adjusts the output format based on the results.

[1782] Operation: The emotion engine analyzes the detected emotion data and adjusts the output format of the summary as needed (e.g., making the summary more concise and softening the voice tone).

[1783] Input: Sentiment data.

[1784] Output: A summary of the adjusted output format.

[1785] Step 10:

[1786] The user specifies the area they want to know more about.

[1787] Action: The user can use gestures or device operations to highlight the summary section they are interested in.

[1788] Input: User gestures or actions.

[1789] Output: The specified summary portion.

[1790] Step 11:

[1791] The terminal recognizes the user's action and requests detailed information about the specified summary from the server.

[1792] Action: The terminal requests detailed information about the specified summary from the server.

[1793] Input: The specified summary portion.

[1794] Output: Request for more information.

[1795] Step 12:

[1796] The server receives a request for detailed information and searches for detailed information about the specified summary section.

[1797] Operation: The server searches the database and identifies relevant details.

[1798] Input: Request for detailed information.

[1799] Output: Relevant details.

[1800] Step 13:

[1801] The server sends detailed information to the terminal.

[1802] Operation: The server sends detailed information to the terminal over the network.

[1803] Input: Detailed information.

[1804] Output: Detailed information sent to the terminal.

[1805] Step 14:

[1806] The terminal then presents the detailed information received from the server to the user again, based on the output format.

[1807] Operation: For example, if the Braille display is selected, the terminal will display detailed information on the Braille display.

[1808] Input: Detailed information.

[1809] Output: Detailed information presented in output format.

[1810] (Application Example 2)

[1811] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1812] It is difficult for visually impaired users to efficiently grasp product information in physical stores, and there is a lack of systems that provide information in an appropriate format according to the user's emotional state. Therefore, it is necessary to provide an environment in physical stores where users can enjoy shopping with peace of mind, and to improve the efficiency of information access and understanding.

[1813] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1814] In this invention, the server includes image recognition means, text extraction means, and summary generation means. This allows for efficient and effective understanding of product information by extracting product information from images taken by visually impaired users and providing that information as a summary. Furthermore, by using emotion recognition means and emotion engine means, dynamic information presentation according to the user's emotional state becomes possible, allowing users to enjoy shopping with peace of mind.

[1815] An "image recognition system" is a system that analyzes image data acquired by a camera and extracts specific information or features from it.

[1816] "Text extraction means" refers to a function that extracts character information contained within an image as text data using optical character recognition (OCR) technology.

[1817] A "summary generation method" is a system that analyzes extracted text data, extracts only the important information, and summarizes it concisely.

[1818] An "output format selection mechanism" is a system for selecting the format of the information to be output according to the user's settings and circumstances.

[1819] "Means of presenting the summary" refers to means of notifying the user of the generated summary information, and includes various methods such as audio, braille display, and sign language video.

[1820] A "detailed information request system" is a system that accepts requests from users who want to know more detailed information than what is presented in the summary.

[1821] A "detailed information provision means" is a system that obtains and provides additional detailed information based on requests received through a detailed information request means.

[1822] A "means for presenting detailed information" refers to a means of presenting the provided detailed information to the user.

[1823] An "emotion recognition system" is a system that identifies a user's emotional state from their facial expressions, tone of voice, gestures, etc.

[1824] An "emotion engine means" is a system that analyzes emotional information detected by an emotion recognition means and dynamically adjusts the output format and information content based on that analysis.

[1825] To implement this invention, it is first necessary to prepare a terminal equipped with an image recognition means. When a user takes a picture of a product label using the camera of this terminal, the terminal analyzes the image data using the image recognition means and extracts the text information within the label using the text extraction means. By using optical character recognition (OCR) technology, it is possible to accurately obtain the text information within the image as text data. For this technology, open-source OCR libraries such as "Tesseract" can be used.

[1826] Next, the extracted text data is sent to a server, where a summary generation mechanism uses text analysis and natural language processing (NLP) algorithms to summarize the important information. The server returns the generated summary to the terminal, which uses an output format selection mechanism to select an output format according to the user's settings and circumstances. These output formats include audio, braille display, and sign language video. For audio output, "gTTS (Google Text-to-Speech)" and other formats can be used.

[1827] When a summary is presented to the user, an emotion recognition system identifies the user's emotional state from their facial expressions, tone of voice, gestures, etc. An emotion engine analyzes this emotional information and dynamically adjusts the output format and content. For example, if the system determines that the user is confused, it makes the summary more concise and softens the tone of the voice output to aid the user's understanding.

[1828] Furthermore, if a user listens to or reads a summary and requires specific details, they can specify that information using the details request mechanism. The terminal requests detailed information for the specified portion from the server, and the server retrieves the relevant details and sends them to the terminal. The terminal then presents the information back to the user via the details provision mechanism. This process allows the user to efficiently and flexibly grasp product information.

[1829] For example, consider a scenario where a user wants to scan a product label in a supermarket. The user takes a picture of the label with their smartphone camera. The device uses OCR technology to extract the text from the label, and a summary is generated on the server. The summary is then returned to the device and presented to the user as audio output. If the emotion engine detects the user's confusion based on their facial expression and tone of voice, the summary becomes more concise and the audio tone softens. If the user shows interest in specific information, additional details are provided.

[1830] Example of a prompt:

[1831] Please briefly summarize the following text:

[1832] Product name: The product in question

[1833] Price: 1000 yen

[1834] Ingredients: Water, sugar, salt

[1835] Manufacturer: That company

[1836] Country of origin: Japan

[1837] Summary: The product in question can be purchased for 1000 yen. Its ingredients include water, sugar, and salt. It is manufactured by the company mentioned, and its country of origin is Japan.

[1838] This allows visually impaired users to efficiently understand product information even in physical stores, enabling them to enjoy shopping with peace of mind.

[1839] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1840] Step 1:

[1841] The user takes a picture of the product label with the device's camera. The user's input is image data, and the device captures and saves this image data. This image data is then processed in the following steps.

[1842] Step 2:

[1843] The device sends the captured image to the OCR module. The OCR module uses optical character recognition technology to extract text information from the image. The input is image data, and the output is extracted text data. This text data is used for summary generation in the next step.

[1844] Step 3:

[1845] The terminal sends text data acquired from the OCR module to the server. The server processes the received text data using a natural language processing algorithm. The input is text data, and the output is generated summary data. The server generates the summary data and sends it back to the terminal.

[1846] Step 4:

[1847] The terminal receives summary data from the server and converts it to the appropriate output format based on user settings. For example, if audio output is selected, the summary text is converted to audio. The input is the summary data and user settings, and the output is audio data.

[1848] Step 5:

[1849] The device presents the user with summary information in an output format. In the case of audio, the device reads the summary aloud. In this step, the input is audio data or other output format data, and the output is the information presented to the user.

[1850] Step 6:

[1851] The device activates an emotion recognition mechanism to detect emotions from the user's facial expressions, tone of voice, and gestures. The input is data of the user's facial expressions and voice, and the output is the analyzed emotion information. The emotion recognition mechanism transmits this to the emotion engine.

[1852] Step 7:

[1853] The emotion engine analyzes the user's emotions and dynamically adjusts the output format and summary content based on the results. The input is emotional information, and the output is the adjusted summary content and output format. For example, if the user is confused, the summary content will be made simpler and the voice tone will be softened.

[1854] Step 8:

[1855] After the user has heard the summary, if they require specific details, they can use the details request mechanism to provide instructions. The input is the user's request for details, and the output is the specific details requested.

[1856] Step 9:

[1857] The terminal sends a specified request for detailed information to the server. The server searches for and retrieves the necessary detailed information. The input is the request for detailed information, and the output is the detailed information data.

[1858] Step 10:

[1859] The server sends detailed information to the terminal, and the terminal converts the detailed information into an output format and presents it to the user. The input consists of detailed information data and output format settings, and the output is the final presentation of the detailed information.

[1860] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1861] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1862] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1863] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1864] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1865] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1866] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1867] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1868] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1869] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1870] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1871] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1872] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1873] 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.

[1874] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1875] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1876] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1877] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1878] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1879] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1880] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[1881] The following is further disclosed regarding the embodiments described above.

[1882] (Claim 1)

[1883] Image recognition means,

[1884] Text extraction method,

[1885] Summary generation means,

[1886] Output format selection means,

[1887] Means of presenting summaries,

[1888] Means for requesting detailed information,

[1889] Means of providing detailed information,

[1890] Means of presenting detailed information,

[1891] A system that includes this.

[1892] (Claim 2)

[1893] The system according to claim 1, wherein the text extraction means extracts text from an image of a document taken by a user using optical character recognition technology.

[1894] (Claim 3)

[1895] The system according to claim 1, wherein the summarization means summarizes important information using text analysis and natural language processing algorithms via a server.

[1896] (Claim 4)

[1897] The system according to claim 1, wherein the output format selection means selects one of the following formats: braille display, speech synthesis, or sign language video.

[1898] (Claim 5)

[1899] The system according to claim 1, wherein the means for providing detailed information again provides detailed information on the summary portion specified by the user.

[1900] "Example 1"

[1901] (Claim 1)

[1902] Image recognition means,

[1903] Optical character recognition means,

[1904] Summary generation means,

[1905] Output format selection means,

[1906] Means of presenting summaries,

[1907] Means for requesting detailed information,

[1908] Means of providing detailed information,

[1909] Means of presenting detailed information,

[1910] A system that includes this.

[1911] (Claim 2)

[1912] The system according to claim 1, wherein the optical character recognition means extracts text from an image of a document taken by a user using optical character recognition technology.

[1913] (Claim 3)

[1914] The system according to claim 1, wherein the summarization generation means analyzes text received by the server using natural language processing technology and summarizes important information.

[1915] "Application Example 1"

[1916] (Claim 1)

[1917] Image recognition means,

[1918] Text extraction method,

[1919] Summary generation means,

[1920] Output format selection means,

[1921] Means of presenting summaries,

[1922] Means for requesting detailed information,

[1923] Means of providing detailed information,

[1924] Means of presenting detailed information,

[1925] A speech synthesis method,

[1926] Voice recognition means and

[1927] A system that includes this.

[1928] (Claim 2)

[1929] The system according to claim 1, wherein the text extraction means extracts text from an image of a document taken by a user using optical character recognition technology.

[1930] (Claim 3)

[1931] The system according to claim 1, wherein the summarization means summarizes important information using text analysis and natural language processing algorithms via a server.

[1932] "Example 2 of combining an emotion engine"

[1933] (Claim 1)

[1934] Image recognition means,

[1935] Text extraction method,

[1936] Summary generation means,

[1937] Output format selection means,

[1938] Means of presenting summaries,

[1939] Means for requesting detailed information,

[1940] Means of providing detailed information,

[1941] Means of presenting detailed information,

[1942] Means of recognizing emotions,

[1943] Emotional engine,

[1944] A system that includes this.

[1945] (Claim 2)

[1946] The system according to claim 1, wherein the text extraction means extracts text from an image of a document taken by a user using optical character recognition technology, and transmits the extracted text data to a server via a network.

[1947] (Claim 3)

[1948] The system according to claim 1, wherein the summarization means uses a server to summarize important information using text analysis and natural language processing algorithms, and transmits the generated summary to a terminal via a network.

[1949] "Application example 2 when combining with an emotional engine"

[1950] (Claim 1)

[1951] Image recognition means,

[1952] Text extraction method,

[1953] Summary generation means,

[1954] Output format selection means,

[1955] Means of presenting summaries,

[1956] Means for requesting detailed information,

[1957] Means of providing detailed information,

[1958] Means of presenting detailed information,

[1959] Means of recognizing emotions,

[1960] Emotional engine means,

[1961] A system that includes this.

[1962] (Claim 2)

[1963] The system according to claim 1, wherein the text extraction means extracts text from an image of content taken by a user using optical character recognition technology.

[1964] (Claim 3)

[1965] The system according to claim 1, wherein the summarization means summarizes important information using text analysis and natural language processing algorithms via a server. [Explanation of Symbols]

[1966] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. An image recognition means that captures images of documents and other materials and analyzes them as digital data, A text extraction means that extracts character information from an image using optical character recognition technology, A summary generation means that analyzes extracted text, extracts important information, and generates a summary, An output format selection means that outputs information in one of the following formats based on user settings: braille display, speech synthesis, or sign language video. A means for presenting summary information to the user in a selected output format, A means for requesting detailed information about a summary section specified by the user, A means of providing detailed information that identifies and provides the requested detailed information, A means for presenting detailed information to the user in a selected output format, A system that includes this.

2. The system according to claim 1, wherein the text extraction means extracts text from an image of a document taken by the user using optical character recognition technology.

3. The system according to claim 1, wherein the summarization means summarizes important information using text analysis and natural language processing algorithms on a server.

4. The system according to claim 1, wherein the output format selection means selects one of the following formats: braille display, speech synthesis, or sign language video.

5. The system according to claim 1, wherein the means for providing detailed information again provides detailed information on the summary portion specified by the user. s

Images