System

Abstract

A system is provided.SOLUTION: The system includes a means for sharing knowledge and technology related to a bias measure through a consortium in which multi-industry organizations participate, a means for designing and implementing a feedback system for providing cultural feedback of a generated image, and a means for giving an authentication mark to a business operator who implements an appropriate bias measure.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Current AI-generated images often contain biases and errors, particularly in Asia. These problems are primarily due to training data collected primarily from Europe and the United States, which is insufficient to accurately reflect Asia's cultural background and diversity. Furthermore, improper use by businesses that lack proper anti-bias measures hinders fair use. Therefore, the purpose of this invention is to reduce bias in AI-generated images in Asia and promote fair use. [Means for solving the problem]

[0005] The present invention solves the above problems by the following means: It establishes a consortium with the participation of institutions from various industries and provides a means for sharing knowledge and technology related to bias countermeasures. It also designs and implements a feedback system that provides cultural feedback on generated images. This system uses an algorithm that analyzes elements within the image and evaluates cultural meaning. It also awards a certification mark to businesses that implement appropriate bias countermeasures. When awarding this certification mark, it includes a means for integrating and analyzing the results of the business's countermeasure evaluation and the feedback system's usage history. In this way, it provides a system that effectively mitigates bias issues in AI-generated images in Asia and promotes fair use.

[0006] A "consortium" is an organization in which institutions from various fields cooperate for a specific purpose and share knowledge and technology.

[0007] "Anti-bias" refers to techniques and measures to mitigate or eliminate inappropriate biases and errors in AI-generated images.

[0008] The "feedback system" is an automated system that provides cultural assessments and corrections to generated AI images.

[0009] An "algorithm" is a mathematical method or set of rules for performing a particular calculation or analysis.

[0010] A "certification mark" is an official mark of recognition given to businesses that implement appropriate anti-bias measures.

[0011] "Image analysis" is the technical process of analyzing a digital image and evaluating its content and elements.

[0012] A "database" is an organized digital storage system for efficiently storing, managing, and retrieving information. [Brief explanation of the drawings]

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

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

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

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

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

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

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

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

[0021] [First embodiment]

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

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

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

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

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

[0027] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

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

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

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

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

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

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

[0034] To implement this invention, the following specific steps are required: First, a consortium is established with the participation of institutions from various industries. The consortium will serve as a forum for sharing knowledge and technology related to bias prevention. Next, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Finally, a system is built to award a certification mark to businesses that implement appropriate bias prevention measures.

[0035] Specific examples of consortium establishment and operation

[0036] Step 1: User Registration

[0037] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[0038] The terminal (company representative) enters the necessary information into the form and completes the registration.

[0039] The server stores the input in a database and sends a confirmation email.

[0040] Step 2: Host a workshop

[0041] Users log in to the consortium web portal and check the event calendar.

[0042] The server displays detailed information about the workshop (title, date and time, instructor information).

[0043] The user registers for the workshop of their choice.

[0044] The server stores the registration information in a database and sends reminder emails.

[0045] Specific examples of feedback system design and implementation

[0046] Step 1: Upload an image

[0047] Users select the AI-generated image they want to receive feedback on from their device.

[0048] The terminal transmits the selected image file to the server.

[0049] The server stores the received images in a database and passes them to the image analysis system.

[0050] Step 2: Generate feedback

[0051] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[0052] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[0053] Step 3: Review and correct your feedback

[0054] The user checks the feedback sent back from the server on the terminal.

[0055] The user makes any necessary corrections based on the feedback and sends the image to the server again.

[0056] Examples of certification marking and management

[0057] Step 1: Evaluate the checklist

[0058] The server will manage anti-bias checklists for each company and make them available for download.

[0059] Users (company representatives) download the checklist and evaluate their company's measures.

[0060] The terminal transmits the self-evaluation result to the server.

[0061] Step 2: Analyze the results

[0062] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0063] The server decides whether to grant a certification mark based on the evaluation results.

[0064] Step 3: Issuance of the certification mark

[0065] The server decides to issue a certification mark to the company to be certified.

[0066] The server transmits the certification mark image data and related information to the target company.

[0067] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[0068] The above is a concrete example of how to implement the present invention. By using this system, we can effectively mitigate bias issues in AI-generated images in Asia and promote fair use.

[0069] The processing flow will be explained below.

[0070] Establishment and operation of the consortium

[0071] Step 1:

[0072] The user enters information such as the company name, the name of the person in charge, and an email address into the new user registration form.

[0073] Step 2:

[0074] The terminal sends the input information to the server.

[0075] Step 3:

[0076] The server stores the submitted information in a database and sends a confirmation email to the user.

[0077] Step 4:

[0078] The user clicks on the link in the confirmation email to complete the registration.

[0079] Step 5:

[0080] The server grants access rights to the consortium web portal.

[0081] Step 6:

[0082] Users log in to a web portal to view the events calendar.

[0083] Step 7:

[0084] The server displays detailed information about the workshop (title, date and time, instructor information).

[0085] Step 8:

[0086] The user registers for the workshop of their choice.

[0087] Step 9:

[0088] The server stores the registration information in a database and sends reminder emails.

[0089] Design and implementation of a feedback system

[0090] Step 1:

[0091] Users select the AI-generated image they want to receive feedback on from their device.

[0092] Step 2:

[0093] The terminal sends the selected image file to the server.

[0094] Step 3:

[0095] The server stores the received images in a database and passes them to the image analysis system.

[0096] Step 4:

[0097] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[0098] Step 5:

[0099] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[0100] Step 6:

[0101] The user checks the feedback sent back from the server on the terminal.

[0102] Step 7:

[0103] The user makes any necessary corrections based on the feedback and sends the image back to the server.

[0104] Certification mark assignment and management

[0105] Step 1:

[0106] The server will manage anti-bias checklists for each company and make them available for download.

[0107] Step 2:

[0108] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[0109] Step 3:

[0110] The terminal transmits the self-evaluation result to the server.

[0111] Step 4:

[0112] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0113] Step 5:

[0114] The server decides whether to grant a certification mark based on the evaluation results.

[0115] Step 6:

[0116] The server decides to issue a certification mark to the company to be certified.

[0117] Step 7:

[0118] The server transmits the certification mark image data and related information to the target company.

[0119] Step 8:

[0120] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[0121] Example 1

[0122] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0123] While advances in AI technology have led to widespread use of generated images, cultural bias in generated images has become a problem. In particular, it is important to share knowledge and technology related to bias countermeasures with the participation of organizations from various industries. There is also a need for certification of businesses that have implemented appropriate bias countermeasures. The purpose of this invention is to provide a means to effectively mitigate cultural bias in AI-generated images and promote fair use.

[0124] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0125] In this invention, the server includes means for hosting a new user registration form, accepting information such as company name, contact person's name, and email address, storing the information in a database, and sending a confirmation email, means for displaying detailed information about workshops, allowing the user to register for the workshop of their choice, and sending reminder emails, means for selecting an AI-generated image for which feedback is desired from a terminal, analyzing elements within the image, generating cultural feedback using an algorithm that evaluates cultural meaning, and returning the analysis results to the user, means for the user to modify the image based on the feedback and re-upload the image, and means for downloading an anti-bias checklist, transmitting self-assessment results to the server, integrating and analyzing the evaluation results with usage history data of the feedback system, and granting a certification mark. This effectively reduces cultural bias in AI-generated images, and enables fair use by granting a certification mark to businesses that have implemented appropriate anti-bias measures.

[0126] A "server" is a computer system that provides services and resources to other computers and devices on a network.

[0127] A "terminal" is an electronic device that a user uses to access a system through an interface and to input or obtain information.

[0128] "User" means any person or entity that registers and accesses the System to use it.

[0129] The "new user registration form" is a web-based input form for entering information such as company name, contact person's name, and email address.

[0130] A "database" is an organized collection of data for efficiently storing, retrieving, and managing information.

[0131] A "confirmation email" is an email sent to notify the user that new registration has been completed.

[0132] A "workshop" is a seminar or training event for sharing and learning knowledge and skills on a specific topic.

[0133] The "Event Calendar" is a calendar-style scheduling feature that lists dates for workshops and other events.

[0134] A "reminder email" is an email that notifies the user in advance of the date and time of event participation and other important matters.

[0135] The "feedback system" is a system for generating and providing cultural feedback to users based on image analysis.

[0136] "AI-generated images" refers to images that are automatically generated using artificial intelligence technology.

[0137] "Cultural meaning" refers to the specific meaning or symbolism that elements such as colors, symbols, and backgrounds have in a particular culture or society.

[0138] "Analysis results" refer to analysis data and evaluation results generated by the image analysis system.

[0139] An "anti-bias checklist" is a list of items that businesses can use to evaluate their anti-bias measures.

[0140] "Self-assessment results" are data on the results of businesses' self-assessment of their own bias countermeasures.

[0141] A "certification mark" is a mark or logo that certifies that fair anti-bias measures are being implemented.

[0142] "Usage History Data" means records of your use of the feedback system and other system features.

[0143] An "algorithm" is a set of steps or formulas for solving a particular problem.

[0144] The present invention relates to a system for effectively reducing cultural bias in AI-generated images and promoting fair use. This system is realized by specific processing steps involving a server, a terminal, and a user.

[0145] To implement this system, the server performs the following functions:

[0146] First, the server hosts a registration form where new users can register. This is built using a web framework such as Django or Ruby on Rails. The user enters the company name, contact name, email address, etc. from the terminal and submits the data. The server receives this data, stores it in a database (e.g., MySQL (registered trademark) or PostgreSQL), and sends a confirmation email.

[0147] The server then displays the details of the workshops and allows users to register for the events they want to attend. This is done using a web portal with an events calendar, where users log in and register. The server stores this information and uses a scheduling tool like crontab or Celery to send reminder emails at set times.

[0148] In the feedback system, users select and upload the AI-generated image they want to receive feedback on from their device. The selected image file is sent to the server and saved in a temporary folder. The server then launches an image analysis system using Tensorflow (registered trademark) or OpenCV to analyze elements in the image and evaluate their cultural meaning. The generated feedback is sent back to the user via the server. For example, it may include feedback such as "red backgrounds may have a specific religious meaning."

[0149] The user can then modify the image based on this feedback and re-upload it, allowing the user to receive new feedback on the altered image.

[0150] The server also provides a bias prevention checklist that users (company personnel) can download and use to conduct a self-assessment. The self-assessment results are sent to the server, which then analyzes them by integrating them with the feedback system's usage history data. Based on the analysis results, businesses that have implemented appropriate bias prevention measures are awarded a certification mark. The certification mark is sent to the company as an image file, which the company can use on its website and marketing materials.

[0151] The following are examples of prompt sentences:

[0152] "Does this image contain any cultural bias?"

[0153] "Please culturally analyze the meaning of the red background."

[0154] Following these steps will help mitigate cultural bias issues in AI-generated images, certify providers who have implemented appropriate anti-bias measures, and promote fair use.

[0155] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0156] Step 1:

[0157] The user enters the required information into the new user registration form and registers.

[0158] Input: Information such as company name, contact person's name, email address, etc.

[0159] Processing: The form data sent from the terminal is received by the server. The server performs initial data validation and stores it in the database if there are no errors.

[0160] Output: The new user information is added to the database and a confirmation email is sent.

[0161] What it does: After the server receives the HTTP request, it validates the form data and stores it in a database using Django ORM and SQL statements. It uses the Python smtplib library to send a confirmation email.

[0162] Step 2:

[0163] Users log in to the consortium web portal, check the event calendar, and register to participate in the workshop.

[0164] Input: User ID and password.

[0165] Processing: The server receives the authentication information and performs authentication. After successful authentication, the event information is retrieved from the database and displayed on the screen. The user selects the workshop they want to participate in, and the server saves that information.

[0166] Output: The user's registration data is saved to the database and a reminder email is scheduled.

[0167] What happens: The server validates the authentication information using a JSON Web Token (JWT). It retrieves the event information from the database using an SQL query and displays a dynamically generated page using an HTML template engine (e.g., Jinja2) to the user. The registration information is stored in the database, and crontab or Celery is configured to send reminder emails.

[0168] Step 3:

[0169] Users select and upload the AI-generated image they want to receive feedback on from their device.

[0170] Input: Image file.

[0171] Processing: The image file is sent from the device to the server, which saves it in a temporary folder and records its path in the database.

[0172] Output: Image files are saved on the server and ready for analysis.

[0173] Specific operation: The user selects an image file using the JavaScript® File API and submits it in a form. The server receives the file using Flask or Django's FileField, and saves it in the OS's temporary directory.

[0174] Step 4:

[0175] The server runs an image analysis system to generate cultural feedback.

[0176] Input: Image file path.

[0177] Processing: The server uses TensorFlow and OpenCV to analyze elements in the image and evaluates cultural meaning using generative AI models.

[0178] Output: Analysis results and feedback.

[0179] How it works: The server runs a Python script, loads a TensorFlow model, and performs image analysis. The resulting data is fed into an AI model, which generates a rating, such as "red backgrounds may have a specific religious meaning."

[0180] Step 5:

[0181] The server returns the analysis results and feedback to the user.

[0182] Input: Analysis results and feedback.

[0183] Processing: The server converts the analysis results into HTML format and returns them to the user via email or a web interface.

[0184] Output: Feedback information is provided to the user.

[0185] Specific operation: The server uses a template engine to format the analysis results into HTML format, and displays them on a dashboard that the user can view after logging in. The same content is also sent via email.

[0186] Step 6:

[0187] The user corrects the image based on the feedback and re-uploads it.

[0188] Input: The modified image file.

[0189] Processing: The corrected image is sent from the device to the server, where it is prepared for reanalysis.

[0190] Output: The corrected image files are saved on the server and can be reanalyzed.

[0191] Specific operation: The user edits the image based on the feedback using image editing software (e.g., Photoshop) and then uploads the image again using step 3.

[0192] Step 7:

[0193] The server provides an anti-bias checklist, and businesses conduct self-assessments.

[0194] Input: User (company representative) checklist download request.

[0195] Processing: The server provides the checklist file and generates a link for the user to download it.

[0196] Output: User downloads the checklist.

[0197] Specific operation: The server retrieves the checklist file from storage, performs access control, and generates a download link. The user clicks the link to download the checklist.

[0198] Step 8:

[0199] The user submits the self-assessment results to the server.

[0200] Input: Self-assessment result data.

[0201] Processing: The device sends the self-evaluation results to the server, which stores the data.

[0202] Output: The self-assessment results are recorded in a database.

[0203] Specific operation: The user evaluates the checklist and sends the results in JSON format to the server using the POST method. The server validates the data and stores it in the database.

[0204] Step 9:

[0205] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0206] Input: Self-assessment result data, feedback usage history data.

[0207] Processing: The server integrates the evaluation data with the historical data and performs data analysis.

[0208] Output: Analysis result data.

[0209] Specific operation: The server uses data analysis libraries such as Pandas and NumPy to comprehensively analyze the evaluation results and usage history and evaluate whether they meet the criteria.

[0210] Step 10:

[0211] The server decides whether to grant a certification mark and transmits the certification mark and related information to the target company.

[0212] Input: Analysis result data.

[0213] Processing: Based on the analysis results, the server determines which companies will be given the certification mark and sends the mark data by email.

[0214] Output: Certification mark image data and related information.

[0215] Specific operation: The server evaluates the analysis results and generates a list of companies that meet the certification criteria. It also sends an image file of the certification mark as an email attachment and provides related information.

[0216] Through these processing steps, it is possible to reduce cultural bias in AI-generated images and award a certification mark to businesses that have implemented appropriate anti-bias measures.

[0217] (Application example 1)

[0218] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0219] There is a need to prevent bias and cultural misunderstandings in generated images and provide fair visual information to users with diverse cultural backgrounds. However, there are limitations to what individual businesses can do independently, and the dissemination of common knowledge on how to counter bias is insufficient. Furthermore, the lack of a system for providing appropriate feedback and certifying images that have been debiased makes it difficult to provide fair images. In particular, there is a growing need for debiasing and fair authentication of images on online shopping sites.

[0220] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0221] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a multi-industry conference, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding a certification mark to businesses that implement appropriate anti-bias measures, a means for evaluating the cultural meaning of images on online shopping sites and generating feedback, a means for users to upload and re-evaluate modified images, and a means for issuing a certification mark based on the evaluation results, thereby enabling the provision and certification of fair, bias-free images.

[0222] The Council is an organization that brings together organizations from various sectors to share knowledge and techniques related to anti-bias measures.

[0223] "Anti-bias" refers to methods and activities to eliminate cultural and social biases contained in generated images and provide fair visual information.

[0224] The "feedback system" is a system that evaluates the cultural meaning of the generated images and provides users with necessary improvements and points of attention.

[0225] A "certification mark" is a seal of approval given to indicate that images and information have been properly anti-biased and culturally fair.

[0226] An "online shopping site" is an online platform where products and services can be purchased over the Internet.

[0227] "Cultural Semantic Evaluation" is the process of analyzing the elements of the generated image and evaluating the cultural and social meanings that these elements possess.

[0228] "Reevaluation" is the process by which a user reevaluates an image that has been modified to ensure an unbiased view.

[0229] "Knowledge and technology sharing" refers to activities in which participating organizations share information and technology related to anti-prejudice measures with each other.

[0230] In order to implement this invention, it is necessary to build a system based on the following contents.

[0231] System configuration

[0232] This system consists of a server, a terminal, and a user. The server operates to share knowledge and techniques related to anti-bias measures through a council and provide cultural feedback on generated images. The terminal is a device through which users upload images and receive feedback. Users can then modify and re-evaluate their images based on the feedback provided by the server.

[0233] Hardware and software used

[0234] Hardware:

[0235] Server: Any hardware specification that supports common servers such as Fedora

[0236] Device: Smart glasses or smartphone

[0237] software:

[0238] Server: Fedora

[0239] Image analysis library: OpenCV

[0240] Generative AI model: Hugging Face transformers library

[0241] Programming language: Python

[0242] Process Overview

[0243] 1. Upload an image

[0244] Users use their devices (smart glasses or smartphones) to take pictures of products and upload them to the server, which receives the images and stores them in a database.

[0245] 2. Generating Cultural Feedback

[0246] The server uses Python scripts to analyze the image. Specifically, it uses OpenCV to analyze elements in the image and then generates cultural feedback using the Hugging Face transformers library. This feedback evaluates the cultural meaning of elements such as the image's colors, symbols, and background, and suggests specific improvements to the user.

[0247] 3. Review and correct feedback

[0248] The user checks the cultural feedback sent from the server on their device, and based on the feedback, the user modifies the image and re-uploads it to the server.

[0249] 4. Issuance of certification mark

[0250] The server then re-evaluates the corrected image to ensure it has been properly treated. It then issues a certification mark and notifies the user. This certification mark can be used on online shopping sites and in marketing materials to demonstrate that the business has properly treated the bias.

[0251] Examples of concrete examples and prompts

[0252] A common example is when an image of a product that uses a lot of red is uploaded. The server provides cultural feedback to the user, prompting them with something like, "This image uses a culturally sensitive color. The use of red may be misleading to certain religions or cultures, so please change it to another color."

[0253] Using this system, users can provide unbiased and fair images and receive proper authentication.

[0254] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0255] Step 1: The user takes a picture of the product using a device (smart glasses or smartphone) and accesses the upload screen. The device sends the captured image data to the server. The input is the image data taken by the user, and the output is the image data received by the server.

[0256] Step 2: The server stores the received image data in a database, preparing it for subsequent processing by the image analysis system. The input is the transmitted image data, and the output is the image data stored in the database.

[0257] Step 3: The server starts the image analysis process. Specifically, it uses Python scripts and OpenCV to analyze elements such as color, symbols, and background in the image. The input is the image data read from the database, and the output is the analyzed data (e.g., color distribution or the detection of specific symbols).

[0258] Step 4: The server generates cultural feedback based on the parsed data using Hugging Face's transformers library. The generative AI model identifies cultural sensitivities and misleading elements and generates specific improvement instructions for the user. The input is the parsed data, and the output is a cultural feedback statement.

[0259] Step 5: The server sends the generated cultural feedback to the user's terminal and notifies the user. The input is the generated feedback sentence, and the output is the notification sent to the user.

[0260] Step 6: The user checks the feedback on the device and makes corrections to the image. After making any necessary corrections, the user re-uploads the image to the server. The input is the feedback statement and the corrected image data, and the output is the corrected image data re-uploaded to the server.

[0261] Step 7: The server re-evaluates the re-uploaded image to ensure proper bias removal. The re-evaluation process also uses OpenCV and the Hugging Face transformers library. The input is the corrected image data, and the output is the re-evaluation result and additional feedback if needed.

[0262] Step 8: Based on the evaluation results, the server issues a certification mark for images that have been properly treated for bias. The issued certification mark is sent to the user. The input is the re-evaluation result and corrected image data, and the output is the certification mark and a notification to the user.

[0263] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0264] To implement this invention, the following specific steps are required: First, a consortium with the participation of various institutions is established to provide a forum for sharing knowledge and technologies related to anti-bias measures. Second, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Furthermore, an emotion engine that recognizes the user's emotions is combined to enhance the effectiveness of the feedback.

[0265] Specific examples of consortium establishment and operation

[0266] Handling user registration

[0267] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[0268] The terminal (company representative) enters the necessary information into the form and completes the registration.

[0269] The server stores the input in a database and sends a confirmation email to the user.

[0270] Users click on the link in the confirmation email to complete registration.

[0271] The server grants access rights to the consortium web portal.

[0272] Holding workshops

[0273] Users log in to the consortium web portal and check the event calendar.

[0274] The server displays detailed information about the workshop (title, date and time, instructor information).

[0275] The user registers for the workshop of their choice.

[0276] The server stores the registration information in a database and sends reminder emails.

[0277] Specific examples of feedback system design and implementation

[0278] Image upload and analysis

[0279] Users select the AI-generated image they want to receive feedback on from their device.

[0280] The terminal sends the selected image file to the server.

[0281] The server stores the received images in a database and passes them to the image analysis system.

[0282] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[0283] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[0284] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[0285] Examples of certification marking and management

[0286] Checklist Evaluation and Certification

[0287] The server will manage anti-bias checklists for each company and make them available for download.

[0288] Users (company representatives) download the checklist and evaluate their company's measures.

[0289] The device sends the self-evaluation results to the server.

[0290] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0291] The server decides whether to grant a certification mark based on the evaluation results.

[0292] The server decides to issue a certification mark to the company to be certified.

[0293] The server transmits the certification mark image data and related information to the target company.

[0294] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[0295] Specific examples of combining emotion engines

[0296] Emotion recognition and feedback regulation

[0297] Users use emotion recognition when receiving feedback.

[0298] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[0299] The server uses the emotion engine to analyze the user's emotion. For example, if the user has a surprised expression, the emotion engine recognizes it as "surprise."

[0300] The server provides emotional data to the feedback system to evaluate the effectiveness of the generated feedback.

[0301] The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[0302] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[0303] The processing flow will be explained below.

[0304] Specific examples of consortium establishment and operation

[0305] Step 1:

[0306] The server hosts a new user registration form and provides an interface for entering information such as company name, contact name, email address, etc.

[0307] Step 2:

[0308] The terminal (company representative) enters the necessary information into the provided form and presses the registration button.

[0309] Step 3:

[0310] The server will store the entered information in a database and send a confirmation email to the appropriate address.

[0311] Step 4:

[0312] The user completes the registration by clicking the link in the confirmation email they receive.

[0313] Step 5:

[0314] The server grants access rights to the consortium web portal to users who have completed registration.

[0315] Step 6:

[0316] The user logs into the web portal using the access rights provided.

[0317] Step 7:

[0318] The server displays detailed information about the workshop (title, date and time, instructor information) on a web portal.

[0319] Step 8:

[0320] The user selects the desired workshop and registers to participate.

[0321] Step 9:

[0322] The server stores the registration information in a database and sends workshop reminder emails to users.

[0323] Specific examples of feedback system design and implementation

[0324] Step 1:

[0325] Users select the AI-generated image they want to receive feedback on from their device.

[0326] Step 2:

[0327] The terminal uploads the selected image file to the server.

[0328] Step 3:

[0329] The server receives the transmitted image file and stores it in a database.

[0330] Step 4:

[0331] The server passes the stored image to the image analysis system and begins analysis.

[0332] Step 5:

[0333] The server generates the analysis results and cultural feedback content from an image analysis system that uses algorithms to analyze elements in an image (such as colors, symbols, and backgrounds) and evaluate their cultural meaning.

[0334] Step 6:

[0335] The server sends the analysis results and feedback to the user's device, such as "red backgrounds may have a specific religious meaning."

[0336] Step 7:

[0337] The user checks the feedback sent back from the server on the terminal.

[0338] Step 8:

[0339] The user makes any necessary corrections based on the feedback and re-uploads the corrected image.

[0340] Examples of certification marking and management

[0341] Step 1:

[0342] The server will manage anti-bias checklists for each company and make them available for download by users.

[0343] Step 2:

[0344] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[0345] Step 3:

[0346] The terminal transmits the self-evaluation result to the server.

[0347] Step 4:

[0348] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0349] Step 5:

[0350] The server decides whether to grant a certification mark based on the integrated analysis results.

[0351] Step 6:

[0352] The server decides to issue a certification mark to the company to be certified.

[0353] Step 7:

[0354] The server transmits the certification mark image data and related information to the target company.

[0355] Step 8:

[0356] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[0357] Specific examples of combining emotion engines

[0358] Step 1:

[0359] The user activates the emotion recognition feature when receiving feedback.

[0360] Step 2:

[0361] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[0362] Step 3:

[0363] The server uses an emotion engine to analyze the user's emotions in real time. For example, if the user shows a surprised expression, it will recognize it as "surprise."

[0364] Step 4:

[0365] The server provides the emotional data to a feedback system to evaluate the effectiveness of the generated feedback.

[0366] Step 5:

[0367] The server adaptively adjusts the feedback content based on changes in the user's emotions. For example, if the user expresses negative emotions, the server adjusts the tone and content of the feedback provided.

[0368] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[0369] Example 2

[0370] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0371] Modern AI-generated image technology runs the risk of producing images that reflect cultural backgrounds and biases. Furthermore, feedback provided without considering user feelings often leads to inappropriate improvements. This poses a challenge, undermining fairness and the user experience surrounding the use of generated images. Furthermore, there is a lack of a system for accurately evaluating and certifying businesses that implement appropriate anti-bias measures.

[0372] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0373] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a consortium of institutions from various industries, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding certification marks to businesses that implement appropriate anti-bias measures, a means for building an emotion recognition system that analyzes user emotions and adjusts the feedback content based on the analysis, and a means for storing and integrating the results of self-evaluation by businesses in a database for evaluation. This makes it possible to provide fair and culturally appropriate feedback and propose improvements that are sensitive to user emotions. It also realizes a system that can accurately evaluate and certify businesses that implement appropriate anti-bias measures.

[0374] A "consortium" is an alliance formed by organizations from various sectors to work together to achieve a common goal.

[0375] "Anti-bias" refers to measures taken to mitigate or eliminate bias or unfairness that may arise from AI systems or algorithms.

[0376] A "feedback system" is a system for providing cultural feedback to generated images, using algorithms that analyze elements within the image and evaluate their cultural meaning.

[0377] An "emotion recognition system" is a system that has the function of analyzing a user's emotions and adjusting the feedback content based on that.

[0378] The "certification mark" is a seal of approval given to businesses that implement appropriate anti-bias measures and indicates fair efforts.

[0379] "Self-evaluation results" refer to the results of an evaluation of the effectiveness and status of anti-bias measures implemented by the business operator itself.

[0380] A "database" is a system that stores structured data and allows it to be efficiently searched and retrieved.

[0381] To implement this invention, the following specific steps are taken: First, a consortium will be established with the participation of institutions from various industries, providing a forum for sharing knowledge and technology related to bias countermeasures. This consortium will share the latest technical information and countermeasure methods, enabling participants to implement effective bias countermeasures.

[0382] Next, we build a feedback system to provide cultural feedback on the generated images, which involves the following steps:

[0383] 1. The user selects the AI-generated image for which they would like to receive feedback from their device.

[0384] 2. The device sends the selected image file to the server.

[0385] 3. The server stores the received images in a database and passes them to the image analysis system.

[0386] 4. The server launches an image analysis system to generate cultural feedback. Specifically, it uses image analysis libraries such as OpenCV and TensorFlow to analyze elements in the image (e.g., colors, symbols, background) and evaluate their cultural meaning.

[0387] 5. The server returns the analysis results and feedback to the user, for example, "The red background may have a specific religious meaning."

[0388] 6. The user checks the feedback sent back from the server on the device and makes any necessary corrections based on the feedback.

[0389] For the granting of the certification mark, the server will manage an anti-bias checklist for each company and make it available for download. Specifically, it will include the following steps:

[0390] 1. The server provides an anti-bias checklist for businesses to download.

[0391] 2. The user (company representative) downloads the checklist and evaluates their company's countermeasures.

[0392] 3. The device sends the evaluation results to the server.

[0393] 4. The server will analyze the self-evaluation results and the feedback system usage history data in order to determine whether to grant the certification mark.

[0394] 5. The server issues a certification mark based on the evaluation results and sends related information to the company.

[0395] 6. The user (certified company representative) downloads the certification mark and uses it on their company's website and marketing materials.

[0396] To combine the emotion recognition system, we will introduce an emotion engine to improve the quality of feedback and user experience. Specifically, we will use emotion analysis software such as Microsoft® Azure® Cognitive Services. The steps are as follows:

[0397] 1. The user chooses to use emotion recognition when receiving feedback.

[0398] 2. The device captures the user's facial expressions and voice and sends the emotional data to the server.

[0399] 3. The server uses an emotion engine to analyze the user's emotions. For example, it can recognize emotions such as "surprise" from the user's facial expressions and voice.

[0400] 4. The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[0401] 5. The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[0402] This system configuration allows us to provide feedback that takes into account the cultural appropriateness of generated images and user emotions, and also enables objective evaluation and certification of businesses that implement anti-bias measures.

[0403] An example of a specific prompt is, "Please analyze the cultural meaning of the image below and provide feedback."

[0404] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0405] Specific examples of consortium establishment and operation

[0406] Handling user registration

[0407] Step 1:

[0408] The user accesses the new user registration form and enters information such as the company name, the name of the person in charge, and an email address.

[0409] Input: Company name, person in charge's name, email address, etc.

[0410] Output: A submission request containing the input information

[0411] Step 2:

[0412] The terminal sends a transmission request to the server.

[0413] Input: A submission request containing input information

[0414] Output: Data sent to the server

[0415] Step 3:

[0416] The server stores the received data in a database and sends the user a confirmation email informing them that their registration is complete.

[0417] Input: Send data

[0418] Data processing: Save input information to a database

[0419] Output: Confirmation email sent

[0420] Step 4:

[0421] The user clicks on the link in the confirmation email to complete the registration.

[0422] Input: Click the link in the confirmation email

[0423] Output: Registration completion notification

[0424] Step 5:

[0425] The server grants the user access to the consortium web portal.

[0426] Data processing: Update user registration status to "Complete"

[0427] Output: Notification of access rights granted

[0428] Holding workshops

[0429] Step 1:

[0430] Users log in to the consortium web portal and check the event calendar.

[0431] Input: User ID and password

[0432] Output: Event calendar view

[0433] Step 2:

[0434] The server displays detailed information about the workshop (title, date and time, instructor information).

[0435] Data processing: Obtaining workshop information from the event database

[0436] Output: Detailed information displayed on a web page

[0437] Step 3:

[0438] The user registers for the workshop of their choice.

[0439] Input: Enter details in the entry form and click the registration button

[0440] Output: Registration request sent

[0441] Step 4:

[0442] The server stores the registration information in a database and sends reminder emails before the workshop begins.

[0443] Data processing: Update participant list, send reminder emails

[0444] Output: Reminder email sent successfully

[0445] Specific examples of feedback system design and implementation

[0446] Image upload and analysis

[0447] Step 1:

[0448] Users select the AI-generated image they want to receive feedback on from their device.

[0449] Input: Select image file

[0450] Output: Selected image file

[0451] Step 2:

[0452] The terminal sends the selected image file to the server.

[0453] Input: Selected image file

[0454] Output: Request to send image file

[0455] Step 3:

[0456] The server stores the received images in a database and passes them to the image analysis system.

[0457] Input: Image file transfer request

[0458] Data processing: Save image data in a database

[0459] Output: Passing image data to an image analysis system

[0460] Step 4:

[0461] The server runs an image analysis system to generate cultural feedback, specifically using OpenCV and TensorFlow to analyze elements in the image and evaluate their cultural meaning.

[0462] Input: Image data

[0463] Data calculation: Analyze elements in the image (color, symbol, background)

[0464] Output: Generated feedback

[0465] Step 5:

[0466] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[0467] Input: Generated feedback

[0468] Data processing: Send feedback to users

[0469] Output: Feedback sent successfully

[0470] Step 6:

[0471] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[0472] Input: Check feedback

[0473] Output: Image correction completed

[0474] Examples of certification marking and management

[0475] Checklist Evaluation and Certification

[0476] Step 1:

[0477] The server will manage anti-bias checklists for each company and make them available for download.

[0478] Input: Checklist PDF file

[0479] Data processing: Upload the checklist to the web portal

[0480] Output: Publish download link

[0481] Step 2:

[0482] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[0483] Input:Download checklist

[0484] Output: Checklist evaluation

[0485] Step 3:

[0486] The terminal transmits the self-evaluation result to the server.

[0487] Input: Self-assessment results

[0488] Output: Send request

[0489] Step 4:

[0490] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0491] Input: Self-assessment results and usage history data

[0492] Data processing: Integrated analysis of evaluation data

[0493] Output: Evaluation results

[0494] Step 5:

[0495] The server decides whether to grant a certification mark based on the evaluation results.

[0496] Input: Evaluation result

[0497] Data processing: Judgment in accordance with certification standards

[0498] Output: Certification mark grant decision

[0499] Step 6:

[0500] The server issues a certification mark to the certified company and transmits related information.

[0501] Input: Certification mark grant decision

[0502] Data processing: generation of certification marks and certificates

[0503] Output: Certification mark sent successfully

[0504] Step 7:

[0505] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[0506] Input: Download certification mark

[0507] Output: Use of certification mark

[0508] Specific examples of combining emotion engines

[0509] Emotion recognition and feedback regulation

[0510] Step 1:

[0511] The user may choose to use emotion recognition functionality when receiving feedback.

[0512] Input: Select emotion recognition feature

[0513] Output: Emotion data collection begins

[0514] Step 2:

[0515] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[0516] Input: User's facial expressions and voice

[0517] Data processing: Real-time capture and data transmission

[0518] Output: Emotion data transmission

[0519] Step 3:

[0520] The server uses an emotion engine to analyze the user's emotions, for example, recognizing emotions such as "surprise" from the user's facial expressions and voice.

[0521] Input: Emotion data

[0522] Data Computing: Sentiment Analysis

[0523] Output: Emotional state

[0524] Step 4:

[0525] The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[0526] Input: Emotional state

[0527] Data processing: Integrated analysis of emotion data and feedback data

[0528] Output: Feedback effect evaluation results

[0529] Step 5:

[0530] The server adaptively adjusts the feedback content based on changes in the user's emotions, for example, softening the tone and content of the feedback if the user expresses negative emotions.

[0531] Input: Feedback effect evaluation results

[0532] Data processing: Adjustment of feedback content

[0533] Output: Adaptively adjusted feedback

[0534] (Application example 2)

[0535] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0536] Advertising materials created by advertising creators often contain bias that ignores cultural background. Such biases can foster misunderstandings and prejudices against specific cultures or ethnicities, hindering the provision of fair advertising. Furthermore, it is difficult to properly capture the emotional responses of users viewing ads and provide feedback, leading to a lack of improvements to maximize the effectiveness of advertising. There is a need to address these issues.

[0537] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0538] In this invention, the server includes means for sharing knowledge and technology related to bias countermeasures through a consortium of institutions from various industries, means for designing and implementing a feedback system that provides cultural feedback on generated images, means for granting a certification mark to businesses that implement appropriate bias countermeasures, means for providing cultural feedback on advertising materials created by advertising creators, and means for analyzing user emotions using an emotion recognition engine and adjusting the feedback content based on the results. This makes it possible to reduce cultural bias in advertising materials and provide appropriate feedback according to user emotions.

[0539] "Multi-industry institutions" is a general term for businesses and organizations in different industries, such as the advertising industry, software development industry, design industry, and marketing industry.

[0540] A "consortium" is a cooperative system formed by multiple companies or institutions working together to achieve a specific purpose or solve a problem.

[0541] "Anti-bias" refers to methods and efforts to eliminate prejudices and preconceptions based on particular cultures or social backgrounds.

[0542] A "feedback system" is a system that provides evaluations and opinions on generated images and advertising materials from a cultural perspective.

[0543] "Elements in an image" are components such as colors, symbols, backgrounds, and objects contained in an image.

[0544] "Cultural feedback" refers to evaluations and opinions about how generated images and advertising materials will be received based on a particular culture or social context.

[0545] A "certification mark" is an official certification symbol given to businesses or products that meet certain standards.

[0546] An "emotion recognition engine" is an algorithm or software that analyzes data such as a user's facial expressions and voice, and classifies and recognizes the user's emotions.

[0547] An "advertising creator" is a professional who creates advertising content and design.

[0548] "User sentiment" refers to the psychological and emotional response a user has when viewing an ad.

[0549] To implement this invention, the following specific system configuration and processes are required: Use the components of the server, terminal, and user, and clarify the roles of each.

[0550] Server Roles

[0551] The server is responsible for the main data processing and analysis functions. High-performance servers are recommended as the hardware to be used. Software such as Python, Flask, Keras, and OpenCV are also required.

[0552] 1. The server receives generated images and advertising materials from the user and stores the data in a database.

[0553] 2. The server passes the stored image data to an analysis algorithm to generate cultural feedback. The algorithm analyzes elements in the image (colors, symbols, background, etc.) and evaluates their cultural meaning.

[0554] 3. The server uses an emotion recognition engine to analyze the user's emotional data (facial expressions and voice) and recognize the user's emotional state. For example, it classifies and recognizes emotions such as "surprise," "happiness," and "sadness."

[0555] Device Role

[0556] A terminal is a device that provides a user interface, such as a smartphone or a head-mounted display.

[0557] 1. The terminal provides an interface for ad creators to upload their created ad materials to the server.

[0558] 2. The device displays the cultural feedback and emotion recognition results returned from the server to the user.

[0559] 3. The device provides tools for modifying advertising materials in real time, helping users improve their ads based on feedback.

[0560] User Roles

[0561] Users are creators who create advertisements and people involved in evaluating advertisements.

[0562] 1. The user uploads advertising material from their device to the server.

[0563] 2. The user checks the cultural feedback and emotion recognition results provided by the server and modifies the advertising material as needed.

[0564] 3. The user decides on the final advertising material based on the feedback and publishes it.

[0565] Specific examples of processing

[0566] For example, when an ad creator uploads a new ad image they've created, they might get feedback like this:

[0567] Cultural feedback: "The image contains a red background, which may have religious significance in parts of Asia. Consider changing the background color or being more considerate of the context."

[0568] Emotion-aware feedback: "We detected a surprised expression from the user. Consider adjusting the exciting elements in your ad."

[0569] This allows ad creators to easily identify specific areas for improvement and create more effective and culturally appropriate ads.

[0570] Through this invention, it is possible to effectively reduce cultural bias in advertising materials and provide appropriate feedback according to the user's emotions, thereby realizing fair and effective advertising creation.

[0571] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0572] Step 1:

[0573] The server receives advertising materials (images or videos) uploaded by users (advertising creators). Specifically, users select advertising materials using the device interface and send them to the server. This input data is stored on the server.

[0574] Step 2:

[0575] The server then passes the stored advertising material to an image analysis algorithm, which generates feedback that includes cultural meaning. Specifically, the algorithm analyzes elements in the image (color, symbols, background, etc.) and evaluates their cultural meaning. The resulting cultural feedback is then output.

[0576] Step 3:

[0577] The server sends emotional data (facial expressions and voice) of users viewing advertisements to the emotion recognition engine. When a user views an advertisement using a device, the device captures the user's facial expressions and voice and sends the data to the server. The emotion recognition engine analyzes this data and recognizes emotional states such as "surprise," "happiness," and "sadness." The analysis results in an emotion label for the user being output.

[0578] Step 4:

[0579] The server integrates the generated cultural feedback with the emotion recognition results and returns feedback to the user, including suggestions for correction and improvement. Specifically, the server combines the cultural feedback and emotion labels and returns them to the device in a format that is easy for the user to understand. For example, specific instructions may be output, such as, "Consider changing the background color because a red background may have a specific religious meaning. Additionally, since the user's surprised expression was detected, please adjust the provocative elements in the advertisement."

[0580] Step 5:

[0581] The user can use their device to check the feedback provided by the server and modify the advertising material as necessary. The user can then edit the advertising material based on the feedback and upload it back to the server. By repeating this process, the optimal advertising material is created.

[0582] Step 6:

[0583] The server will award a certification mark to businesses that have implemented appropriate anti-bias measures. Specifically, the server will integrate and analyze the evaluation results of the businesses' measures and their usage history of the feedback system, and will issue a certification mark if the standards are met. This certification mark will be provided to the certified businesses to display in their advertisements and on their websites.

[0584] Following this trend, advertising creators will be able to eliminate cultural bias and create optimal advertising materials that take user emotions into consideration.

[0585] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0586] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0587] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0588] [Second embodiment]

[0589] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0590] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0591] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0592] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0593] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0594] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0595] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0596] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0597] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0598] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0599] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0600] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0601] To implement this invention, the following specific steps are required: First, a consortium is established with the participation of institutions from various industries. The consortium will serve as a forum for sharing knowledge and technology related to bias prevention. Next, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Finally, a system is built to award a certification mark to businesses that implement appropriate bias prevention measures.

[0602] Specific examples of consortium establishment and operation

[0603] Step 1: User Registration

[0604] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[0605] The terminal (company representative) enters the necessary information into the form and completes the registration.

[0606] The server stores the input in a database and sends a confirmation email.

[0607] Step 2: Host a workshop

[0608] Users log in to the consortium web portal and check the event calendar.

[0609] The server displays detailed information about the workshop (title, date and time, instructor information).

[0610] The user registers for the workshop of their choice.

[0611] The server stores the registration information in a database and sends reminder emails.

[0612] Specific examples of feedback system design and implementation

[0613] Step 1: Upload an image

[0614] Users select the AI-generated image they want to receive feedback on from their device.

[0615] The terminal transmits the selected image file to the server.

[0616] The server stores the received images in a database and passes them to the image analysis system.

[0617] Step 2: Generate feedback

[0618] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[0619] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[0620] Step 3: Review and correct your feedback

[0621] The user checks the feedback sent back from the server on the terminal.

[0622] The user makes any necessary corrections based on the feedback and sends the image to the server again.

[0623] Examples of certification marking and management

[0624] Step 1: Evaluate the checklist

[0625] The server will manage anti-bias checklists for each company and make them available for download.

[0626] Users (company representatives) download the checklist and evaluate their company's measures.

[0627] The terminal transmits the self-evaluation result to the server.

[0628] Step 2: Analyze the results

[0629] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0630] The server decides whether to grant a certification mark based on the evaluation results.

[0631] Step 3: Issuance of the certification mark

[0632] The server decides to issue a certification mark to the company to be certified.

[0633] The server transmits the certification mark image data and related information to the target company.

[0634] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[0635] The above is a concrete example of how to implement the present invention. By using this system, we can effectively mitigate bias issues in AI-generated images in Asia and promote fair use.

[0636] The processing flow will be explained below.

[0637] Establishment and operation of the consortium

[0638] Step 1:

[0639] The user enters information such as the company name, the name of the person in charge, and an email address into the new user registration form.

[0640] Step 2:

[0641] The terminal sends the input information to the server.

[0642] Step 3:

[0643] The server stores the submitted information in a database and sends a confirmation email to the user.

[0644] Step 4:

[0645] The user clicks on the link in the confirmation email to complete the registration.

[0646] Step 5:

[0647] The server grants access rights to the consortium web portal.

[0648] Step 6:

[0649] Users log in to a web portal to view the events calendar.

[0650] Step 7:

[0651] The server displays detailed information about the workshop (title, date and time, instructor information).

[0652] Step 8:

[0653] The user registers for the workshop of their choice.

[0654] Step 9:

[0655] The server stores the registration information in a database and sends reminder emails.

[0656] Design and implementation of a feedback system

[0657] Step 1:

[0658] Users select the AI-generated image they want to receive feedback on from their device.

[0659] Step 2:

[0660] The terminal sends the selected image file to the server.

[0661] Step 3:

[0662] The server stores the received images in a database and passes them to the image analysis system.

[0663] Step 4:

[0664] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[0665] Step 5:

[0666] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[0667] Step 6:

[0668] The user checks the feedback sent back from the server on the terminal.

[0669] Step 7:

[0670] The user makes any necessary corrections based on the feedback and sends the image back to the server.

[0671] Certification mark assignment and management

[0672] Step 1:

[0673] The server will manage anti-bias checklists for each company and make them available for download.

[0674] Step 2:

[0675] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[0676] Step 3:

[0677] The terminal transmits the self-evaluation result to the server.

[0678] Step 4:

[0679] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0680] Step 5:

[0681] The server decides whether to grant a certification mark based on the evaluation results.

[0682] Step 6:

[0683] The server decides to issue a certification mark to the company to be certified.

[0684] Step 7:

[0685] The server transmits the certification mark image data and related information to the target company.

[0686] Step 8:

[0687] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[0688] Example 1

[0689] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0690] While advances in AI technology have led to widespread use of generated images, cultural bias in generated images has become a problem. In particular, it is important to share knowledge and technology related to bias countermeasures with the participation of organizations from various industries. There is also a need for certification of businesses that have implemented appropriate bias countermeasures. The purpose of this invention is to provide a means to effectively mitigate cultural bias in AI-generated images and promote fair use.

[0691] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0692] In this invention, the server includes means for hosting a new user registration form, accepting information such as company name, contact person's name, and email address, storing the information in a database, and sending a confirmation email, means for displaying detailed information about workshops, allowing the user to register for the workshop of their choice, and sending reminder emails, means for selecting an AI-generated image for which feedback is desired from a terminal, analyzing elements within the image, generating cultural feedback using an algorithm that evaluates cultural meaning, and returning the analysis results to the user, means for the user to modify the image based on the feedback and re-upload the image, and means for downloading an anti-bias checklist, transmitting self-assessment results to the server, integrating and analyzing the evaluation results with usage history data of the feedback system, and granting a certification mark. This effectively reduces cultural bias in AI-generated images, and enables fair use by granting a certification mark to businesses that have implemented appropriate anti-bias measures.

[0693] A "server" is a computer system that provides services and resources to other computers and devices on a network.

[0694] A "terminal" is an electronic device that a user uses to access a system through an interface and to input or obtain information.

[0695] "User" means any person or entity that registers and accesses the System to use it.

[0696] The "new user registration form" is a web-based input form for entering information such as company name, contact person's name, and email address.

[0697] A "database" is an organized collection of data for efficiently storing, retrieving, and managing information.

[0698] A "confirmation email" is an email sent to notify the user that new registration has been completed.

[0699] A "workshop" is a seminar or training event for sharing and learning knowledge and skills on a specific topic.

[0700] The "Event Calendar" is a calendar-style scheduling feature that lists dates for workshops and other events.

[0701] A "reminder email" is an email that notifies the user in advance of the date and time of event participation and other important matters.

[0702] The "feedback system" is a system for generating and providing cultural feedback to users based on image analysis.

[0703] "AI-generated images" refers to images that are automatically generated using artificial intelligence technology.

[0704] "Cultural meaning" refers to the specific meaning or symbolism that elements such as colors, symbols, and backgrounds have in a particular culture or society.

[0705] "Analysis results" refer to analysis data and evaluation results generated by the image analysis system.

[0706] An "anti-bias checklist" is a list of items that businesses can use to evaluate their anti-bias measures.

[0707] "Self-assessment results" are data on the results of businesses' self-assessment of their own bias countermeasures.

[0708] A "certification mark" is a mark or logo that certifies that fair anti-bias measures are being implemented.

[0709] "Usage History Data" means records of your use of the feedback system and other system features.

[0710] An "algorithm" is a set of steps or formulas for solving a particular problem.

[0711] The present invention relates to a system for effectively reducing cultural bias in AI-generated images and promoting fair use. This system is realized by specific processing steps involving a server, a terminal, and a user.

[0712] To implement this system, the server performs the following functions:

[0713] First, the server hosts a registration form where new users can register. This is built using a web framework such as Django or Ruby on Rails. The user enters the company name, contact name, email address, etc. from the terminal and submits the data. The server receives this, stores it in a database (e.g., MySQL or PostgreSQL), and sends a confirmation email.

[0714] The server then displays the details of the workshops and allows users to register for the events they want to attend. This is done using a web portal with an events calendar, where users log in and register. The server stores this information and uses a scheduling tool like crontab or Celery to send reminder emails at set times.

[0715] In the feedback system, users select and upload the AI-generated image they want to receive feedback on from their device. The selected image file is sent to the server and saved in a temporary folder. The server then launches an image analysis system using TensorFlow, OpenCV, etc. to analyze elements in the image and evaluate their cultural meaning. The generated feedback is sent back to the user via the server. For example, it may include feedback such as, "Red backgrounds may have a specific religious meaning."

[0716] The user can then modify the image based on this feedback and re-upload it, allowing the user to receive new feedback on the altered image.

[0717] The server also provides a bias prevention checklist that users (company personnel) can download and use to conduct a self-assessment. The self-assessment results are sent to the server, which then analyzes them by integrating them with the feedback system's usage history data. Based on the analysis results, businesses that have implemented appropriate bias prevention measures are awarded a certification mark. The certification mark is sent to the company as an image file, which the company can use on its website and marketing materials.

[0718] The following are examples of prompt sentences:

[0719] "Does this image contain any cultural bias?"

[0720] "Please culturally analyze the meaning of the red background."

[0721] Following these steps will help mitigate cultural bias issues in AI-generated images, certify providers who have implemented appropriate anti-bias measures, and promote fair use.

[0722] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0723] Step 1:

[0724] The user enters the required information into the new user registration form and registers.

[0725] Input: Information such as company name, contact person's name, email address, etc.

[0726] Processing: The form data sent from the terminal is received by the server. The server performs initial data validation and stores it in the database if there are no errors.

[0727] Output: The new user information is added to the database and a confirmation email is sent.

[0728] What it does: After the server receives the HTTP request, it validates the form data and stores it in a database using Django ORM and SQL statements. It uses the Python smtplib library to send a confirmation email.

[0729] Step 2:

[0730] Users log in to the consortium web portal, check the event calendar, and register to participate in the workshop.

[0731] Input: User ID and password.

[0732] Processing: The server receives the authentication information and performs authentication. After successful authentication, the event information is retrieved from the database and displayed on the screen. The user selects the workshop they want to participate in, and the server saves that information.

[0733] Output: The user's registration data is saved to the database and a reminder email is scheduled.

[0734] What happens: The server validates the authentication information using a JSON Web Token (JWT). It retrieves the event information from the database using an SQL query and displays a dynamically generated page using an HTML template engine (e.g., Jinja2) to the user. The registration information is stored in the database, and crontab or Celery is configured to send reminder emails.

[0735] Step 3:

[0736] Users select and upload the AI-generated image they want to receive feedback on from their device.

[0737] Input: Image file.

[0738] Processing: The image file is sent from the device to the server, which saves it in a temporary folder and records its path in the database.

[0739] Output: Image files are saved on the server and ready for analysis.

[0740] Specific operation: The user selects an image file using JavaScript's File API and submits it in a form. The server receives the file using Flask or Django's FileField and saves it in the OS's temporary directory.

[0741] Step 4:

[0742] The server runs an image analysis system to generate cultural feedback.

[0743] Input: Image file path.

[0744] Processing: The server uses TensorFlow and OpenCV to analyze elements in the image and evaluates cultural meaning using generative AI models.

[0745] Output: Analysis results and feedback.

[0746] How it works: The server runs a Python script, loads a TensorFlow model, and performs image analysis. The resulting data is fed into an AI model, which generates a rating, such as "red backgrounds may have a specific religious meaning."

[0747] Step 5:

[0748] The server returns the analysis results and feedback to the user.

[0749] Input: Analysis results and feedback.

[0750] Processing: The server converts the analysis results into HTML format and returns them to the user via email or a web interface.

[0751] Output: Feedback information is provided to the user.

[0752] Specific operation: The server uses a template engine to format the analysis results into HTML format, and displays them on a dashboard that the user can view after logging in. The same content is also sent via email.

[0753] Step 6:

[0754] The user corrects the image based on the feedback and re-uploads it.

[0755] Input: The modified image file.

[0756] Processing: The corrected image is sent from the device to the server, where it is prepared for reanalysis.

[0757] Output: The corrected image files are saved on the server and can be reanalyzed.

[0758] Specific operation: The user edits the image based on the feedback using image editing software (e.g., Photoshop) and then uploads the image again using step 3.

[0759] Step 7:

[0760] The server provides an anti-bias checklist, and businesses conduct self-assessments.

[0761] Input: User (company representative) checklist download request.

[0762] Processing: The server provides the checklist file and generates a link for the user to download it.

[0763] Output: User downloads the checklist.

[0764] Specific operation: The server retrieves the checklist file from storage, performs access control, and generates a download link. The user clicks the link to download the checklist.

[0765] Step 8:

[0766] The user submits the self-assessment results to the server.

[0767] Input: Self-assessment result data.

[0768] Processing: The device sends the self-evaluation results to the server, which stores the data.

[0769] Output: The self-assessment results are recorded in a database.

[0770] Specific operation: The user evaluates the checklist and sends the results in JSON format to the server using the POST method. The server validates the data and stores it in the database.

[0771] Step 9:

[0772] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0773] Input: Self-assessment result data, feedback usage history data.

[0774] Processing: The server integrates the evaluation data with the historical data and performs data analysis.

[0775] Output: Analysis result data.

[0776] Specific operation: The server uses data analysis libraries such as Pandas and NumPy to comprehensively analyze the evaluation results and usage history and evaluate whether they meet the criteria.

[0777] Step 10:

[0778] The server decides whether to grant a certification mark and transmits the certification mark and related information to the target company.

[0779] Input: Analysis result data.

[0780] Processing: Based on the analysis results, the server determines which companies will be given the certification mark and sends the mark data by email.

[0781] Output: Certification mark image data and related information.

[0782] Specific operation: The server evaluates the analysis results and generates a list of companies that meet the certification criteria. It also sends an image file of the certification mark as an email attachment and provides related information.

[0783] Through these processing steps, it is possible to reduce cultural bias in AI-generated images and award a certification mark to businesses that have implemented appropriate anti-bias measures.

[0784] (Application example 1)

[0785] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0786] There is a need to prevent bias and cultural misunderstandings in generated images and provide fair visual information to users with diverse cultural backgrounds. However, there are limitations to what individual businesses can do independently, and the dissemination of common knowledge on how to counter bias is insufficient. Furthermore, the lack of a system for providing appropriate feedback and certifying images that have been debiased makes it difficult to provide fair images. In particular, there is a growing need for debiasing and fair authentication of images on online shopping sites.

[0787] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0788] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a multi-industry conference, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding a certification mark to businesses that implement appropriate anti-bias measures, a means for evaluating the cultural meaning of images on online shopping sites and generating feedback, a means for users to upload and re-evaluate modified images, and a means for issuing a certification mark based on the evaluation results, thereby enabling the provision and certification of fair, bias-free images.

[0789] The Council is an organization that brings together organizations from various sectors to share knowledge and techniques related to anti-bias measures.

[0790] "Anti-bias" refers to methods and activities to eliminate cultural and social biases contained in generated images and provide fair visual information.

[0791] The "feedback system" is a system that evaluates the cultural meaning of the generated images and provides users with necessary improvements and points of attention.

[0792] A "certification mark" is a seal of approval given to indicate that images and information have been properly anti-biased and culturally fair.

[0793] An "online shopping site" is an online platform where products and services can be purchased over the Internet.

[0794] "Cultural Semantic Evaluation" is the process of analyzing the elements of the generated image and evaluating the cultural and social meanings that these elements possess.

[0795] "Reevaluation" is the process by which a user reevaluates an image that has been modified to ensure an unbiased view.

[0796] "Knowledge and technology sharing" refers to activities in which participating organizations share information and technology related to anti-prejudice measures with each other.

[0797] In order to implement this invention, it is necessary to build a system based on the following contents.

[0798] System configuration

[0799] This system consists of a server, a terminal, and a user. The server operates to share knowledge and techniques related to anti-bias measures through a council and provide cultural feedback on generated images. The terminal is a device through which users upload images and receive feedback. Users can then modify and re-evaluate their images based on the feedback provided by the server.

[0800] Hardware and software used

[0801] Hardware:

[0802] Server: Any hardware specification that supports common servers such as Fedora

[0803] Device: Smart glasses or smartphone

[0804] software:

[0805] Server: Fedora

[0806] Image analysis library: OpenCV

[0807] Generative AI model: Hugging Face transformers library

[0808] Programming language: Python

[0809] Process Overview

[0810] 1. Upload an image

[0811] Users use their devices (smart glasses or smartphones) to take pictures of products and upload them to the server, which receives the images and stores them in a database.

[0812] 2. Generating Cultural Feedback

[0813] The server uses Python scripts to analyze the image. Specifically, it uses OpenCV to analyze elements in the image and then generates cultural feedback using the Hugging Face transformers library. This feedback evaluates the cultural meaning of elements such as the image's colors, symbols, and background, and suggests specific improvements to the user.

[0814] 3. Review and correct feedback

[0815] The user checks the cultural feedback sent from the server on their device, and based on the feedback, the user modifies the image and re-uploads it to the server.

[0816] 4. Issuance of certification mark

[0817] The server then re-evaluates the corrected image to ensure it has been properly treated. It then issues a certification mark and notifies the user. This certification mark can be used on online shopping sites and in marketing materials to demonstrate that the business has properly treated the bias.

[0818] Examples of concrete examples and prompts

[0819] A common example is when an image of a product that uses a lot of red is uploaded. The server provides cultural feedback to the user, prompting them with something like, "This image uses a culturally sensitive color. The use of red may be misleading to certain religions or cultures, so please change it to another color."

[0820] Using this system, users can provide unbiased and fair images and receive proper authentication.

[0821] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0822] Step 1: The user takes a picture of the product using a device (smart glasses or smartphone) and accesses the upload screen. The device sends the captured image data to the server. The input is the image data taken by the user, and the output is the image data received by the server.

[0823] Step 2: The server stores the received image data in a database, preparing it for subsequent processing by the image analysis system. The input is the transmitted image data, and the output is the image data stored in the database.

[0824] Step 3: The server starts the image analysis process. Specifically, it uses Python scripts and OpenCV to analyze elements such as color, symbols, and background in the image. The input is the image data read from the database, and the output is the analyzed data (e.g., color distribution or the detection of specific symbols).

[0825] Step 4: The server generates cultural feedback based on the parsed data using Hugging Face's transformers library. The generative AI model identifies cultural sensitivities and misleading elements and generates specific improvement instructions for the user. The input is the parsed data, and the output is a cultural feedback statement.

[0826] Step 5: The server sends the generated cultural feedback to the user's terminal and notifies the user. The input is the generated feedback sentence, and the output is the notification sent to the user.

[0827] Step 6: The user checks the feedback on the device and makes corrections to the image. After making any necessary corrections, the user re-uploads the image to the server. The input is the feedback statement and the corrected image data, and the output is the corrected image data re-uploaded to the server.

[0828] Step 7: The server re-evaluates the re-uploaded image to ensure proper bias removal. The re-evaluation process also uses OpenCV and the Hugging Face transformers library. The input is the corrected image data, and the output is the re-evaluation result and additional feedback if needed.

[0829] Step 8: Based on the evaluation results, the server issues a certification mark for images that have been properly treated for bias. The issued certification mark is sent to the user. The input is the re-evaluation result and corrected image data, and the output is the certification mark and a notification to the user.

[0830] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0831] To implement this invention, the following specific steps are required: First, a consortium with the participation of various institutions is established to provide a forum for sharing knowledge and technologies related to anti-bias measures. Second, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Furthermore, an emotion engine that recognizes the user's emotions is combined to enhance the effectiveness of the feedback.

[0832] Specific examples of consortium establishment and operation

[0833] Handling user registration

[0834] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[0835] The terminal (company representative) enters the necessary information into the form and completes the registration.

[0836] The server stores the input in a database and sends a confirmation email to the user.

[0837] Users click on the link in the confirmation email to complete registration.

[0838] The server grants access rights to the consortium web portal.

[0839] Holding workshops

[0840] Users log in to the consortium web portal and check the event calendar.

[0841] The server displays detailed information about the workshop (title, date and time, instructor information).

[0842] The user registers for the workshop of their choice.

[0843] The server stores the registration information in a database and sends reminder emails.

[0844] Specific examples of feedback system design and implementation

[0845] Image upload and analysis

[0846] Users select the AI-generated image they want to receive feedback on from their device.

[0847] The terminal sends the selected image file to the server.

[0848] The server stores the received images in a database and passes them to the image analysis system.

[0849] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[0850] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[0851] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[0852] Examples of certification marking and management

[0853] Checklist Evaluation and Certification

[0854] The server will manage anti-bias checklists for each company and make them available for download.

[0855] Users (company representatives) download the checklist and evaluate their company's measures.

[0856] The device sends the self-evaluation results to the server.

[0857] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0858] The server decides whether to grant a certification mark based on the evaluation results.

[0859] The server decides to issue a certification mark to the company to be certified.

[0860] The server transmits the certification mark image data and related information to the target company.

[0861] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[0862] Specific examples of combining emotion engines

[0863] Emotion recognition and feedback regulation

[0864] Users use emotion recognition when receiving feedback.

[0865] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[0866] The server uses the emotion engine to analyze the user's emotion. For example, if the user has a surprised expression, the emotion engine recognizes it as "surprise."

[0867] The server provides emotional data to the feedback system to evaluate the effectiveness of the generated feedback.

[0868] The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[0869] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[0870] The processing flow will be explained below.

[0871] Specific examples of consortium establishment and operation

[0872] Step 1:

[0873] The server hosts a new user registration form and provides an interface for entering information such as company name, contact name, email address, etc.

[0874] Step 2:

[0875] The terminal (company representative) enters the necessary information into the provided form and presses the registration button.

[0876] Step 3:

[0877] The server will store the entered information in a database and send a confirmation email to the appropriate address.

[0878] Step 4:

[0879] The user completes the registration by clicking the link in the confirmation email they receive.

[0880] Step 5:

[0881] The server grants access rights to the consortium web portal to users who have completed registration.

[0882] Step 6:

[0883] The user logs into the web portal using the access rights provided.

[0884] Step 7:

[0885] The server displays detailed information about the workshop (title, date and time, instructor information) on a web portal.

[0886] Step 8:

[0887] The user selects the desired workshop and registers to participate.

[0888] Step 9:

[0889] The server stores the registration information in a database and sends workshop reminder emails to users.

[0890] Specific examples of feedback system design and implementation

[0891] Step 1:

[0892] Users select the AI-generated image they want to receive feedback on from their device.

[0893] Step 2:

[0894] The terminal uploads the selected image file to the server.

[0895] Step 3:

[0896] The server receives the transmitted image file and stores it in a database.

[0897] Step 4:

[0898] The server passes the stored image to the image analysis system and begins analysis.

[0899] Step 5:

[0900] The server generates the analysis results and cultural feedback content from an image analysis system that uses algorithms to analyze elements in an image (such as colors, symbols, and backgrounds) and evaluate their cultural meaning.

[0901] Step 6:

[0902] The server sends the analysis results and feedback to the user's device, such as "red backgrounds may have a specific religious meaning."

[0903] Step 7:

[0904] The user checks the feedback sent back from the server on the terminal.

[0905] Step 8:

[0906] The user makes any necessary corrections based on the feedback and re-uploads the corrected image.

[0907] Examples of certification marking and management

[0908] Step 1:

[0909] The server will manage anti-bias checklists for each company and make them available for download by users.

[0910] Step 2:

[0911] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[0912] Step 3:

[0913] The terminal transmits the self-evaluation result to the server.

[0914] Step 4:

[0915] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[0916] Step 5:

[0917] The server decides whether to grant a certification mark based on the integrated analysis results.

[0918] Step 6:

[0919] The server decides to issue a certification mark to the company to be certified.

[0920] Step 7:

[0921] The server transmits the certification mark image data and related information to the target company.

[0922] Step 8:

[0923] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[0924] Specific examples of combining emotion engines

[0925] Step 1:

[0926] The user activates the emotion recognition feature when receiving feedback.

[0927] Step 2:

[0928] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[0929] Step 3:

[0930] The server uses an emotion engine to analyze the user's emotions in real time. For example, if the user shows a surprised expression, it will recognize it as "surprise."

[0931] Step 4:

[0932] The server provides the emotional data to a feedback system to evaluate the effectiveness of the generated feedback.

[0933] Step 5:

[0934] The server adaptively adjusts the feedback content based on changes in the user's emotions. For example, if the user expresses negative emotions, the server adjusts the tone and content of the feedback provided.

[0935] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[0936] Example 2

[0937] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0938] Modern AI-generated image technology runs the risk of producing images that reflect cultural backgrounds and biases. Furthermore, feedback provided without considering user feelings often leads to inappropriate improvements. This poses a challenge, undermining fairness and the user experience surrounding the use of generated images. Furthermore, there is a lack of a system for accurately evaluating and certifying businesses that implement appropriate anti-bias measures.

[0939] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0940] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a consortium of institutions from various industries, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding certification marks to businesses that implement appropriate anti-bias measures, a means for building an emotion recognition system that analyzes user emotions and adjusts the feedback content based on the analysis, and a means for storing and integrating the results of self-evaluation by businesses in a database for evaluation. This makes it possible to provide fair and culturally appropriate feedback and propose improvements that are sensitive to user emotions. It also realizes a system that can accurately evaluate and certify businesses that implement appropriate anti-bias measures.

[0941] A "consortium" is an alliance formed by organizations from various sectors to work together to achieve a common goal.

[0942] "Anti-bias" refers to measures taken to mitigate or eliminate bias or unfairness that may arise from AI systems or algorithms.

[0943] A "feedback system" is a system for providing cultural feedback to generated images, using algorithms that analyze elements within the image and evaluate their cultural meaning.

[0944] An "emotion recognition system" is a system that has the function of analyzing a user's emotions and adjusting the feedback content based on that.

[0945] The "certification mark" is a seal of approval given to businesses that implement appropriate anti-bias measures and indicates fair efforts.

[0946] "Self-evaluation results" refer to the results of an evaluation of the effectiveness and status of anti-bias measures implemented by the business operator itself.

[0947] A "database" is a system that stores structured data and allows it to be efficiently searched and retrieved.

[0948] To implement this invention, the following specific steps are taken: First, a consortium will be established with the participation of institutions from various industries, providing a forum for sharing knowledge and technology related to bias countermeasures. This consortium will share the latest technical information and countermeasure methods, enabling participants to implement effective bias countermeasures.

[0949] Next, we build a feedback system to provide cultural feedback on the generated images, which involves the following steps:

[0950] 1. The user selects the AI-generated image for which they would like to receive feedback from their device.

[0951] 2. The device sends the selected image file to the server.

[0952] 3. The server stores the received images in a database and passes them to the image analysis system.

[0953] 4. The server launches an image analysis system to generate cultural feedback. Specifically, it uses image analysis libraries such as OpenCV and TensorFlow to analyze elements in the image (e.g., colors, symbols, background) and evaluate their cultural meaning.

[0954] 5. The server returns the analysis results and feedback to the user, for example, "The red background may have a specific religious meaning."

[0955] 6. The user checks the feedback sent back from the server on the device and makes any necessary corrections based on the feedback.

[0956] For the granting of the certification mark, the server will manage an anti-bias checklist for each company and make it available for download. Specifically, it will include the following steps:

[0957] 1. The server provides an anti-bias checklist for businesses to download.

[0958] 2. The user (company representative) downloads the checklist and evaluates their company's countermeasures.

[0959] 3. The device sends the evaluation results to the server.

[0960] 4. The server will analyze the self-evaluation results and the feedback system usage history data in order to determine whether to grant the certification mark.

[0961] 5. The server issues a certification mark based on the evaluation results and sends related information to the company.

[0962] 6. The user (certified company representative) downloads the certification mark and uses it on their company's website and marketing materials.

[0963] For the combination of emotion recognition systems, we will introduce an emotion engine to improve the quality of feedback and user experience. Specifically, we will use emotion analysis software such as Microsoft Azure Cognitive Services. The steps are as follows:

[0964] 1. The user chooses to use emotion recognition when receiving feedback.

[0965] 2. The device captures the user's facial expressions and voice and sends the emotional data to the server.

[0966] 3. The server uses an emotion engine to analyze the user's emotions. For example, it can recognize emotions such as "surprise" from the user's facial expressions and voice.

[0967] 4. The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[0968] 5. The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[0969] This system configuration allows us to provide feedback that takes into account the cultural appropriateness of generated images and user emotions, and also enables objective evaluation and certification of businesses that implement anti-bias measures.

[0970] An example of a specific prompt is, "Please analyze the cultural meaning of the image below and provide feedback."

[0971] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0972] Specific examples of consortium establishment and operation

[0973] Handling user registration

[0974] Step 1:

[0975] The user accesses the new user registration form and enters information such as the company name, the name of the person in charge, and an email address.

[0976] Input: Company name, person in charge's name, email address, etc.

[0977] Output: A submission request containing the input information

[0978] Step 2:

[0979] The terminal sends a transmission request to the server.

[0980] Input: A submission request containing input information

[0981] Output: Data sent to the server

[0982] Step 3:

[0983] The server stores the received data in a database and sends the user a confirmation email informing them that their registration is complete.

[0984] Input: Send data

[0985] Data processing: Save input information to a database

[0986] Output: Confirmation email sent

[0987] Step 4:

[0988] The user clicks on the link in the confirmation email to complete the registration.

[0989] Input: Click the link in the confirmation email

[0990] Output: Registration completion notification

[0991] Step 5:

[0992] The server grants the user access to the consortium web portal.

[0993] Data processing: Update user registration status to "Complete"

[0994] Output: Notification of access rights granted

[0995] Holding workshops

[0996] Step 1:

[0997] Users log in to the consortium web portal and check the event calendar.

[0998] Input: User ID and password

[0999] Output: Event calendar view

[1000] Step 2:

[1001] The server displays detailed information about the workshop (title, date and time, instructor information).

[1002] Data processing: Obtaining workshop information from the event database

[1003] Output: Detailed information displayed on a web page

[1004] Step 3:

[1005] The user registers for the workshop of their choice.

[1006] Input: Enter details in the entry form and click the registration button

[1007] Output: Registration request sent

[1008] Step 4:

[1009] The server stores the registration information in a database and sends reminder emails before the workshop begins.

[1010] Data processing: Update participant list, send reminder emails

[1011] Output: Reminder email sent successfully

[1012] Specific examples of feedback system design and implementation

[1013] Image upload and analysis

[1014] Step 1:

[1015] Users select the AI-generated image they want to receive feedback on from their device.

[1016] Input: Select image file

[1017] Output: Selected image file

[1018] Step 2:

[1019] The terminal sends the selected image file to the server.

[1020] Input: Selected image file

[1021] Output: Request to send image file

[1022] Step 3:

[1023] The server stores the received images in a database and passes them to the image analysis system.

[1024] Input: Image file transfer request

[1025] Data processing: Save image data in a database

[1026] Output: Passing image data to an image analysis system

[1027] Step 4:

[1028] The server runs an image analysis system to generate cultural feedback, specifically using OpenCV and TensorFlow to analyze elements in the image and evaluate their cultural meaning.

[1029] Input: Image data

[1030] Data calculation: Analyze elements in the image (color, symbol, background)

[1031] Output: Generated feedback

[1032] Step 5:

[1033] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[1034] Input: Generated feedback

[1035] Data processing: Send feedback to users

[1036] Output: Feedback sent successfully

[1037] Step 6:

[1038] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[1039] Input: Check feedback

[1040] Output: Image correction completed

[1041] Examples of certification marking and management

[1042] Checklist Evaluation and Certification

[1043] Step 1:

[1044] The server will manage anti-bias checklists for each company and make them available for download.

[1045] Input: Checklist PDF file

[1046] Data processing: Upload the checklist to the web portal

[1047] Output: Publish download link

[1048] Step 2:

[1049] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[1050] Input:Download checklist

[1051] Output: Checklist evaluation

[1052] Step 3:

[1053] The terminal transmits the self-evaluation result to the server.

[1054] Input: Self-assessment results

[1055] Output: Send request

[1056] Step 4:

[1057] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1058] Input: Self-assessment results and usage history data

[1059] Data processing: Integrated analysis of evaluation data

[1060] Output: Evaluation results

[1061] Step 5:

[1062] The server decides whether to grant a certification mark based on the evaluation results.

[1063] Input: Evaluation result

[1064] Data processing: Judgment in accordance with certification standards

[1065] Output: Certification mark grant decision

[1066] Step 6:

[1067] The server issues a certification mark to the certified company and transmits related information.

[1068] Input: Certification mark grant decision

[1069] Data processing: generation of certification marks and certificates

[1070] Output: Certification mark sent successfully

[1071] Step 7:

[1072] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[1073] Input: Download certification mark

[1074] Output: Use of certification mark

[1075] Specific examples of combining emotion engines

[1076] Emotion recognition and feedback regulation

[1077] Step 1:

[1078] The user may choose to use emotion recognition functionality when receiving feedback.

[1079] Input: Select emotion recognition feature

[1080] Output: Emotion data collection begins

[1081] Step 2:

[1082] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[1083] Input: User's facial expressions and voice

[1084] Data processing: Real-time capture and data transmission

[1085] Output: Emotion data transmission

[1086] Step 3:

[1087] The server uses an emotion engine to analyze the user's emotions, for example, recognizing emotions such as "surprise" from the user's facial expressions and voice.

[1088] Input: Emotion data

[1089] Data Computing: Sentiment Analysis

[1090] Output: Emotional state

[1091] Step 4:

[1092] The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[1093] Input: Emotional state

[1094] Data processing: Integrated analysis of emotion data and feedback data

[1095] Output: Feedback effect evaluation results

[1096] Step 5:

[1097] The server adaptively adjusts the feedback content based on changes in the user's emotions, for example, softening the tone and content of the feedback if the user expresses negative emotions.

[1098] Input: Feedback effect evaluation results

[1099] Data processing: Adjustment of feedback content

[1100] Output: Adaptively adjusted feedback

[1101] (Application example 2)

[1102] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1103] Advertising materials created by advertising creators often contain bias that ignores cultural background. Such biases can foster misunderstandings and prejudices against specific cultures or ethnicities, hindering the provision of fair advertising. Furthermore, it is difficult to properly capture the emotional responses of users viewing ads and provide feedback, leading to a lack of improvements to maximize the effectiveness of advertising. There is a need to address these issues.

[1104] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1105] In this invention, the server includes means for sharing knowledge and technology related to bias countermeasures through a consortium of institutions from various industries, means for designing and implementing a feedback system that provides cultural feedback on generated images, means for granting a certification mark to businesses that implement appropriate bias countermeasures, means for providing cultural feedback on advertising materials created by advertising creators, and means for analyzing user emotions using an emotion recognition engine and adjusting the feedback content based on the results. This makes it possible to reduce cultural bias in advertising materials and provide appropriate feedback according to user emotions.

[1106] "Multi-industry institutions" is a general term for businesses and organizations in different industries, such as the advertising industry, software development industry, design industry, and marketing industry.

[1107] A "consortium" is a cooperative system formed by multiple companies or institutions working together to achieve a specific purpose or solve a problem.

[1108] "Anti-bias" refers to methods and efforts to eliminate prejudices and preconceptions based on particular cultures or social backgrounds.

[1109] A "feedback system" is a system that provides evaluations and opinions on generated images and advertising materials from a cultural perspective.

[1110] "Elements in an image" are components such as colors, symbols, backgrounds, and objects contained in an image.

[1111] "Cultural feedback" refers to evaluations and opinions about how generated images and advertising materials will be received based on a particular culture or social context.

[1112] A "certification mark" is an official certification symbol given to businesses or products that meet certain standards.

[1113] An "emotion recognition engine" is an algorithm or software that analyzes data such as a user's facial expressions and voice, and classifies and recognizes the user's emotions.

[1114] An "advertising creator" is a professional who creates advertising content and design.

[1115] "User sentiment" refers to the psychological and emotional response a user has when viewing an ad.

[1116] To implement this invention, the following specific system configuration and processes are required: Use the components of the server, terminal, and user, and clarify the roles of each.

[1117] Server Roles

[1118] The server is responsible for the main data processing and analysis functions. High-performance servers are recommended as the hardware to be used. Software such as Python, Flask, Keras, and OpenCV are also required.

[1119] 1. The server receives generated images and advertising materials from the user and stores the data in a database.

[1120] 2. The server passes the stored image data to an analysis algorithm to generate cultural feedback. The algorithm analyzes elements in the image (colors, symbols, background, etc.) and evaluates their cultural meaning.

[1121] 3. The server uses an emotion recognition engine to analyze the user's emotional data (facial expressions and voice) and recognize the user's emotional state. For example, it classifies and recognizes emotions such as "surprise," "happiness," and "sadness."

[1122] Device Role

[1123] A terminal is a device that provides a user interface, such as a smartphone or a head-mounted display.

[1124] 1. The terminal provides an interface for ad creators to upload their created ad materials to the server.

[1125] 2. The device displays the cultural feedback and emotion recognition results returned from the server to the user.

[1126] 3. The device provides tools for modifying advertising materials in real time, helping users improve their ads based on feedback.

[1127] User Roles

[1128] Users are creators who create advertisements and people involved in evaluating advertisements.

[1129] 1. The user uploads advertising material from their device to the server.

[1130] 2. The user checks the cultural feedback and emotion recognition results provided by the server and modifies the advertising material as needed.

[1131] 3. The user decides on the final advertising material based on the feedback and publishes it.

[1132] Specific examples of processing

[1133] For example, when an ad creator uploads a new ad image they've created, they might get feedback like this:

[1134] Cultural feedback: "The image contains a red background, which may have religious significance in parts of Asia. Consider changing the background color or being more considerate of the context."

[1135] Emotion-aware feedback: "We detected a surprised expression from the user. Consider adjusting the exciting elements in your ad."

[1136] This allows ad creators to easily identify specific areas for improvement and create more effective and culturally appropriate ads.

[1137] Through this invention, it is possible to effectively reduce cultural bias in advertising materials and provide appropriate feedback according to the user's emotions, thereby realizing fair and effective advertising creation.

[1138] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1139] Step 1:

[1140] The server receives advertising materials (images or videos) uploaded by users (advertising creators). Specifically, users select advertising materials using the device interface and send them to the server. This input data is stored on the server.

[1141] Step 2:

[1142] The server then passes the stored advertising material to an image analysis algorithm, which generates feedback that includes cultural meaning. Specifically, the algorithm analyzes elements in the image (color, symbols, background, etc.) and evaluates their cultural meaning. The resulting cultural feedback is then output.

[1143] Step 3:

[1144] The server sends emotional data (facial expressions and voice) of users viewing advertisements to the emotion recognition engine. When a user views an advertisement using a device, the device captures the user's facial expressions and voice and sends the data to the server. The emotion recognition engine analyzes this data and recognizes emotional states such as "surprise," "happiness," and "sadness." The analysis results in an emotion label for the user being output.

[1145] Step 4:

[1146] The server integrates the generated cultural feedback with the emotion recognition results and returns feedback to the user, including suggestions for correction and improvement. Specifically, the server combines the cultural feedback and emotion labels and returns them to the device in a format that is easy for the user to understand. For example, specific instructions may be output, such as, "Consider changing the background color because a red background may have a specific religious meaning. Additionally, since the user's surprised expression was detected, please adjust the provocative elements in the advertisement."

[1147] Step 5:

[1148] The user can use their device to check the feedback provided by the server and modify the advertising material as necessary. The user can then edit the advertising material based on the feedback and upload it back to the server. By repeating this process, the optimal advertising material is created.

[1149] Step 6:

[1150] The server will award a certification mark to businesses that have implemented appropriate anti-bias measures. Specifically, the server will integrate and analyze the evaluation results of the businesses' measures and their usage history of the feedback system, and will issue a certification mark if the standards are met. This certification mark will be provided to the certified businesses to display in their advertisements and on their websites.

[1151] Following this trend, advertising creators will be able to eliminate cultural bias and create optimal advertising materials that take user emotions into consideration.

[1152] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1153] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1154] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1155] [Third embodiment]

[1156] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1157] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1158] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1159] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1160] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1161] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1162] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1163] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1164] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1165] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1166] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1167] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1168] To implement this invention, the following specific steps are required: First, a consortium is established with the participation of institutions from various industries. The consortium will serve as a forum for sharing knowledge and technology related to bias prevention. Next, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Finally, a system is built to award a certification mark to businesses that implement appropriate bias prevention measures.

[1169] Specific examples of consortium establishment and operation

[1170] Step 1: User Registration

[1171] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[1172] The terminal (company representative) enters the necessary information into the form and completes the registration.

[1173] The server stores the input in a database and sends a confirmation email.

[1174] Step 2: Host a workshop

[1175] Users log in to the consortium web portal and check the event calendar.

[1176] The server displays detailed information about the workshop (title, date and time, instructor information).

[1177] The user registers for the workshop of their choice.

[1178] The server stores the registration information in a database and sends reminder emails.

[1179] Specific examples of feedback system design and implementation

[1180] Step 1: Upload an image

[1181] Users select the AI-generated image they want to receive feedback on from their device.

[1182] The terminal transmits the selected image file to the server.

[1183] The server stores the received images in a database and passes them to the image analysis system.

[1184] Step 2: Generate feedback

[1185] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[1186] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[1187] Step 3: Review and correct your feedback

[1188] The user checks the feedback sent back from the server on the terminal.

[1189] The user makes any necessary corrections based on the feedback and sends the image to the server again.

[1190] Examples of certification marking and management

[1191] Step 1: Evaluate the checklist

[1192] The server will manage anti-bias checklists for each company and make them available for download.

[1193] Users (company representatives) download the checklist and evaluate their company's measures.

[1194] The terminal transmits the self-evaluation result to the server.

[1195] Step 2: Analyze the results

[1196] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1197] The server decides whether to grant a certification mark based on the evaluation results.

[1198] Step 3: Issuance of the certification mark

[1199] The server decides to issue a certification mark to the company to be certified.

[1200] The server transmits the certification mark image data and related information to the target company.

[1201] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[1202] The above is a concrete example of how to implement the present invention. By using this system, we can effectively mitigate bias issues in AI-generated images in Asia and promote fair use.

[1203] The processing flow will be explained below.

[1204] Establishment and operation of the consortium

[1205] Step 1:

[1206] The user enters information such as the company name, the name of the person in charge, and an email address into the new user registration form.

[1207] Step 2:

[1208] The terminal sends the input information to the server.

[1209] Step 3:

[1210] The server stores the submitted information in a database and sends a confirmation email to the user.

[1211] Step 4:

[1212] The user clicks on the link in the confirmation email to complete the registration.

[1213] Step 5:

[1214] The server grants access rights to the consortium web portal.

[1215] Step 6:

[1216] Users log in to a web portal to view the events calendar.

[1217] Step 7:

[1218] The server displays detailed information about the workshop (title, date and time, instructor information).

[1219] Step 8:

[1220] The user registers for the workshop of their choice.

[1221] Step 9:

[1222] The server stores the registration information in a database and sends reminder emails.

[1223] Design and implementation of a feedback system

[1224] Step 1:

[1225] Users select the AI-generated image they want to receive feedback on from their device.

[1226] Step 2:

[1227] The terminal sends the selected image file to the server.

[1228] Step 3:

[1229] The server stores the received images in a database and passes them to the image analysis system.

[1230] Step 4:

[1231] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[1232] Step 5:

[1233] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[1234] Step 6:

[1235] The user checks the feedback sent back from the server on the terminal.

[1236] Step 7:

[1237] The user makes any necessary corrections based on the feedback and sends the image back to the server.

[1238] Certification mark assignment and management

[1239] Step 1:

[1240] The server will manage anti-bias checklists for each company and make them available for download.

[1241] Step 2:

[1242] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[1243] Step 3:

[1244] The terminal transmits the self-evaluation result to the server.

[1245] Step 4:

[1246] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1247] Step 5:

[1248] The server decides whether to grant a certification mark based on the evaluation results.

[1249] Step 6:

[1250] The server decides to issue a certification mark to the company to be certified.

[1251] Step 7:

[1252] The server transmits the certification mark image data and related information to the target company.

[1253] Step 8:

[1254] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[1255] Example 1

[1256] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1257] While advances in AI technology have led to widespread use of generated images, cultural bias in generated images has become a problem. In particular, it is important to share knowledge and technology related to bias countermeasures with the participation of organizations from various industries. There is also a need for certification of businesses that have implemented appropriate bias countermeasures. The purpose of this invention is to provide a means to effectively mitigate cultural bias in AI-generated images and promote fair use.

[1258] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1259] In this invention, the server includes means for hosting a new user registration form, accepting information such as company name, contact person's name, and email address, storing the information in a database, and sending a confirmation email, means for displaying detailed information about workshops, allowing the user to register for the workshop of their choice, and sending reminder emails, means for selecting an AI-generated image for which feedback is desired from a terminal, analyzing elements within the image, generating cultural feedback using an algorithm that evaluates cultural meaning, and returning the analysis results to the user, means for the user to modify the image based on the feedback and re-upload the image, and means for downloading an anti-bias checklist, transmitting self-assessment results to the server, integrating and analyzing the evaluation results with usage history data of the feedback system, and granting a certification mark. This effectively reduces cultural bias in AI-generated images, and enables fair use by granting a certification mark to businesses that have implemented appropriate anti-bias measures.

[1260] A "server" is a computer system that provides services and resources to other computers and devices on a network.

[1261] A "terminal" is an electronic device that a user uses to access a system through an interface and to input or obtain information.

[1262] "User" means any person or entity that registers and accesses the System to use it.

[1263] The "new user registration form" is a web-based input form for entering information such as company name, contact person's name, and email address.

[1264] A "database" is an organized collection of data for efficiently storing, retrieving, and managing information.

[1265] A "confirmation email" is an email sent to notify the user that new registration has been completed.

[1266] A "workshop" is a seminar or training event for sharing and learning knowledge and skills on a specific topic.

[1267] The "Event Calendar" is a calendar-style scheduling feature that lists dates for workshops and other events.

[1268] A "reminder email" is an email that notifies the user in advance of the date and time of event participation and other important matters.

[1269] The "feedback system" is a system for generating and providing cultural feedback to users based on image analysis.

[1270] "AI-generated images" refers to images that are automatically generated using artificial intelligence technology.

[1271] "Cultural meaning" refers to the specific meaning or symbolism that elements such as colors, symbols, and backgrounds have in a particular culture or society.

[1272] "Analysis results" refer to analysis data and evaluation results generated by the image analysis system.

[1273] An "anti-bias checklist" is a list of items that businesses can use to evaluate their anti-bias measures.

[1274] "Self-assessment results" are data on the results of businesses' self-assessment of their own bias countermeasures.

[1275] A "certification mark" is a mark or logo that certifies that fair anti-bias measures are being implemented.

[1276] "Usage History Data" means records of your use of the feedback system and other system features.

[1277] An "algorithm" is a set of steps or formulas for solving a particular problem.

[1278] The present invention relates to a system for effectively reducing cultural bias in AI-generated images and promoting fair use. This system is realized by specific processing steps involving a server, a terminal, and a user.

[1279] To implement this system, the server performs the following functions:

[1280] First, the server hosts a registration form where new users can register. This is built using a web framework such as Django or Ruby on Rails. The user enters the company name, contact name, email address, etc. from the terminal and submits the data. The server receives this, stores it in a database (e.g., MySQL or PostgreSQL), and sends a confirmation email.

[1281] The server then displays the details of the workshops and allows users to register for the events they want to attend. This is done using a web portal with an events calendar, where users log in and register. The server stores this information and uses a scheduling tool like crontab or Celery to send reminder emails at set times.

[1282] In the feedback system, users select and upload the AI-generated image they want to receive feedback on from their device. The selected image file is sent to the server and saved in a temporary folder. The server then launches an image analysis system using TensorFlow, OpenCV, etc. to analyze elements in the image and evaluate their cultural meaning. The generated feedback is sent back to the user via the server. For example, it may include feedback such as, "Red backgrounds may have a specific religious meaning."

[1283] The user can then modify the image based on this feedback and re-upload it, allowing the user to receive new feedback on the altered image.

[1284] The server also provides a bias prevention checklist that users (company personnel) can download and use to conduct a self-assessment. The self-assessment results are sent to the server, which then analyzes them by integrating them with the feedback system's usage history data. Based on the analysis results, businesses that have implemented appropriate bias prevention measures are awarded a certification mark. The certification mark is sent to the company as an image file, which the company can use on its website and marketing materials.

[1285] The following are examples of prompt sentences:

[1286] "Does this image contain any cultural bias?"

[1287] "Please culturally analyze the meaning of the red background."

[1288] Following these steps will help mitigate cultural bias issues in AI-generated images, certify providers who have implemented appropriate anti-bias measures, and promote fair use.

[1289] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1290] Step 1:

[1291] The user enters the required information into the new user registration form and registers.

[1292] Input: Information such as company name, contact person's name, email address, etc.

[1293] Processing: The form data sent from the terminal is received by the server. The server performs initial data validation and stores it in the database if there are no errors.

[1294] Output: The new user information is added to the database and a confirmation email is sent.

[1295] What it does: After the server receives the HTTP request, it validates the form data and stores it in a database using Django ORM and SQL statements. It uses the Python smtplib library to send a confirmation email.

[1296] Step 2:

[1297] Users log in to the consortium web portal, check the event calendar, and register to participate in the workshop.

[1298] Input: User ID and password.

[1299] Processing: The server receives the authentication information and performs authentication. After successful authentication, the event information is retrieved from the database and displayed on the screen. The user selects the workshop they want to participate in, and the server saves that information.

[1300] Output: The user's registration data is saved to the database and a reminder email is scheduled.

[1301] What happens: The server validates the authentication information using a JSON Web Token (JWT). It retrieves the event information from the database using an SQL query and displays a dynamically generated page using an HTML template engine (e.g., Jinja2) to the user. The registration information is stored in the database, and crontab or Celery is configured to send reminder emails.

[1302] Step 3:

[1303] Users select and upload the AI-generated image they want to receive feedback on from their device.

[1304] Input: Image file.

[1305] Processing: The image file is sent from the device to the server, which saves it in a temporary folder and records its path in the database.

[1306] Output: Image files are saved on the server and ready for analysis.

[1307] Specific operation: The user selects an image file using JavaScript's File API and submits it in a form. The server receives the file using Flask or Django's FileField and saves it in the OS's temporary directory.

[1308] Step 4:

[1309] The server runs an image analysis system to generate cultural feedback.

[1310] Input: Image file path.

[1311] Processing: The server uses TensorFlow and OpenCV to analyze elements in the image and evaluates cultural meaning using generative AI models.

[1312] Output: Analysis results and feedback.

[1313] How it works: The server runs a Python script, loads a TensorFlow model, and performs image analysis. The resulting data is fed into an AI model, which generates a rating, such as "red backgrounds may have a specific religious meaning."

[1314] Step 5:

[1315] The server returns the analysis results and feedback to the user.

[1316] Input: Analysis results and feedback.

[1317] Processing: The server converts the analysis results into HTML format and returns them to the user via email or a web interface.

[1318] Output: Feedback information is provided to the user.

[1319] Specific operation: The server uses a template engine to format the analysis results into HTML format, and displays them on a dashboard that the user can view after logging in. The same content is also sent via email.

[1320] Step 6:

[1321] The user corrects the image based on the feedback and re-uploads it.

[1322] Input: The modified image file.

[1323] Processing: The corrected image is sent from the device to the server, where it is prepared for reanalysis.

[1324] Output: The corrected image files are saved on the server and can be reanalyzed.

[1325] Specific operation: The user edits the image based on the feedback using image editing software (e.g., Photoshop) and then uploads the image again using step 3.

[1326] Step 7:

[1327] The server provides an anti-bias checklist, and businesses conduct self-assessments.

[1328] Input: User (company representative) checklist download request.

[1329] Processing: The server provides the checklist file and generates a link for the user to download it.

[1330] Output: User downloads the checklist.

[1331] Specific operation: The server retrieves the checklist file from storage, performs access control, and generates a download link. The user clicks the link to download the checklist.

[1332] Step 8:

[1333] The user submits the self-assessment results to the server.

[1334] Input: Self-assessment result data.

[1335] Processing: The device sends the self-evaluation results to the server, which stores the data.

[1336] Output: The self-assessment results are recorded in a database.

[1337] Specific operation: The user evaluates the checklist and sends the results in JSON format to the server using the POST method. The server validates the data and stores it in the database.

[1338] Step 9:

[1339] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1340] Input: Self-assessment result data, feedback usage history data.

[1341] Processing: The server integrates the evaluation data with the historical data and performs data analysis.

[1342] Output: Analysis result data.

[1343] Specific operation: The server uses data analysis libraries such as Pandas and NumPy to comprehensively analyze the evaluation results and usage history and evaluate whether they meet the criteria.

[1344] Step 10:

[1345] The server decides whether to grant a certification mark and transmits the certification mark and related information to the target company.

[1346] Input: Analysis result data.

[1347] Processing: Based on the analysis results, the server determines which companies will be given the certification mark and sends the mark data by email.

[1348] Output: Certification mark image data and related information.

[1349] Specific operation: The server evaluates the analysis results and generates a list of companies that meet the certification criteria. It also sends an image file of the certification mark as an email attachment and provides related information.

[1350] Through these processing steps, it is possible to reduce cultural bias in AI-generated images and award a certification mark to businesses that have implemented appropriate anti-bias measures.

[1351] (Application example 1)

[1352] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1353] There is a need to prevent bias and cultural misunderstandings in generated images and provide fair visual information to users with diverse cultural backgrounds. However, there are limitations to what individual businesses can do independently, and the dissemination of common knowledge on how to counter bias is insufficient. Furthermore, the lack of a system for providing appropriate feedback and certifying images that have been debiased makes it difficult to provide fair images. In particular, there is a growing need for debiasing and fair authentication of images on online shopping sites.

[1354] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1355] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a multi-industry conference, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding a certification mark to businesses that implement appropriate anti-bias measures, a means for evaluating the cultural meaning of images on online shopping sites and generating feedback, a means for users to upload and re-evaluate modified images, and a means for issuing a certification mark based on the evaluation results, thereby enabling the provision and certification of fair, bias-free images.

[1356] The Council is an organization that brings together organizations from various sectors to share knowledge and techniques related to anti-bias measures.

[1357] "Anti-bias" refers to methods and activities to eliminate cultural and social biases contained in generated images and provide fair visual information.

[1358] The "feedback system" is a system that evaluates the cultural meaning of the generated images and provides users with necessary improvements and points of attention.

[1359] A "certification mark" is a seal of approval given to indicate that images and information have been properly anti-biased and culturally fair.

[1360] An "online shopping site" is an online platform where products and services can be purchased over the Internet.

[1361] "Cultural Semantic Evaluation" is the process of analyzing the elements of the generated image and evaluating the cultural and social meanings that these elements possess.

[1362] "Reevaluation" is the process by which a user reevaluates an image that has been modified to ensure an unbiased view.

[1363] "Knowledge and technology sharing" refers to activities in which participating organizations share information and technology related to anti-prejudice measures with each other.

[1364] In order to implement this invention, it is necessary to build a system based on the following contents.

[1365] System configuration

[1366] This system consists of a server, a terminal, and a user. The server operates to share knowledge and techniques related to anti-bias measures through a council and provide cultural feedback on generated images. The terminal is a device through which users upload images and receive feedback. Users can then modify and re-evaluate their images based on the feedback provided by the server.

[1367] Hardware and software used

[1368] Hardware:

[1369] Server: Any hardware specification that supports common servers such as Fedora

[1370] Device: Smart glasses or smartphone

[1371] software:

[1372] Server: Fedora

[1373] Image analysis library: OpenCV

[1374] Generative AI model: Hugging Face transformers library

[1375] Programming language: Python

[1376] Process Overview

[1377] 1. Upload an image

[1378] Users use their devices (smart glasses or smartphones) to take pictures of products and upload them to the server, which receives the images and stores them in a database.

[1379] 2. Generating Cultural Feedback

[1380] The server uses Python scripts to analyze the image. Specifically, it uses OpenCV to analyze elements in the image and then generates cultural feedback using the Hugging Face transformers library. This feedback evaluates the cultural meaning of elements such as the image's colors, symbols, and background, and suggests specific improvements to the user.

[1381] 3. Review and correct feedback

[1382] The user checks the cultural feedback sent from the server on their device, and based on the feedback, the user modifies the image and re-uploads it to the server.

[1383] 4. Issuance of certification mark

[1384] The server then re-evaluates the corrected image to ensure it has been properly treated. It then issues a certification mark and notifies the user. This certification mark can be used on online shopping sites and in marketing materials to demonstrate that the business has properly treated the bias.

[1385] Examples of concrete examples and prompts

[1386] A common example is when an image of a product that uses a lot of red is uploaded. The server provides cultural feedback to the user, prompting them with something like, "This image uses a culturally sensitive color. The use of red may be misleading to certain religions or cultures, so please change it to another color."

[1387] Using this system, users can provide unbiased and fair images and receive proper authentication.

[1388] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1389] Step 1: The user takes a picture of the product using a device (smart glasses or smartphone) and accesses the upload screen. The device sends the captured image data to the server. The input is the image data taken by the user, and the output is the image data received by the server.

[1390] Step 2: The server stores the received image data in a database, preparing it for subsequent processing by the image analysis system. The input is the transmitted image data, and the output is the image data stored in the database.

[1391] Step 3: The server starts the image analysis process. Specifically, it uses Python scripts and OpenCV to analyze elements such as color, symbols, and background in the image. The input is the image data read from the database, and the output is the analyzed data (e.g., color distribution or the detection of specific symbols).

[1392] Step 4: The server generates cultural feedback based on the parsed data using Hugging Face's transformers library. The generative AI model identifies cultural sensitivities and misleading elements and generates specific improvement instructions for the user. The input is the parsed data, and the output is a cultural feedback statement.

[1393] Step 5: The server sends the generated cultural feedback to the user's terminal and notifies the user. The input is the generated feedback sentence, and the output is the notification sent to the user.

[1394] Step 6: The user checks the feedback on the device and makes corrections to the image. After making any necessary corrections, the user re-uploads the image to the server. The input is the feedback statement and the corrected image data, and the output is the corrected image data re-uploaded to the server.

[1395] Step 7: The server re-evaluates the re-uploaded image to ensure proper bias removal. The re-evaluation process also uses OpenCV and the Hugging Face transformers library. The input is the corrected image data, and the output is the re-evaluation result and additional feedback if needed.

[1396] Step 8: Based on the evaluation results, the server issues a certification mark for images that have been properly treated for bias. The issued certification mark is sent to the user. The input is the re-evaluation result and corrected image data, and the output is the certification mark and a notification to the user.

[1397] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1398] To implement this invention, the following specific steps are required: First, a consortium with the participation of various institutions is established to provide a forum for sharing knowledge and technologies related to anti-bias measures. Second, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Furthermore, an emotion engine that recognizes the user's emotions is combined to enhance the effectiveness of the feedback.

[1399] Specific examples of consortium establishment and operation

[1400] Handling user registration

[1401] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[1402] The terminal (company representative) enters the necessary information into the form and completes the registration.

[1403] The server stores the input in a database and sends a confirmation email to the user.

[1404] Users click on the link in the confirmation email to complete registration.

[1405] The server grants access rights to the consortium web portal.

[1406] Holding workshops

[1407] Users log in to the consortium web portal and check the event calendar.

[1408] The server displays detailed information about the workshop (title, date and time, instructor information).

[1409] The user registers for the workshop of their choice.

[1410] The server stores the registration information in a database and sends reminder emails.

[1411] Specific examples of feedback system design and implementation

[1412] Image upload and analysis

[1413] Users select the AI-generated image they want to receive feedback on from their device.

[1414] The terminal sends the selected image file to the server.

[1415] The server stores the received images in a database and passes them to the image analysis system.

[1416] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[1417] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[1418] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[1419] Examples of certification marking and management

[1420] Checklist Evaluation and Certification

[1421] The server will manage anti-bias checklists for each company and make them available for download.

[1422] Users (company representatives) download the checklist and evaluate their company's measures.

[1423] The device sends the self-evaluation results to the server.

[1424] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1425] The server decides whether to grant a certification mark based on the evaluation results.

[1426] The server decides to issue a certification mark to the company to be certified.

[1427] The server transmits the certification mark image data and related information to the target company.

[1428] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[1429] Specific examples of combining emotion engines

[1430] Emotion recognition and feedback regulation

[1431] Users use emotion recognition when receiving feedback.

[1432] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[1433] The server uses the emotion engine to analyze the user's emotion. For example, if the user has a surprised expression, the emotion engine recognizes it as "surprise."

[1434] The server provides emotional data to the feedback system to evaluate the effectiveness of the generated feedback.

[1435] The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[1436] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[1437] The processing flow will be explained below.

[1438] Specific examples of consortium establishment and operation

[1439] Step 1:

[1440] The server hosts a new user registration form and provides an interface for entering information such as company name, contact name, email address, etc.

[1441] Step 2:

[1442] The terminal (company representative) enters the necessary information into the provided form and presses the registration button.

[1443] Step 3:

[1444] The server will store the entered information in a database and send a confirmation email to the appropriate address.

[1445] Step 4:

[1446] The user completes the registration by clicking the link in the confirmation email they receive.

[1447] Step 5:

[1448] The server grants access rights to the consortium web portal to users who have completed registration.

[1449] Step 6:

[1450] The user logs into the web portal using the access rights provided.

[1451] Step 7:

[1452] The server displays detailed information about the workshop (title, date and time, instructor information) on a web portal.

[1453] Step 8:

[1454] The user selects the desired workshop and registers to participate.

[1455] Step 9:

[1456] The server stores the registration information in a database and sends workshop reminder emails to users.

[1457] Specific examples of feedback system design and implementation

[1458] Step 1:

[1459] Users select the AI-generated image they want to receive feedback on from their device.

[1460] Step 2:

[1461] The terminal uploads the selected image file to the server.

[1462] Step 3:

[1463] The server receives the transmitted image file and stores it in a database.

[1464] Step 4:

[1465] The server passes the stored image to the image analysis system and begins analysis.

[1466] Step 5:

[1467] The server generates the analysis results and cultural feedback content from an image analysis system that uses algorithms to analyze elements in an image (such as colors, symbols, and backgrounds) and evaluate their cultural meaning.

[1468] Step 6:

[1469] The server sends the analysis results and feedback to the user's device, such as "red backgrounds may have a specific religious meaning."

[1470] Step 7:

[1471] The user checks the feedback sent back from the server on the terminal.

[1472] Step 8:

[1473] The user makes any necessary corrections based on the feedback and re-uploads the corrected image.

[1474] Examples of certification marking and management

[1475] Step 1:

[1476] The server will manage anti-bias checklists for each company and make them available for download by users.

[1477] Step 2:

[1478] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[1479] Step 3:

[1480] The terminal transmits the self-evaluation result to the server.

[1481] Step 4:

[1482] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1483] Step 5:

[1484] The server decides whether to grant a certification mark based on the integrated analysis results.

[1485] Step 6:

[1486] The server decides to issue a certification mark to the company to be certified.

[1487] Step 7:

[1488] The server transmits the certification mark image data and related information to the target company.

[1489] Step 8:

[1490] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[1491] Specific examples of combining emotion engines

[1492] Step 1:

[1493] The user activates the emotion recognition feature when receiving feedback.

[1494] Step 2:

[1495] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[1496] Step 3:

[1497] The server uses an emotion engine to analyze the user's emotions in real time. For example, if the user shows a surprised expression, it will recognize it as "surprise."

[1498] Step 4:

[1499] The server provides the emotional data to a feedback system to evaluate the effectiveness of the generated feedback.

[1500] Step 5:

[1501] The server adaptively adjusts the feedback content based on changes in the user's emotions. For example, if the user expresses negative emotions, the server adjusts the tone and content of the feedback provided.

[1502] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[1503] Example 2

[1504] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1505] Modern AI-generated image technology runs the risk of producing images that reflect cultural backgrounds and biases. Furthermore, feedback provided without considering user feelings often leads to inappropriate improvements. This poses a challenge, undermining fairness and the user experience surrounding the use of generated images. Furthermore, there is a lack of a system for accurately evaluating and certifying businesses that implement appropriate anti-bias measures.

[1506] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1507] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a consortium of institutions from various industries, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding certification marks to businesses that implement appropriate anti-bias measures, a means for building an emotion recognition system that analyzes user emotions and adjusts the feedback content based on the analysis, and a means for storing and integrating the results of self-evaluation by businesses in a database for evaluation. This makes it possible to provide fair and culturally appropriate feedback and propose improvements that are sensitive to user emotions. It also realizes a system that can accurately evaluate and certify businesses that implement appropriate anti-bias measures.

[1508] A "consortium" is an alliance formed by organizations from various sectors to work together to achieve a common goal.

[1509] "Anti-bias" refers to measures taken to mitigate or eliminate bias or unfairness that may arise from AI systems or algorithms.

[1510] A "feedback system" is a system for providing cultural feedback to generated images, using algorithms that analyze elements within the image and evaluate their cultural meaning.

[1511] An "emotion recognition system" is a system that has the function of analyzing a user's emotions and adjusting the feedback content based on that.

[1512] The "certification mark" is a seal of approval given to businesses that implement appropriate anti-bias measures and indicates fair efforts.

[1513] "Self-evaluation results" refer to the results of an evaluation of the effectiveness and status of anti-bias measures implemented by the business operator itself.

[1514] A "database" is a system that stores structured data and allows it to be efficiently searched and retrieved.

[1515] To implement this invention, the following specific steps are taken: First, a consortium will be established with the participation of institutions from various industries, providing a forum for sharing knowledge and technology related to bias countermeasures. This consortium will share the latest technical information and countermeasure methods, enabling participants to implement effective bias countermeasures.

[1516] Next, we build a feedback system to provide cultural feedback on the generated images, which involves the following steps:

[1517] 1. The user selects the AI-generated image for which they would like to receive feedback from their device.

[1518] 2. The device sends the selected image file to the server.

[1519] 3. The server stores the received images in a database and passes them to the image analysis system.

[1520] 4. The server launches an image analysis system to generate cultural feedback. Specifically, it uses image analysis libraries such as OpenCV and TensorFlow to analyze elements in the image (e.g., colors, symbols, background) and evaluate their cultural meaning.

[1521] 5. The server returns the analysis results and feedback to the user, for example, "The red background may have a specific religious meaning."

[1522] 6. The user checks the feedback sent back from the server on the device and makes any necessary corrections based on the feedback.

[1523] For the granting of the certification mark, the server will manage an anti-bias checklist for each company and make it available for download. Specifically, it will include the following steps:

[1524] 1. The server provides an anti-bias checklist for businesses to download.

[1525] 2. The user (company representative) downloads the checklist and evaluates their company's countermeasures.

[1526] 3. The device sends the evaluation results to the server.

[1527] 4. The server will analyze the self-evaluation results and the feedback system usage history data in order to determine whether to grant the certification mark.

[1528] 5. The server issues a certification mark based on the evaluation results and sends related information to the company.

[1529] 6. The user (certified company representative) downloads the certification mark and uses it on their company's website and marketing materials.

[1530] For the combination of emotion recognition systems, we will introduce an emotion engine to improve the quality of feedback and user experience. Specifically, we will use emotion analysis software such as Microsoft Azure Cognitive Services. The steps are as follows:

[1531] 1. The user chooses to use emotion recognition when receiving feedback.

[1532] 2. The device captures the user's facial expressions and voice and sends the emotional data to the server.

[1533] 3. The server uses an emotion engine to analyze the user's emotions. For example, it can recognize emotions such as "surprise" from the user's facial expressions and voice.

[1534] 4. The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[1535] 5. The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[1536] This system configuration allows us to provide feedback that takes into account the cultural appropriateness of generated images and user emotions, and also enables objective evaluation and certification of businesses that implement anti-bias measures.

[1537] An example of a specific prompt is, "Please analyze the cultural meaning of the image below and provide feedback."

[1538] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1539] Specific examples of consortium establishment and operation

[1540] Handling user registration

[1541] Step 1:

[1542] The user accesses the new user registration form and enters information such as the company name, the name of the person in charge, and an email address.

[1543] Input: Company name, person in charge's name, email address, etc.

[1544] Output: A submission request containing the input information

[1545] Step 2:

[1546] The terminal sends a transmission request to the server.

[1547] Input: A submission request containing input information

[1548] Output: Data sent to the server

[1549] Step 3:

[1550] The server stores the received data in a database and sends the user a confirmation email informing them that their registration is complete.

[1551] Input: Send data

[1552] Data processing: Save input information to a database

[1553] Output: Confirmation email sent

[1554] Step 4:

[1555] The user clicks on the link in the confirmation email to complete the registration.

[1556] Input: Click the link in the confirmation email

[1557] Output: Registration completion notification

[1558] Step 5:

[1559] The server grants the user access to the consortium web portal.

[1560] Data processing: Update user registration status to "Complete"

[1561] Output: Notification of access rights granted

[1562] Holding workshops

[1563] Step 1:

[1564] Users log in to the consortium web portal and check the event calendar.

[1565] Input: User ID and password

[1566] Output: Event calendar view

[1567] Step 2:

[1568] The server displays detailed information about the workshop (title, date and time, instructor information).

[1569] Data processing: Obtaining workshop information from the event database

[1570] Output: Detailed information displayed on a web page

[1571] Step 3:

[1572] The user registers for the workshop of their choice.

[1573] Input: Enter details in the entry form and click the registration button

[1574] Output: Registration request sent

[1575] Step 4:

[1576] The server stores the registration information in a database and sends reminder emails before the workshop begins.

[1577] Data processing: Update participant list, send reminder emails

[1578] Output: Reminder email sent successfully

[1579] Specific examples of feedback system design and implementation

[1580] Image upload and analysis

[1581] Step 1:

[1582] Users select the AI-generated image they want to receive feedback on from their device.

[1583] Input: Select image file

[1584] Output: Selected image file

[1585] Step 2:

[1586] The terminal sends the selected image file to the server.

[1587] Input: Selected image file

[1588] Output: Request to send image file

[1589] Step 3:

[1590] The server stores the received images in a database and passes them to the image analysis system.

[1591] Input: Image file transfer request

[1592] Data processing: Save image data in a database

[1593] Output: Passing image data to an image analysis system

[1594] Step 4:

[1595] The server runs an image analysis system to generate cultural feedback, specifically using OpenCV and TensorFlow to analyze elements in the image and evaluate their cultural meaning.

[1596] Input: Image data

[1597] Data calculation: Analyze elements in the image (color, symbol, background)

[1598] Output: Generated feedback

[1599] Step 5:

[1600] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[1601] Input: Generated feedback

[1602] Data processing: Send feedback to users

[1603] Output: Feedback sent successfully

[1604] Step 6:

[1605] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[1606] Input: Check feedback

[1607] Output: Image correction completed

[1608] Examples of certification marking and management

[1609] Checklist Evaluation and Certification

[1610] Step 1:

[1611] The server will manage anti-bias checklists for each company and make them available for download.

[1612] Input: Checklist PDF file

[1613] Data processing: Upload the checklist to the web portal

[1614] Output: Publish download link

[1615] Step 2:

[1616] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[1617] Input:Download checklist

[1618] Output: Checklist evaluation

[1619] Step 3:

[1620] The terminal transmits the self-evaluation result to the server.

[1621] Input: Self-assessment results

[1622] Output: Send request

[1623] Step 4:

[1624] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1625] Input: Self-assessment results and usage history data

[1626] Data processing: Integrated analysis of evaluation data

[1627] Output: Evaluation results

[1628] Step 5:

[1629] The server decides whether to grant a certification mark based on the evaluation results.

[1630] Input: Evaluation result

[1631] Data processing: Judgment in accordance with certification standards

[1632] Output: Certification mark grant decision

[1633] Step 6:

[1634] The server issues a certification mark to the certified company and transmits related information.

[1635] Input: Certification mark grant decision

[1636] Data processing: generation of certification marks and certificates

[1637] Output: Certification mark sent successfully

[1638] Step 7:

[1639] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[1640] Input: Download certification mark

[1641] Output: Use of certification mark

[1642] Specific examples of combining emotion engines

[1643] Emotion recognition and feedback regulation

[1644] Step 1:

[1645] The user may choose to use emotion recognition functionality when receiving feedback.

[1646] Input: Select emotion recognition feature

[1647] Output: Emotion data collection begins

[1648] Step 2:

[1649] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[1650] Input: User's facial expressions and voice

[1651] Data processing: Real-time capture and data transmission

[1652] Output: Emotion data transmission

[1653] Step 3:

[1654] The server uses an emotion engine to analyze the user's emotions, for example, recognizing emotions such as "surprise" from the user's facial expressions and voice.

[1655] Input: Emotion data

[1656] Data Computing: Sentiment Analysis

[1657] Output: Emotional state

[1658] Step 4:

[1659] The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[1660] Input: Emotional state

[1661] Data processing: Integrated analysis of emotion data and feedback data

[1662] Output: Feedback effect evaluation results

[1663] Step 5:

[1664] The server adaptively adjusts the feedback content based on changes in the user's emotions, for example, softening the tone and content of the feedback if the user expresses negative emotions.

[1665] Input: Feedback effect evaluation results

[1666] Data processing: Adjustment of feedback content

[1667] Output: Adaptively adjusted feedback

[1668] (Application example 2)

[1669] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1670] Advertising materials created by advertising creators often contain bias that ignores cultural background. Such biases can foster misunderstandings and prejudices against specific cultures or ethnicities, hindering the provision of fair advertising. Furthermore, it is difficult to properly capture the emotional responses of users viewing ads and provide feedback, leading to a lack of improvements to maximize the effectiveness of advertising. There is a need to address these issues.

[1671] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1672] In this invention, the server includes means for sharing knowledge and technology related to bias countermeasures through a consortium of institutions from various industries, means for designing and implementing a feedback system that provides cultural feedback on generated images, means for granting a certification mark to businesses that implement appropriate bias countermeasures, means for providing cultural feedback on advertising materials created by advertising creators, and means for analyzing user emotions using an emotion recognition engine and adjusting the feedback content based on the results. This makes it possible to reduce cultural bias in advertising materials and provide appropriate feedback according to user emotions.

[1673] "Multi-industry institutions" is a general term for businesses and organizations in different industries, such as the advertising industry, software development industry, design industry, and marketing industry.

[1674] A "consortium" is a cooperative system formed by multiple companies or institutions working together to achieve a specific purpose or solve a problem.

[1675] "Anti-bias" refers to methods and efforts to eliminate prejudices and preconceptions based on particular cultures or social backgrounds.

[1676] A "feedback system" is a system that provides evaluations and opinions on generated images and advertising materials from a cultural perspective.

[1677] "Elements in an image" are components such as colors, symbols, backgrounds, and objects contained in an image.

[1678] "Cultural feedback" refers to evaluations and opinions about how generated images and advertising materials will be received based on a particular culture or social context.

[1679] A "certification mark" is an official certification symbol given to businesses or products that meet certain standards.

[1680] An "emotion recognition engine" is an algorithm or software that analyzes data such as a user's facial expressions and voice, and classifies and recognizes the user's emotions.

[1681] An "advertising creator" is a professional who creates advertising content and design.

[1682] "User sentiment" refers to the psychological and emotional response a user has when viewing an ad.

[1683] To implement this invention, the following specific system configuration and processes are required: Use the components of the server, terminal, and user, and clarify the roles of each.

[1684] Server Roles

[1685] The server is responsible for the main data processing and analysis functions. High-performance servers are recommended as the hardware to be used. Software such as Python, Flask, Keras, and OpenCV are also required.

[1686] 1. The server receives generated images and advertising materials from the user and stores the data in a database.

[1687] 2. The server passes the stored image data to an analysis algorithm to generate cultural feedback. The algorithm analyzes elements in the image (colors, symbols, background, etc.) and evaluates their cultural meaning.

[1688] 3. The server uses an emotion recognition engine to analyze the user's emotional data (facial expressions and voice) and recognize the user's emotional state. For example, it classifies and recognizes emotions such as "surprise," "happiness," and "sadness."

[1689] Device Role

[1690] A terminal is a device that provides a user interface, such as a smartphone or a head-mounted display.

[1691] 1. The terminal provides an interface for ad creators to upload their created ad materials to the server.

[1692] 2. The device displays the cultural feedback and emotion recognition results returned from the server to the user.

[1693] 3. The device provides tools for modifying advertising materials in real time, helping users improve their ads based on feedback.

[1694] User Roles

[1695] Users are creators who create advertisements and people involved in evaluating advertisements.

[1696] 1. The user uploads advertising material from their device to the server.

[1697] 2. The user checks the cultural feedback and emotion recognition results provided by the server and modifies the advertising material as needed.

[1698] 3. The user decides on the final advertising material based on the feedback and publishes it.

[1699] Specific examples of processing

[1700] For example, when an ad creator uploads a new ad image they've created, they might get feedback like this:

[1701] Cultural feedback: "The image contains a red background, which may have religious significance in parts of Asia. Consider changing the background color or being more considerate of the context."

[1702] Emotion-aware feedback: "We detected a surprised expression from the user. Consider adjusting the exciting elements in your ad."

[1703] This allows ad creators to easily identify specific areas for improvement and create more effective and culturally appropriate ads.

[1704] Through this invention, it is possible to effectively reduce cultural bias in advertising materials and provide appropriate feedback according to the user's emotions, thereby realizing fair and effective advertising creation.

[1705] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1706] Step 1:

[1707] The server receives advertising materials (images or videos) uploaded by users (advertising creators). Specifically, users select advertising materials using the device interface and send them to the server. This input data is stored on the server.

[1708] Step 2:

[1709] The server then passes the stored advertising material to an image analysis algorithm, which generates feedback that includes cultural meaning. Specifically, the algorithm analyzes elements in the image (color, symbols, background, etc.) and evaluates their cultural meaning. The resulting cultural feedback is then output.

[1710] Step 3:

[1711] The server sends emotional data (facial expressions and voice) of users viewing advertisements to the emotion recognition engine. When a user views an advertisement using a device, the device captures the user's facial expressions and voice and sends the data to the server. The emotion recognition engine analyzes this data and recognizes emotional states such as "surprise," "happiness," and "sadness." The analysis results in an emotion label for the user being output.

[1712] Step 4:

[1713] The server integrates the generated cultural feedback with the emotion recognition results and returns feedback to the user, including suggestions for correction and improvement. Specifically, the server combines the cultural feedback and emotion labels and returns them to the device in a format that is easy for the user to understand. For example, specific instructions may be output, such as, "Consider changing the background color because a red background may have a specific religious meaning. Additionally, since the user's surprised expression was detected, please adjust the provocative elements in the advertisement."

[1714] Step 5:

[1715] The user can use their device to check the feedback provided by the server and modify the advertising material as necessary. The user can then edit the advertising material based on the feedback and upload it back to the server. By repeating this process, the optimal advertising material is created.

[1716] Step 6:

[1717] The server will award a certification mark to businesses that have implemented appropriate anti-bias measures. Specifically, the server will integrate and analyze the evaluation results of the businesses' measures and their usage history of the feedback system, and will issue a certification mark if the standards are met. This certification mark will be provided to the certified businesses to display in their advertisements and on their websites.

[1718] Following this trend, advertising creators will be able to eliminate cultural bias and create optimal advertising materials that take user emotions into consideration.

[1719] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1720] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1721] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1722] [Fourth embodiment]

[1723] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1724] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1725] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1726] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1727] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1728] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1729] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1730] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1731] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1732] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1733] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1734] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1735] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1736] To implement this invention, the following specific steps are required: First, a consortium is established with the participation of institutions from various industries. The consortium will serve as a forum for sharing knowledge and technology related to bias prevention. Next, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Finally, a system is built to award a certification mark to businesses that implement appropriate bias prevention measures.

[1737] Specific examples of consortium establishment and operation

[1738] Step 1: User Registration

[1739] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[1740] The terminal (company representative) enters the necessary information into the form and completes the registration.

[1741] The server stores the input in a database and sends a confirmation email.

[1742] Step 2: Host a workshop

[1743] Users log in to the consortium web portal and check the event calendar.

[1744] The server displays detailed information about the workshop (title, date and time, instructor information).

[1745] The user registers for the workshop of their choice.

[1746] The server stores the registration information in a database and sends reminder emails.

[1747] Specific examples of feedback system design and implementation

[1748] Step 1: Upload an image

[1749] Users select the AI-generated image they want to receive feedback on from their device.

[1750] The terminal transmits the selected image file to the server.

[1751] The server stores the received images in a database and passes them to the image analysis system.

[1752] Step 2: Generate feedback

[1753] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[1754] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[1755] Step 3: Review and correct your feedback

[1756] The user checks the feedback sent back from the server on the terminal.

[1757] The user makes any necessary corrections based on the feedback and sends the image to the server again.

[1758] Examples of certification marking and management

[1759] Step 1: Evaluate the checklist

[1760] The server will manage anti-bias checklists for each company and make them available for download.

[1761] Users (company representatives) download the checklist and evaluate their company's measures.

[1762] The terminal transmits the self-evaluation result to the server.

[1763] Step 2: Analyze the results

[1764] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1765] The server decides whether to grant a certification mark based on the evaluation results.

[1766] Step 3: Issuance of the certification mark

[1767] The server decides to issue a certification mark to the company to be certified.

[1768] The server transmits the certification mark image data and related information to the target company.

[1769] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[1770] The above is a concrete example of how to implement the present invention. By using this system, we can effectively mitigate bias issues in AI-generated images in Asia and promote fair use.

[1771] The processing flow will be explained below.

[1772] Establishment and operation of the consortium

[1773] Step 1:

[1774] The user enters information such as the company name, the name of the person in charge, and an email address into the new user registration form.

[1775] Step 2:

[1776] The terminal sends the input information to the server.

[1777] Step 3:

[1778] The server stores the submitted information in a database and sends a confirmation email to the user.

[1779] Step 4:

[1780] The user clicks on the link in the confirmation email to complete the registration.

[1781] Step 5:

[1782] The server grants access rights to the consortium web portal.

[1783] Step 6:

[1784] Users log in to a web portal to view the events calendar.

[1785] Step 7:

[1786] The server displays detailed information about the workshop (title, date and time, instructor information).

[1787] Step 8:

[1788] The user registers for the workshop of their choice.

[1789] Step 9:

[1790] The server stores the registration information in a database and sends reminder emails.

[1791] Design and implementation of a feedback system

[1792] Step 1:

[1793] Users select the AI-generated image they want to receive feedback on from their device.

[1794] Step 2:

[1795] The terminal sends the selected image file to the server.

[1796] Step 3:

[1797] The server stores the received images in a database and passes them to the image analysis system.

[1798] Step 4:

[1799] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[1800] Step 5:

[1801] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[1802] Step 6:

[1803] The user checks the feedback sent back from the server on the terminal.

[1804] Step 7:

[1805] The user makes any necessary corrections based on the feedback and sends the image back to the server.

[1806] Certification mark assignment and management

[1807] Step 1:

[1808] The server will manage anti-bias checklists for each company and make them available for download.

[1809] Step 2:

[1810] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[1811] Step 3:

[1812] The terminal transmits the self-evaluation result to the server.

[1813] Step 4:

[1814] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1815] Step 5:

[1816] The server decides whether to grant a certification mark based on the evaluation results.

[1817] Step 6:

[1818] The server decides to issue a certification mark to the company to be certified.

[1819] Step 7:

[1820] The server transmits the certification mark image data and related information to the target company.

[1821] Step 8:

[1822] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[1823] Example 1

[1824] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1825] While advances in AI technology have led to widespread use of generated images, cultural bias in generated images has become a problem. In particular, it is important to share knowledge and technology related to bias countermeasures with the participation of organizations from various industries. There is also a need for certification of businesses that have implemented appropriate bias countermeasures. The purpose of this invention is to provide a means to effectively mitigate cultural bias in AI-generated images and promote fair use.

[1826] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1827] In this invention, the server includes means for hosting a new user registration form, accepting information such as company name, contact person's name, and email address, storing the information in a database, and sending a confirmation email, means for displaying detailed information about workshops, allowing the user to register for the workshop of their choice, and sending reminder emails, means for selecting an AI-generated image for which feedback is desired from a terminal, analyzing elements within the image, generating cultural feedback using an algorithm that evaluates cultural meaning, and returning the analysis results to the user, means for the user to modify the image based on the feedback and re-upload the image, and means for downloading an anti-bias checklist, transmitting self-assessment results to the server, integrating and analyzing the evaluation results with usage history data of the feedback system, and granting a certification mark. This effectively reduces cultural bias in AI-generated images, and enables fair use by granting a certification mark to businesses that have implemented appropriate anti-bias measures.

[1828] A "server" is a computer system that provides services and resources to other computers and devices on a network.

[1829] A "terminal" is an electronic device that a user uses to access a system through an interface and to input or obtain information.

[1830] "User" means any person or entity that registers and accesses the System to use it.

[1831] The "new user registration form" is a web-based input form for entering information such as company name, contact person's name, and email address.

[1832] A "database" is an organized collection of data for efficiently storing, retrieving, and managing information.

[1833] A "confirmation email" is an email sent to notify the user that new registration has been completed.

[1834] A "workshop" is a seminar or training event for sharing and learning knowledge and skills on a specific topic.

[1835] The "Event Calendar" is a calendar-style scheduling feature that lists dates for workshops and other events.

[1836] A "reminder email" is an email that notifies the user in advance of the date and time of event participation and other important matters.

[1837] The "feedback system" is a system for generating and providing cultural feedback to users based on image analysis.

[1838] "AI-generated images" refers to images that are automatically generated using artificial intelligence technology.

[1839] "Cultural meaning" refers to the specific meaning or symbolism that elements such as colors, symbols, and backgrounds have in a particular culture or society.

[1840] "Analysis results" refer to analysis data and evaluation results generated by the image analysis system.

[1841] An "anti-bias checklist" is a list of items that businesses can use to evaluate their anti-bias measures.

[1842] "Self-assessment results" are data on the results of businesses' self-assessment of their own bias countermeasures.

[1843] A "certification mark" is a mark or logo that certifies that fair anti-bias measures are being implemented.

[1844] "Usage History Data" means records of your use of the feedback system and other system features.

[1845] An "algorithm" is a set of steps or formulas for solving a particular problem.

[1846] The present invention relates to a system for effectively reducing cultural bias in AI-generated images and promoting fair use. This system is realized by specific processing steps involving a server, a terminal, and a user.

[1847] To implement this system, the server performs the following functions:

[1848] First, the server hosts a registration form where new users can register. This is built using a web framework such as Django or Ruby on Rails. The user enters the company name, contact name, email address, etc. from the terminal and submits the data. The server receives this, stores it in a database (e.g., MySQL or PostgreSQL), and sends a confirmation email.

[1849] The server then displays the details of the workshops and allows users to register for the events they want to attend. This is done using a web portal with an events calendar, where users log in and register. The server stores this information and uses a scheduling tool like crontab or Celery to send reminder emails at set times.

[1850] In the feedback system, users select and upload the AI-generated image they want to receive feedback on from their device. The selected image file is sent to the server and saved in a temporary folder. The server then launches an image analysis system using TensorFlow, OpenCV, etc. to analyze elements in the image and evaluate their cultural meaning. The generated feedback is sent back to the user via the server. For example, it may include feedback such as, "Red backgrounds may have a specific religious meaning."

[1851] The user can then modify the image based on this feedback and re-upload it, allowing the user to receive new feedback on the altered image.

[1852] The server also provides a bias prevention checklist that users (company personnel) can download and use to conduct a self-assessment. The self-assessment results are sent to the server, which then analyzes them by integrating them with the feedback system's usage history data. Based on the analysis results, businesses that have implemented appropriate bias prevention measures are awarded a certification mark. The certification mark is sent to the company as an image file, which the company can use on its website and marketing materials.

[1853] The following are examples of prompt sentences:

[1854] "Does this image contain any cultural bias?"

[1855] "Please culturally analyze the meaning of the red background."

[1856] Following these steps will help mitigate cultural bias issues in AI-generated images, certify providers who have implemented appropriate anti-bias measures, and promote fair use.

[1857] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1858] Step 1:

[1859] The user enters the required information into the new user registration form and registers.

[1860] Input: Information such as company name, contact person's name, email address, etc.

[1861] Processing: The form data sent from the terminal is received by the server. The server performs initial data validation and stores it in the database if there are no errors.

[1862] Output: The new user information is added to the database and a confirmation email is sent.

[1863] What it does: After the server receives the HTTP request, it validates the form data and stores it in a database using Django ORM and SQL statements. It uses the Python smtplib library to send a confirmation email.

[1864] Step 2:

[1865] Users log in to the consortium web portal, check the event calendar, and register to participate in the workshop.

[1866] Input: User ID and password.

[1867] Processing: The server receives the authentication information and performs authentication. After successful authentication, the event information is retrieved from the database and displayed on the screen. The user selects the workshop they want to participate in, and the server saves that information.

[1868] Output: The user's registration data is saved to the database and a reminder email is scheduled.

[1869] What happens: The server validates the authentication information using a JSON Web Token (JWT). It retrieves the event information from the database using an SQL query and displays a dynamically generated page using an HTML template engine (e.g., Jinja2) to the user. The registration information is stored in the database, and crontab or Celery is configured to send reminder emails.

[1870] Step 3:

[1871] Users select and upload the AI-generated image they want to receive feedback on from their device.

[1872] Input: Image file.

[1873] Processing: The image file is sent from the device to the server, which saves it in a temporary folder and records its path in the database.

[1874] Output: Image files are saved on the server and ready for analysis.

[1875] Specific operation: The user selects an image file using JavaScript's File API and submits it in a form. The server receives the file using Flask or Django's FileField and saves it in the OS's temporary directory.

[1876] Step 4:

[1877] The server runs an image analysis system to generate cultural feedback.

[1878] Input: Image file path.

[1879] Processing: The server uses TensorFlow and OpenCV to analyze elements in the image and evaluates cultural meaning using generative AI models.

[1880] Output: Analysis results and feedback.

[1881] How it works: The server runs a Python script, loads a TensorFlow model, and performs image analysis. The resulting data is fed into an AI model, which generates a rating, such as "red backgrounds may have a specific religious meaning."

[1882] Step 5:

[1883] The server returns the analysis results and feedback to the user.

[1884] Input: Analysis results and feedback.

[1885] Processing: The server converts the analysis results into HTML format and returns them to the user via email or a web interface.

[1886] Output: Feedback information is provided to the user.

[1887] Specific operation: The server uses a template engine to format the analysis results into HTML format, and displays them on a dashboard that the user can view after logging in. The same content is also sent via email.

[1888] Step 6:

[1889] The user corrects the image based on the feedback and re-uploads it.

[1890] Input: The modified image file.

[1891] Processing: The corrected image is sent from the device to the server, where it is prepared for reanalysis.

[1892] Output: The corrected image files are saved on the server and can be reanalyzed.

[1893] Specific operation: The user edits the image based on the feedback using image editing software (e.g., Photoshop) and then uploads the image again using step 3.

[1894] Step 7:

[1895] The server provides an anti-bias checklist, and businesses conduct self-assessments.

[1896] Input: User (company representative) checklist download request.

[1897] Processing: The server provides the checklist file and generates a link for the user to download it.

[1898] Output: User downloads the checklist.

[1899] Specific operation: The server retrieves the checklist file from storage, performs access control, and generates a download link. The user clicks the link to download the checklist.

[1900] Step 8:

[1901] The user submits the self-assessment results to the server.

[1902] Input: Self-assessment result data.

[1903] Processing: The device sends the self-evaluation results to the server, which stores the data.

[1904] Output: The self-assessment results are recorded in a database.

[1905] Specific operation: The user evaluates the checklist and sends the results in JSON format to the server using the POST method. The server validates the data and stores it in the database.

[1906] Step 9:

[1907] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1908] Input: Self-assessment result data, feedback usage history data.

[1909] Processing: The server integrates the evaluation data with the historical data and performs data analysis.

[1910] Output: Analysis result data.

[1911] Specific operation: The server uses data analysis libraries such as Pandas and NumPy to comprehensively analyze the evaluation results and usage history and evaluate whether they meet the criteria.

[1912] Step 10:

[1913] The server decides whether to grant a certification mark and transmits the certification mark and related information to the target company.

[1914] Input: Analysis result data.

[1915] Processing: Based on the analysis results, the server determines which companies will be given the certification mark and sends the mark data by email.

[1916] Output: Certification mark image data and related information.

[1917] Specific operation: The server evaluates the analysis results and generates a list of companies that meet the certification criteria. It also sends an image file of the certification mark as an email attachment and provides related information.

[1918] Through these processing steps, it is possible to reduce cultural bias in AI-generated images and award a certification mark to businesses that have implemented appropriate anti-bias measures.

[1919] (Application example 1)

[1920] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1921] There is a need to prevent bias and cultural misunderstandings in generated images and provide fair visual information to users with diverse cultural backgrounds. However, there are limitations to what individual businesses can do independently, and the dissemination of common knowledge on how to counter bias is insufficient. Furthermore, the lack of a system for providing appropriate feedback and certifying images that have been debiased makes it difficult to provide fair images. In particular, there is a growing need for debiasing and fair authentication of images on online shopping sites.

[1922] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1923] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a multi-industry conference, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding a certification mark to businesses that implement appropriate anti-bias measures, a means for evaluating the cultural meaning of images on online shopping sites and generating feedback, a means for users to upload and re-evaluate modified images, and a means for issuing a certification mark based on the evaluation results, thereby enabling the provision and certification of fair, bias-free images.

[1924] The Council is an organization that brings together organizations from various sectors to share knowledge and techniques related to anti-bias measures.

[1925] "Anti-bias" refers to methods and activities to eliminate cultural and social biases contained in generated images and provide fair visual information.

[1926] The "feedback system" is a system that evaluates the cultural meaning of the generated images and provides users with necessary improvements and points of attention.

[1927] A "certification mark" is a seal of approval given to indicate that images and information have been properly anti-biased and culturally fair.

[1928] An "online shopping site" is an online platform where products and services can be purchased over the Internet.

[1929] "Cultural Semantic Evaluation" is the process of analyzing the elements of the generated image and evaluating the cultural and social meanings that these elements possess.

[1930] "Reevaluation" is the process by which a user reevaluates an image that has been modified to ensure an unbiased view.

[1931] "Knowledge and technology sharing" refers to activities in which participating organizations share information and technology related to anti-prejudice measures with each other.

[1932] In order to implement this invention, it is necessary to build a system based on the following contents.

[1933] System configuration

[1934] This system consists of a server, a terminal, and a user. The server operates to share knowledge and techniques related to anti-bias measures through a council and provide cultural feedback on generated images. The terminal is a device through which users upload images and receive feedback. Users can then modify and re-evaluate their images based on the feedback provided by the server.

[1935] Hardware and software used

[1936] Hardware:

[1937] Server: Any hardware specification that supports common servers such as Fedora

[1938] Device: Smart glasses or smartphone

[1939] software:

[1940] Server: Fedora

[1941] Image analysis library: OpenCV

[1942] Generative AI model: Hugging Face transformers library

[1943] Programming language: Python

[1944] Process Overview

[1945] 1. Upload an image

[1946] Users use their devices (smart glasses or smartphones) to take pictures of products and upload them to the server, which receives the images and stores them in a database.

[1947] 2. Generating Cultural Feedback

[1948] The server uses Python scripts to analyze the image. Specifically, it uses OpenCV to analyze elements in the image and then generates cultural feedback using the Hugging Face transformers library. This feedback evaluates the cultural meaning of elements such as the image's colors, symbols, and background, and suggests specific improvements to the user.

[1949] 3. Review and correct feedback

[1950] The user checks the cultural feedback sent from the server on their device, and based on the feedback, the user modifies the image and re-uploads it to the server.

[1951] 4. Issuance of certification mark

[1952] The server then re-evaluates the corrected image to ensure it has been properly treated. It then issues a certification mark and notifies the user. This certification mark can be used on online shopping sites and in marketing materials to demonstrate that the business has properly treated the bias.

[1953] Examples of concrete examples and prompts

[1954] A common example is when an image of a product that uses a lot of red is uploaded. The server provides cultural feedback to the user, prompting them with something like, "This image uses a culturally sensitive color. The use of red may be misleading to certain religions or cultures, so please change it to another color."

[1955] Using this system, users can provide unbiased and fair images and receive proper authentication.

[1956] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1957] Step 1: The user takes a picture of the product using a device (smart glasses or smartphone) and accesses the upload screen. The device sends the captured image data to the server. The input is the image data taken by the user, and the output is the image data received by the server.

[1958] Step 2: The server stores the received image data in a database, preparing it for subsequent processing by the image analysis system. The input is the transmitted image data, and the output is the image data stored in the database.

[1959] Step 3: The server starts the image analysis process. Specifically, it uses Python scripts and OpenCV to analyze elements such as color, symbols, and background in the image. The input is the image data read from the database, and the output is the analyzed data (e.g., color distribution or the detection of specific symbols).

[1960] Step 4: The server generates cultural feedback based on the parsed data using Hugging Face's transformers library. The generative AI model identifies cultural sensitivities and misleading elements and generates specific improvement instructions for the user. The input is the parsed data, and the output is a cultural feedback statement.

[1961] Step 5: The server sends the generated cultural feedback to the user's terminal and notifies the user. The input is the generated feedback sentence, and the output is the notification sent to the user.

[1962] Step 6: The user checks the feedback on the device and makes corrections to the image. After making any necessary corrections, the user re-uploads the image to the server. The input is the feedback statement and the corrected image data, and the output is the corrected image data re-uploaded to the server.

[1963] Step 7: The server re-evaluates the re-uploaded image to ensure proper bias removal. The re-evaluation process also uses OpenCV and the Hugging Face transformers library. The input is the corrected image data, and the output is the re-evaluation result and additional feedback if needed.

[1964] Step 8: Based on the evaluation results, the server issues a certification mark for images that have been properly treated for bias. The issued certification mark is sent to the user. The input is the re-evaluation result and corrected image data, and the output is the certification mark and a notification to the user.

[1965] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1966] To implement this invention, the following specific steps are required: First, a consortium with the participation of various institutions is established to provide a forum for sharing knowledge and technologies related to anti-bias measures. Second, a feedback system is built to provide cultural feedback on generated images. This feedback system uses an algorithm that analyzes elements within the image and evaluates their cultural meaning. Furthermore, an emotion engine that recognizes the user's emotions is combined to enhance the effectiveness of the feedback.

[1967] Specific examples of consortium establishment and operation

[1968] Handling user registration

[1969] The server hosts a new user registration form, accepting information such as company name, contact name, and email address.

[1970] The terminal (company representative) enters the necessary information into the form and completes the registration.

[1971] The server stores the input in a database and sends a confirmation email to the user.

[1972] Users click on the link in the confirmation email to complete registration.

[1973] The server grants access rights to the consortium web portal.

[1974] Holding workshops

[1975] Users log in to the consortium web portal and check the event calendar.

[1976] The server displays detailed information about the workshop (title, date and time, instructor information).

[1977] The user registers for the workshop of their choice.

[1978] The server stores the registration information in a database and sends reminder emails.

[1979] Specific examples of feedback system design and implementation

[1980] Image upload and analysis

[1981] Users select the AI-generated image they want to receive feedback on from their device.

[1982] The terminal sends the selected image file to the server.

[1983] The server stores the received images in a database and passes them to the image analysis system.

[1984] The server runs an image analysis system to generate cultural feedback. This system uses algorithms to analyze elements in an image (colors, symbols, background, etc.) and evaluate their cultural meaning.

[1985] The server sends the analysis results and feedback back to the user, for example, "The red background may have a specific religious meaning."

[1986] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[1987] Examples of certification marking and management

[1988] Checklist Evaluation and Certification

[1989] The server will manage anti-bias checklists for each company and make them available for download.

[1990] Users (company representatives) download the checklist and evaluate their company's measures.

[1991] The device sends the self-evaluation results to the server.

[1992] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[1993] The server decides whether to grant a certification mark based on the evaluation results.

[1994] The server decides to issue a certification mark to the company to be certified.

[1995] The server transmits the certification mark image data and related information to the target company.

[1996] Users (certified company representatives) download the certification mark and use it on their company's website and marketing materials.

[1997] Specific examples of combining emotion engines

[1998] Emotion recognition and feedback regulation

[1999] Users use emotion recognition when receiving feedback.

[2000] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[2001] The server uses the emotion engine to analyze the user's emotion. For example, if the user has a surprised expression, the emotion engine recognizes it as "surprise."

[2002] The server provides emotional data to the feedback system to evaluate the effectiveness of the generated feedback.

[2003] The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[2004] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[2005] The processing flow will be explained below.

[2006] Specific examples of consortium establishment and operation

[2007] Step 1:

[2008] The server hosts a new user registration form and provides an interface for entering information such as company name, contact name, email address, etc.

[2009] Step 2:

[2010] The terminal (company representative) enters the necessary information into the provided form and presses the registration button.

[2011] Step 3:

[2012] The server will store the entered information in a database and send a confirmation email to the appropriate address.

[2013] Step 4:

[2014] The user completes the registration by clicking the link in the confirmation email they receive.

[2015] Step 5:

[2016] The server grants access rights to the consortium web portal to users who have completed registration.

[2017] Step 6:

[2018] The user logs into the web portal using the access rights provided.

[2019] Step 7:

[2020] The server displays detailed information about the workshop (title, date and time, instructor information) on a web portal.

[2021] Step 8:

[2022] The user selects the desired workshop and registers to participate.

[2023] Step 9:

[2024] The server stores the registration information in a database and sends workshop reminder emails to users.

[2025] Specific examples of feedback system design and implementation

[2026] Step 1:

[2027] Users select the AI-generated image they want to receive feedback on from their device.

[2028] Step 2:

[2029] The terminal uploads the selected image file to the server.

[2030] Step 3:

[2031] The server receives the transmitted image file and stores it in a database.

[2032] Step 4:

[2033] The server passes the stored image to the image analysis system and begins analysis.

[2034] Step 5:

[2035] The server generates the analysis results and cultural feedback content from an image analysis system that uses algorithms to analyze elements in an image (such as colors, symbols, and backgrounds) and evaluate their cultural meaning.

[2036] Step 6:

[2037] The server sends the analysis results and feedback to the user's device, such as "red backgrounds may have a specific religious meaning."

[2038] Step 7:

[2039] The user checks the feedback sent back from the server on the terminal.

[2040] Step 8:

[2041] The user makes any necessary corrections based on the feedback and re-uploads the corrected image.

[2042] Examples of certification marking and management

[2043] Step 1:

[2044] The server will manage anti-bias checklists for each company and make them available for download by users.

[2045] Step 2:

[2046] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[2047] Step 3:

[2048] The terminal transmits the self-evaluation result to the server.

[2049] Step 4:

[2050] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[2051] Step 5:

[2052] The server decides whether to grant a certification mark based on the integrated analysis results.

[2053] Step 6:

[2054] The server decides to issue a certification mark to the company to be certified.

[2055] Step 7:

[2056] The server transmits the certification mark image data and related information to the target company.

[2057] Step 8:

[2058] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[2059] Specific examples of combining emotion engines

[2060] Step 1:

[2061] The user activates the emotion recognition feature when receiving feedback.

[2062] Step 2:

[2063] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[2064] Step 3:

[2065] The server uses an emotion engine to analyze the user's emotions in real time. For example, if the user shows a surprised expression, it will recognize it as "surprise."

[2066] Step 4:

[2067] The server provides the emotional data to a feedback system to evaluate the effectiveness of the generated feedback.

[2068] Step 5:

[2069] The server adaptively adjusts the feedback content based on changes in the user's emotions. For example, if the user expresses negative emotions, the server adjusts the tone and content of the feedback provided.

[2070] The above is a specific example of how to implement the present invention. Using this system can effectively mitigate bias issues in AI-generated images in Asia and promote fair use. Furthermore, by combining it with an emotion engine, it is possible to further improve the user experience.

[2071] Example 2

[2072] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2073] Modern AI-generated image technology runs the risk of producing images that reflect cultural backgrounds and biases. Furthermore, feedback provided without considering user feelings often leads to inappropriate improvements. This poses a challenge, undermining fairness and the user experience surrounding the use of generated images. Furthermore, there is a lack of a system for accurately evaluating and certifying businesses that implement appropriate anti-bias measures.

[2074] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[2075] In this invention, the server includes a means for sharing knowledge and technology related to anti-bias measures through a consortium of institutions from various industries, a means for designing and implementing a feedback system that provides cultural feedback on generated images, a means for awarding certification marks to businesses that implement appropriate anti-bias measures, a means for building an emotion recognition system that analyzes user emotions and adjusts the feedback content based on the analysis, and a means for storing and integrating the results of self-evaluation by businesses in a database for evaluation. This makes it possible to provide fair and culturally appropriate feedback and propose improvements that are sensitive to user emotions. It also realizes a system that can accurately evaluate and certify businesses that implement appropriate anti-bias measures.

[2076] A "consortium" is an alliance formed by organizations from various sectors to work together to achieve a common goal.

[2077] "Anti-bias" refers to measures taken to mitigate or eliminate bias or unfairness that may arise from AI systems or algorithms.

[2078] A "feedback system" is a system for providing cultural feedback to generated images, using algorithms that analyze elements within the image and evaluate their cultural meaning.

[2079] An "emotion recognition system" is a system that has the function of analyzing a user's emotions and adjusting the feedback content based on that.

[2080] The "certification mark" is a seal of approval given to businesses that implement appropriate anti-bias measures and indicates fair efforts.

[2081] "Self-evaluation results" refer to the results of an evaluation of the effectiveness and status of anti-bias measures implemented by the business operator itself.

[2082] A "database" is a system that stores structured data and allows it to be efficiently searched and retrieved.

[2083] To implement this invention, the following specific steps are taken: First, a consortium will be established with the participation of institutions from various industries, providing a forum for sharing knowledge and technology related to bias countermeasures. This consortium will share the latest technical information and countermeasure methods, enabling participants to implement effective bias countermeasures.

[2084] Next, we build a feedback system to provide cultural feedback on the generated images, which involves the following steps:

[2085] 1. The user selects the AI-generated image for which they would like to receive feedback from their device.

[2086] 2. The device sends the selected image file to the server.

[2087] 3. The server stores the received images in a database and passes them to the image analysis system.

[2088] 4. The server launches an image analysis system to generate cultural feedback. Specifically, it uses image analysis libraries such as OpenCV and TensorFlow to analyze elements in the image (e.g., colors, symbols, background) and evaluate their cultural meaning.

[2089] 5. The server returns the analysis results and feedback to the user, for example, "The red background may have a specific religious meaning."

[2090] 6. The user checks the feedback sent back from the server on the device and makes any necessary corrections based on the feedback.

[2091] For the granting of the certification mark, the server will manage an anti-bias checklist for each company and make it available for download. Specifically, it will include the following steps:

[2092] 1. The server provides an anti-bias checklist for businesses to download.

[2093] 2. The user (company representative) downloads the checklist and evaluates their company's countermeasures.

[2094] 3. The device sends the evaluation results to the server.

[2095] 4. The server will analyze the self-evaluation results and the feedback system usage history data in order to determine whether to grant the certification mark.

[2096] 5. The server issues a certification mark based on the evaluation results and sends related information to the company.

[2097] 6. The user (certified company representative) downloads the certification mark and uses it on their company's website and marketing materials.

[2098] For the combination of emotion recognition systems, we will introduce an emotion engine to improve the quality of feedback and user experience. Specifically, we will use emotion analysis software such as Microsoft Azure Cognitive Services. The steps are as follows:

[2099] 1. The user chooses to use emotion recognition when receiving feedback.

[2100] 2. The device captures the user's facial expressions and voice and sends the emotional data to the server.

[2101] 3. The server uses an emotion engine to analyze the user's emotions. For example, it can recognize emotions such as "surprise" from the user's facial expressions and voice.

[2102] 4. The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[2103] 5. The server adaptively adjusts the feedback content based on the user's emotional changes. For example, if the user expresses negative emotions, the server softens the tone and content of the feedback.

[2104] This system configuration allows us to provide feedback that takes into account the cultural appropriateness of generated images and user emotions, and also enables objective evaluation and certification of businesses that implement anti-bias measures.

[2105] An example of a specific prompt is, "Please analyze the cultural meaning of the image below and provide feedback."

[2106] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2107] Specific examples of consortium establishment and operation

[2108] Handling user registration

[2109] Step 1:

[2110] The user accesses the new user registration form and enters information such as the company name, the name of the person in charge, and an email address.

[2111] Input: Company name, person in charge's name, email address, etc.

[2112] Output: A submission request containing the input information

[2113] Step 2:

[2114] The terminal sends a transmission request to the server.

[2115] Input: A submission request containing input information

[2116] Output: Data sent to the server

[2117] Step 3:

[2118] The server stores the received data in a database and sends the user a confirmation email informing them that their registration is complete.

[2119] Input: Send data

[2120] Data processing: Save input information to a database

[2121] Output: Confirmation email sent

[2122] Step 4:

[2123] The user clicks on the link in the confirmation email to complete the registration.

[2124] Input: Click the link in the confirmation email

[2125] Output: Registration completion notification

[2126] Step 5:

[2127] The server grants the user access to the consortium web portal.

[2128] Data processing: Update user registration status to "Complete"

[2129] Output: Notification of access rights granted

[2130] Holding workshops

[2131] Step 1:

[2132] Users log in to the consortium web portal and check the event calendar.

[2133] Input: User ID and password

[2134] Output: Event calendar view

[2135] Step 2:

[2136] The server displays detailed information about the workshop (title, date and time, instructor information).

[2137] Data processing: Obtaining workshop information from the event database

[2138] Output: Detailed information displayed on a web page

[2139] Step 3:

[2140] The user registers for the workshop of their choice.

[2141] Input: Enter details in the entry form and click the registration button

[2142] Output: Registration request sent

[2143] Step 4:

[2144] The server stores the registration information in a database and sends reminder emails before the workshop begins.

[2145] Data processing: Update participant list, send reminder emails

[2146] Output: Reminder email sent successfully

[2147] Specific examples of feedback system design and implementation

[2148] Image upload and analysis

[2149] Step 1:

[2150] Users select the AI-generated image they want to receive feedback on from their device.

[2151] Input: Select image file

[2152] Output: Selected image file

[2153] Step 2:

[2154] The terminal sends the selected image file to the server.

[2155] Input: Selected image file

[2156] Output: Request to send image file

[2157] Step 3:

[2158] The server stores the received images in a database and passes them to the image analysis system.

[2159] Input: Image file transfer request

[2160] Data processing: Save image data in a database

[2161] Output: Passing image data to an image analysis system

[2162] Step 4:

[2163] The server runs an image analysis system to generate cultural feedback, specifically using OpenCV and TensorFlow to analyze elements in the image and evaluate their cultural meaning.

[2164] Input: Image data

[2165] Data calculation: Analyze elements in the image (color, symbol, background)

[2166] Output: Generated feedback

[2167] Step 5:

[2168] The server sends the analysis results and feedback back to the user, such as "red backgrounds may have a specific religious meaning."

[2169] Input: Generated feedback

[2170] Data processing: Send feedback to users

[2171] Output: Feedback sent successfully

[2172] Step 6:

[2173] The user checks the feedback sent back from the server on the terminal and makes any necessary corrections based on the feedback.

[2174] Input: Check feedback

[2175] Output: Image correction completed

[2176] Examples of certification marking and management

[2177] Checklist Evaluation and Certification

[2178] Step 1:

[2179] The server will manage anti-bias checklists for each company and make them available for download.

[2180] Input: Checklist PDF file

[2181] Data processing: Upload the checklist to the web portal

[2182] Output: Publish download link

[2183] Step 2:

[2184] Users (company personnel) download the checklist and evaluate their company's countermeasures.

[2185] Input:Download checklist

[2186] Output: Checklist evaluation

[2187] Step 3:

[2188] The terminal transmits the self-evaluation result to the server.

[2189] Input: Self-assessment results

[2190] Output: Send request

[2191] Step 4:

[2192] The server integrates and analyzes the self-assessment results and the feedback system usage history data.

[2193] Input: Self-assessment results and usage history data

[2194] Data processing: Integrated analysis of evaluation data

[2195] Output: Evaluation results

[2196] Step 5:

[2197] The server decides whether to grant a certification mark based on the evaluation results.

[2198] Input: Evaluation result

[2199] Data processing: Judgment in accordance with certification standards

[2200] Output: Certification mark grant decision

[2201] Step 6:

[2202] The server issues a certification mark to the certified company and transmits related information.

[2203] Input: Certification mark grant decision

[2204] Data processing: generation of certification marks and certificates

[2205] Output: Certification mark sent successfully

[2206] Step 7:

[2207] Users (certified company representatives) can download the certification mark and use it on their company's website and marketing materials.

[2208] Input: Download certification mark

[2209] Output: Use of certification mark

[2210] Specific examples of combining emotion engines

[2211] Emotion recognition and feedback regulation

[2212] Step 1:

[2213] The user may choose to use emotion recognition functionality when receiving feedback.

[2214] Input: Select emotion recognition feature

[2215] Output: Emotion data collection begins

[2216] Step 2:

[2217] The device captures the user's facial expressions and voice and transmits the emotional data to the server.

[2218] Input: User's facial expressions and voice

[2219] Data processing: Real-time capture and data transmission

[2220] Output: Emotion data transmission

[2221] Step 3:

[2222] The server uses an emotion engine to analyze the user's emotions, for example, recognizing emotions such as "surprise" from the user's facial expressions and voice.

[2223] Input: Emotion data

[2224] Data Computing: Sentiment Analysis

[2225] Output: Emotional state

[2226] Step 4:

[2227] The server provides the emotion data to the feedback system and evaluates the effectiveness of the generated feedback.

[2228] Input: Emotional state

[2229] Data processing: Integrated analysis of emotion data and feedback data

[2230] Output: Feedback effect evaluation results

[2231] Step 5:

[2232] The server adaptively adjusts the feedback content based on changes in the user's emotions, for example, softening the tone and content of the feedback if the user expresses negative emotions.

[2233] Input: Feedback effect evaluation results

[2234] Data processing: Adjustment of feedback content

[2235] Output: Adaptively adjusted feedback

[2236] (Application example 2)

[2237] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2238] Advertising materials created by advertising creators often contain bias that ignores cultural background. Such biases can foster misunderstandings and prejudices against specific cultures or ethnicities, hindering the provision of fair advertising. Furthermore, it is difficult to properly capture the emotional responses of users viewing ads and provide feedback, leading to a lack of improvements to maximize the effectiveness of advertising. There is a need to address these issues.

[2239] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[2240] In this invention, the server includes means for sharing knowledge and technology related to bias countermeasures through a consortium of institutions from various industries, means for designing and implementing a feedback system that provides cultural feedback on generated images, means for granting a certification mark to businesses that implement appropriate bias countermeasures, means for providing cultural feedback on advertising materials created by advertising creators, and means for analyzing user emotions using an emotion recognition engine and adjusting the feedback content based on the results. This makes it possible to reduce cultural bias in advertising materials and provide appropriate feedback according to user emotions.

[2241] "Multi-industry institutions" is a general term for businesses and organizations in different industries, such as the advertising industry, software development industry, design industry, and marketing industry.

[2242] A "consortium" is a cooperative system formed by multiple companies or institutions working together to achieve a specific purpose or solve a problem.

[2243] "Anti-bias" refers to methods and efforts to eliminate prejudices and preconceptions based on particular cultures or social backgrounds.

[2244] A "feedback system" is a system that provides evaluations and opinions on generated images and advertising materials from a cultural perspective.

[2245] "Elements in an image" are components such as colors, symbols, backgrounds, and objects contained in an image.

[2246] "Cultural feedback" refers to evaluations and opinions about how generated images and advertising materials will be received based on a particular culture or social context.

[2247] A "certification mark" is an official certification symbol given to businesses or products that meet certain standards.

[2248] An "emotion recognition engine" is an algorithm or software that analyzes data such as a user's facial expressions and voice, and classifies and recognizes the user's emotions.

[2249] An "advertising creator" is a professional who creates advertising content and design.

[2250] "User sentiment" refers to the psychological and emotional response a user has when viewing an ad.

[2251] To implement this invention, the following specific system configuration and processes are required: Use the components of the server, terminal, and user, and clarify the roles of each.

[2252] Server Roles

[2253] The server is responsible for the main data processing and analysis functions. High-performance servers are recommended as the hardware to be used. Software such as Python, Flask, Keras, and OpenCV are also required.

[2254] 1. The server receives generated images and advertising materials from the user and stores the data in a database.

[2255] 2. The server passes the stored image data to an analysis algorithm to generate cultural feedback. The algorithm analyzes elements in the image (colors, symbols, background, etc.) and evaluates their cultural meaning.

[2256] 3. The server uses an emotion recognition engine to analyze the user's emotional data (facial expressions and voice) and recognize the user's emotional state. For example, it classifies and recognizes emotions such as "surprise," "happiness," and "sadness."

[2257] Device Role

[2258] A terminal is a device that provides a user interface, such as a smartphone or a head-mounted display.

[2259] 1. The terminal provides an interface for ad creators to upload their created ad materials to the server.

[2260] 2. The device displays the cultural feedback and emotion recognition results returned from the server to the user.

[2261] 3. The device provides tools for modifying advertising materials in real time, helping users improve their ads based on feedback.

[2262] User Roles

[2263] Users are creators who create advertisements and people involved in evaluating advertisements.

[2264] 1. The user uploads advertising material from their device to the server.

[2265] 2. The user checks the cultural feedback and emotion recognition results provided by the server and modifies the advertising material as needed.

[2266] 3. The user decides on the final advertising material based on the feedback and publishes it.

[2267] Specific examples of processing

[2268] For example, when an ad creator uploads a new ad image they've created, they might get feedback like this:

[2269] Cultural feedback: "The image contains a red background, which may have religious significance in parts of Asia. Consider changing the background color or being more considerate of the context."

[2270] Emotion-aware feedback: "We detected a surprised expression from the user. Consider adjusting the exciting elements in your ad."

[2271] This allows ad creators to easily identify specific areas for improvement and create more effective and culturally appropriate ads.

[2272] Through this invention, it is possible to effectively reduce cultural bias in advertising materials and provide appropriate feedback according to the user's emotions, thereby realizing fair and effective advertising creation.

[2273] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2274] Step 1:

[2275] The server receives advertising materials (images or videos) uploaded by users (advertising creators). Specifically, users select advertising materials using the device interface and send them to the server. This input data is stored on the server.

[2276] Step 2:

[2277] The server then passes the stored advertising material to an image analysis algorithm, which generates feedback that includes cultural meaning. Specifically, the algorithm analyzes elements in the image (color, symbols, background, etc.) and evaluates their cultural meaning. The resulting cultural feedback is then output.

[2278] Step 3:

[2279] The server sends emotional data (facial expressions and voice) of users viewing advertisements to the emotion recognition engine. When a user views an advertisement using a device, the device captures the user's facial expressions and voice and sends the data to the server. The emotion recognition engine analyzes this data and recognizes emotional states such as "surprise," "happiness," and "sadness." The analysis results in an emotion label for the user being output.

[2280] Step 4:

[2281] The server integrates the generated cultural feedback with the emotion recognition results and returns feedback to the user, including suggestions for correction and improvement. Specifically, the server combines the cultural feedback and emotion labels and returns them to the device in a format that is easy for the user to understand. For example, specific instructions may be output, such as, "Consider changing the background color because a red background may have a specific religious meaning. Additionally, since the user's surprised expression was detected, please adjust the provocative elements in the advertisement."

[2282] Step 5:

[2283] The user can use their device to check the feedback provided by the server and modify the advertising material as necessary. The user can then edit the advertising material based on the feedback and upload it back to the server. By repeating this process, the optimal advertising material is created.

[2284] Step 6:

[2285] The server will award a certification mark to businesses that have implemented appropriate anti-bias measures. Specifically, the server will integrate and analyze the evaluation results of the businesses' measures and their usage history of the feedback system, and will issue a certification mark if the standards are met. This certification mark will be provided to the certified businesses to display in their advertisements and on their websites.

[2286] Following this trend, advertising creators will be able to eliminate cultural bias and create optimal advertising materials that take user emotions into consideration.

[2287] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2288] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[2289] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2290] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2291] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2292] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2293] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2294] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2295] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2296] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2297] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2298] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2299] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2300] 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.

[2301] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[2302] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2303] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[2304] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[2305] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[2306] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[2307] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[2308] The following is further disclosed regarding the above embodiment.

[2309] (Claim 1)

[2310] A means of sharing knowledge and techniques on anti-bias measures through a multi-institutional consortium;

[2311] A means of designing and implementing a feedback system that provides cultural feedback on the generated images;

[2312] A means of granting a certification mark to businesses that implement appropriate anti-bias measures;

[2313] A system including:

[2314] (Claim 2)

[2315] The system of claim 1 , wherein the feedback system uses an algorithm that analyzes elements within an image and assesses cultural meaning.

[2316] (Claim 3)

[2317] 2. The system according to claim 1, further comprising means for integrating and analyzing the evaluation results of the measures taken by the business and the usage history of the feedback system when granting the certification mark.

[2318] "Example 1"

[2319] (Claim 1)

[2320] A means to host a new user registration form, accept information such as company name, contact name, and email address, store it in a database, and send a confirmation email;

[2321] A means for displaying detailed information about workshops, allowing users to register for the workshops they wish to attend, and sending reminder emails;

[2322] A means for selecting an AI-generated image for which feedback is desired from a terminal, analyzing elements within the image, generating cultural feedback using an algorithm that evaluates cultural meaning, and returning the analysis results to the user;

[2323] A means for users to modify images based on feedback and re-upload the images;

[2324] A means for downloading an anti-bias checklist, transmitting the self-assessment results to a server, integrating and analyzing the assessment results and the usage history dat...

Claims

1. A means of sharing knowledge and techniques on anti-bias measures through a multi-institutional consortium; A means of designing and implementing a feedback system that provides cultural feedback on the generated images; A means of granting a certification mark to businesses that implement appropriate anti-bias measures; A system including:

2. The system of claim 1 , wherein the feedback system uses an algorithm that analyzes elements within an image and assesses cultural meaning.

3. 2. The system according to claim 1, further comprising means for integrating and analyzing the results of the evaluation of the measures taken by the business and the usage history of the feedback system when granting the certification mark.

Images