System

Abstract

A system is provided.SOLUTION: A system comprising: means for receiving a particular data set from a user; means for analyzing the received data set and extracting characteristics; means for automatically setting production parameters based on the extracted characteristics; means for automatically creating a product using the set production parameters; means for evaluating the quality of the created product and adjusting the production parameters as needed; and means for providing the final product to the user.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] Conventional generative AI technologies have the drawback of requiring extensive manual adjustments to create high-quality products from specific datasets, resulting in unstable product quality. Furthermore, optimizing the generation process and adjusting generation parameters in real time is difficult, often resulting in ineffective generation. The present invention aims to solve these problems and automatically generate high-quality products efficiently. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by providing a system including the following means: means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This allows the user to obtain a high-quality product with minimal effort and ensures an efficient generation process.

[0006] A "dataset" refers to a collection of data that is the subject of a particular analysis or generation.

[0007] "User" means a user who utilizes the system to provide a dataset and receive the resulting data.

[0008] "Analysis" is the process of examining the information contained in a data set and extracting characteristics and patterns.

[0009] "Characteristics" refer to the individual characteristics or patterns within a dataset, and are the information obtained through analysis.

[0010] "Generation parameters" refer to the settings and conditions used by a generative AI model when creating a product based on a dataset.

[0011] "Automatically set" means that the system determines the optimal parameters by itself without user intervention.

[0012] "Product" refers to the final output created by a generative AI model, such as an image, text, or music.

[0013] "Quality evaluation" is the process of evaluating the quality of a product, determining whether the product meets a certain quality standard based on criteria.

[0014] "Adjustment" means modifying the production parameters according to the results of the product quality evaluation and reproducing the product.

[0015] "Providing" means making a product available to users. [Brief explanation of the drawings]

[0016] [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

[0017] 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.

[0018] First, the terms used in the following description will be explained.

[0019] 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).

[0020] 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.

[0021] 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.

[0022] 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.

[0023] 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."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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."

[0037] As an embodiment of the present invention, a system for automatically generating high-quality products from a specific data set will be described. This system operates in cooperation between a server, a terminal, and a user, and executes the necessary processing at each step.

[0038] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[0039] The server then analyzes the received dataset. It applies analysis algorithms to extract features within the dataset. For image datasets, this could be color, shape, or pattern. For text datasets, this could be keywords, themes, etc. These features are then used to optimize the generation parameters, as described below.

[0040] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0041] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[0042] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[0043] Finally, the server prepares the product for delivery to the user. The server stores the product and generates a download link for the user, who then obtains the high-quality product through the link. This system enables the efficient generation of high-quality products using a specific dataset, and each step is automated, minimizing user effort.

[0044] A concrete example is a process where a user uploads a text dataset of literary works, and a server generates new literary fragments based on that dataset. In this example, the server analyzes keywords and themes, selects optimal sentence structures and vocabulary, generates new text, evaluates and adjusts the quality, and ultimately delivers a high-quality output to the user. Similarly, a similar process can be applied to generate new music using a dataset of music tracks.

[0045] The processing flow will be explained below.

[0046] Step 1:

[0047] Upload Dataset (User)

[0048] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The device then sends the dataset to the server via an HTTP POST request.

[0049] Step 2:

[0050] Data reception and storage (server)

[0051] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[0052] Step 3:

[0053] Dataset analysis (server)

[0054] The server analyzes the stored dataset. In the case of an image dataset, it extracts features such as color, shape, and pattern. In the case of a text dataset, it extracts keywords, themes, and writing style. This is done using various analysis algorithms and libraries (e.g., OpenCV and NLTK).

[0055] Step 4:

[0056] Storage of characteristic data (server)

[0057] The server stores the extracted characteristic data in a database, which is used for subsequent optimization of the generation parameters.

[0058] Step 5:

[0059] Optimizing generation parameters (server)

[0060] The server calculates the optimal generation parameters based on the stored characteristic data. For example, in the case of image generation, it automatically sets the type of filter and parameters to be used, and in the case of text generation, it automatically sets the sentence structure and vocabulary selection. This is done using optimization algorithms and machine learning models.

[0061] Step 6:

[0062] Initialization of the generative AI model (server)

[0063] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[0064] Step 7:

[0065] Execution of the generation process (server)

[0066] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., preprocessing the data, inputting it into the model, and running the generation).

[0067] Step 8:

[0068] Real-time evaluation of artifacts (server)

[0069] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[0070] Step 9:

[0071] Parameter adjustment and regeneration (server)

[0072] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[0073] Step 10:

[0074] Storing and providing the generated data (server)

[0075] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[0076] Example 1

[0077] 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."

[0078] In today's information society, there is a growing demand for systems that allow users to efficiently create high-quality products based on their own datasets. However, conventional systems often require users to analyze large amounts of data, set optimal generation parameters, and evaluate and adjust the quality of the products, which is a manual process that places a heavy burden on users. Another issue is that the quality of the products cannot be maintained consistently at a high level.

[0079] 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.

[0080] In this invention, the server includes means for receiving a specific dataset from a user, means for analyzing the received dataset and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This allows users to easily obtain high-quality products automatically. Furthermore, the server includes means for applying a generation algorithm used to set the generation parameters, means for implementing an analysis algorithm for extracting specific generation parameters, and means for defining product evaluation criteria and making adjustments based on the quality criteria. This automates the analysis, generation, and evaluation processes, ensuring efficient and consistent product quality.

[0081] A "specific dataset" is a collection of data, regardless of type, such as images, text, or audio, that is input or provided by a user.

[0082] The "means for receiving data from a user" refers to the interface that a user uses to provide data and the function that a server uses to receive it.

[0083] "Analysis and characterization means" are algorithms or software used to understand the content of the data and extract important elements or features.

[0084] The "means for automatically setting generation parameters" is a function that automatically determines the setting values ​​for optimal operation of the generation algorithm based on the extracted characteristics.

[0085] The "means for automatically generating a product" is a function for generating new data or a product for a user based on generation parameters. The product is generated using an artificial intelligence model.

[0086] The "means for evaluating quality and adjusting generation parameters as necessary" is a function for evaluating the quality of the generated deliverables and resetting and regenerating them if they do not meet the standards.

[0087] The "means for providing the final product to the user" is a function for delivering the completed product to the user, and is usually an operation including saving the file and generating a download link.

[0088] A "generative algorithm" is a computational method or program used to create new products from data.

[0089] An "analysis algorithm" is a computational method or program used to analyze data and extract characteristics.

[0090] "Metrics" are standards or indicators established to measure the quality of a product.

[0091] The "means for adjusting based on quality standards" is a function for adjusting generation parameters so as to satisfy evaluation standards and performing regeneration.

[0092] The system of this invention is designed to automatically generate high-quality products from a specific dataset. The system operates in cooperation with a server, terminals, and users, and performs the necessary processing at each step.

[0093] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user might upload a folder of images or a text file with hundreds of pages. The hardware used here is the user's device, and the software is a web browser or a specialized upload application.

[0094] The server then analyzes the received dataset by running Python scripts or specific libraries (e.g., Pandas, OpenCV) to extract color tones, shapes, and patterns in the case of image datasets, and keywords and themes in the case of text datasets. Specifically, the server uses Python's PIL library to analyze the color distribution of the images, and an NLP library (e.g., NLTK) to extract keywords from the text.

[0095] Based on the analysis results, the server automatically sets generation parameters. For example, it generates instructions for the text generation model GPT-3 on what sentence structure and vocabulary to use. Based on the generated theme, the server generates prompt sentences such as "Explain in detail new trends in the market." For image generation, instructions include "Apply a filter that emphasizes blue tones."

[0096] Next, the server creates artifacts from the dataset using the configured generation parameters. The server initializes an appropriate generative AI model (e.g., DALL-E or GPT-3) and runs the model with specific generation parameters. For example, the server uses the DALL-E model to generate new images and the GPT-3 model to generate new sentences. Specifically, the server starts the generation process by calling the function "model.generate(prompt)" in Python code.

[0097] The quality of the generated content is evaluated in real time. The server uses a quality evaluation algorithm (e.g., a library that calculates the Mean Opinion Score or BLEU score) to quantify the quality of the generated content and regenerates it if necessary. Specifically, the server executes the function "quality_score = evaluate(generated_content)" against the algorithm.

[0098] Finally, the server prepares the product to be provided to the user. It saves the product and generates a download link that the user can access. The user can download the product via that link. For example, the server creates a URL for the generated file and sends the link to the user's email address. Specifically, the server executes the function "file_url = generate_download_link(generated_file)" and sends the link using "send_email(user_email, file_url)".

[0099] For example, a user can upload a text dataset of literary works and generate a fragment of a new literary work based on that dataset. The server analyzes keywords and themes, selects the most appropriate sentence structure and vocabulary, and generates the new text. An example of a prompt might be, "Write a poem in 15th-century English incorporating elements from Shakespeare's works."

[0100] The system allows users to obtain high-quality products with minimal effort, and utilizes specialized algorithms and techniques at each step to ensure efficient and consistent product quality.

[0101] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0102] Step 1:

[0103] The user selects a dataset and uploads it to the server. The user selects a specific dataset (e.g., a folder of image files, a text file with several hundred pages, etc.) on their own device. To do this, they operate the file selection dialog in their web browser and click the "Upload" button. The input is the dataset on the user's device, and the output is the dataset file sent to the server. Specifically, after selecting the file, "upload" is executed and an HTTP request is sent to the server.

[0104] Step 2:

[0105] The server analyzes the received dataset. When the server receives the dataset, it begins analysis using Python scripts and libraries such as Pandas and OpenCV. It receives the dataset file as input and outputs the analyzed characteristics (color tone, shape, and pattern for image datasets, and keywords and themes for text datasets). Specifically, the server uses OpenCV to analyze images and NLTK to extract keywords from text. Processes such as "image = cv2.imread('uploaded_image.jpg')" and "keywords = nltk.word_tokenize(text)" are executed.

[0106] Step 3:

[0107] The server automatically sets the generation parameters. Based on the analyzed features, the server determines the generation parameters. The input is the analyzed features, and the output is the generation parameters. Specifically, for image generation, this could be a filter or conversion method, and for text generation, it could be sentence structure or vocabulary selection. For example, "emphasize blue tones" or "describe market trends in detail" are generated. The function "parameters = generate_parameters(extracted_features)" is called.

[0108] Step 4:

[0109] The server automatically creates the product. It creates the product from the dataset based on the set generation parameters. The input is the generation parameters and the original dataset, and the output is the newly generated product. The server initializes an appropriate generative AI model (DALL-E, GPT-3, etc.) and executes the generation process. For example, code such as "generated_image = dalle.generate(image, parameters)" and "generated_text = gpt3.generate(prompt)" is executed.

[0110] Step 5:

[0111] The server evaluates the quality of the generated image. It applies a quality evaluation algorithm to evaluate the quality of the generated image. The input is the new image and the output is a quality score. For example, "quality_score = evaluate(generated_image)" or "text_quality_score = evaluate_text(generated_text)" are executed. If the evaluation criteria are not met, the server adjusts the generation parameters to regenerate the image. It calls the function "adjusted_parameters = adjust_parameters(quality_score)".

[0112] Step 6:

[0113] The server provides the product to the user. After the final product is completed, the server saves the product and generates a download link that the user can access. The input is the final product, and the output is the download link. Specifically, a URL is created for the generated file and the link is sent to the user's email address. Code such as "file_url = generate_download_link(generated_file)" and "send_email(user_email, file_url)" is executed. The user clicks the link to download the product.

[0114] (Application example 1)

[0115] 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."

[0116] In the automatic generation of advertising banners, manual design and quality control require time and effort, and there is no guarantee that high-quality results will be produced. In particular, the quality of the results varies depending on the data set uploaded by the user. Furthermore, unless the generation parameters are adjusted in real time, regeneration to improve quality may not be performed smoothly.

[0117] 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.

[0118] In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, means for providing the final product to the user, and means for providing the created advertising banner, thereby enabling the creation of high-quality advertising banners based on data provided by the user.

[0119] A "specific dataset" is data, such as images or text, provided by a user, that serves as the basis for the analysis and generation process used by the AI ​​model.

[0120] The "receiving means" refers to an interface or function for importing a data set provided by a user into a device such as a server.

[0121] "Means for analyzing and extracting characteristics" refers to algorithms or functions that analyze the received data set to find characteristics such as color tone, shape, keywords, etc.

[0122] The "means for automatically setting generation parameters" is a function that calculates and sets optimal parameters to be applied to the generation model based on the extracted characteristics.

[0123] "Means for automatically creating products" refers to a function that uses set generation parameters to create products such as images and text using an AI model.

[0124] The "means for evaluating the quality of the product" refers to an algorithm or function that evaluates the quality of the generated product in real time and quantifies the evaluation results.

[0125] The "means for adjusting the production parameters" is a function for resetting the production parameters and re-executing the production process based on the results of the product quality evaluation.

[0126] The "means for providing the final product to the user" is a function for storing the quality-guaranteed product in storage and generating a link that allows the user to download it.

[0127] The "means for providing generated advertising banners" is a function for displaying or saving high-quality advertising banners generated based on a data set provided by the user and providing them directly to the user.

[0128] The present invention provides a system for automatically generating high-quality advertising banners from a specific data set. This system operates in cooperation with a server, a terminal, and a user, and performs the necessary processing at each step.

[0129] First, a user selects a specific data set (e.g., product images and product description text) from their device and uploads it to the server. The data provided by the user becomes the raw material for creating advertising banners. The device communicates with the server via a high-speed Internet connection.

[0130] The server then analyzes the received dataset. The server applies an analysis algorithm to extract characteristics within the dataset. For image data, this includes color, shape, and patterns. For text data, this includes key keywords and themes. The analyzed characteristics are used to optimize the generation parameters, as described below.

[0131] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0132] The server then automatically creates an advertising banner using the set generation parameters. The server initializes an appropriate generative AI model and executes the generation process. For example, it generates high-quality banner ads based on product images and description text. Examples of AI models used include OpenAI's GPT-3 and DALL-E.

[0133] After the product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality advertising banners.

[0134] Finally, the server provides the generated advertisement banner to the user, stores the result, and generates a download link for the user, through which the user can obtain a high-quality advertisement banner.

[0135] For example, if a user uploads a product image and the description text "This is a great product that solves many issues," the server extracts the "keyword" and identifies the color scheme "vivid," and generates an advertising banner based on the following: "keywords: great product, issues, layout: banner, color scheme: vivid."

[0136] Example prompt sentence:

[0137] Create a banner with keywords: great product, issues, layout: banner, and color scheme: vivid.

[0138] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0139] Step 1:

[0140] The user selects a specific dataset (product images and product description text) from the terminal and uploads it to the server. The input data are image files and text files. The terminal sends these files to the server via a high-speed Internet connection. The output is the image and text data received on the server side.

[0141] Step 2:

[0142] The server analyzes the received dataset, applying analysis algorithms to extract color tones, shapes, patterns, etc. in the case of images, and key keywords and themes in the case of text. The input data are the received image and text data, and the output data are the extracted image features and text keywords.

[0143] Step 3:

[0144] The server automatically sets generation parameters based on the extracted characteristics. For example, it determines filters and conversion methods based on image characteristics, and determines the style and structure of the ad based on extracted keywords. The input is the characteristic data extracted in step 2, and the output is the optimal generation parameter settings.

[0145] Step 4:

[0146] The server automatically creates a product (advertising banner) using the set generation parameters. It initializes a generative AI model (e.g., OpenAI's GPT-3 or DALL-E) and runs the generation process. The inputs are the generation parameters, the original image, and text data, and the output is the created advertising banner.

[0147] Step 5:

[0148] The server evaluates the quality of the created product in real time. It applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. The input is the created advertising banner, and the output is the quality evaluation result. If necessary, the generation parameters are readjusted based on the evaluation result.

[0149] Step 6:

[0150] The server prepares the final product for the user, stores it and generates a link for the user to download it. The input is a high-quality advertising banner, and the output is a download link for the product available to the user.

[0151] In this way, a series of automated processes can be used to generate high-quality advertising banners based on user-provided data.

[0152] 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.

[0153] As an embodiment of the present invention, we will explain a configuration that combines a system that automatically generates high-quality products from a specific data set with an emotion engine that recognizes user emotions. This system operates in cooperation between a server, a terminal, and a user, and performs the necessary processing at each step.

[0154] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[0155] The server then analyzes the received dataset. The server applies analysis algorithms to extract characteristics within the dataset. For image datasets, this includes color, shape, and pattern. For text datasets, this includes keywords, themes, and style. These characteristics are then used to optimize the generation parameters, as described below.

[0156] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0157] Furthermore, the present invention incorporates an emotion engine that recognizes user emotions. When a user uploads a dataset, the emotion engine analyzes the user's facial expressions, voice, text input, etc. to collect emotion data. For example, it can recognize the user's emotions, such as happiness or sadness, in real time.

[0158] Based on the emotion engine's recognition, the server adjusts the generation parameters. For example, if the user expresses sadness, the server selects warmer colors and softer tones as the generation parameters. Therefore, the generated image will be more suitable for the user's emotion.

[0159] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[0160] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[0161] Finally, the server prepares the generated results for delivery to the user. The server stores the results and generates a download link for the user. The user can then obtain a high-quality result via the link. This system enables the efficient generation of high-quality results using a specific dataset and user emotion data, and each step is automated to minimize user effort.

[0162] For example, if the emotion engine determines that the user is happy, it will use that data to generate brightly colored images and positive text. In this way, the system can provide optimal results based on the user's emotions. The emotion engine can also analyze the user's emotions from voice input and adjust the results in real time. In this way, customization based on emotions is possible.

[0163] The processing flow will be explained below.

[0164] Step 1:

[0165] Upload Dataset (User)

[0166] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The user's device then sends the dataset to the server via an HTTP POST request.

[0167] Step 2:

[0168] Data reception and storage (server)

[0169] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[0170] Step 3:

[0171] Emotion data collection (terminal, server)

[0172] The device analyzes the user's emotions, such as facial expressions, voice, and text input, using an emotion engine to obtain emotion data in real time. For example, it analyzes facial expression data from a webcam and voice data from a microphone. The obtained emotion data is sent to a server.

[0173] Step 4:

[0174] Emotion data storage (server)

[0175] The server stores the emotion data received from the device in a database. The emotion data is associated with a session ID and a timestamp and is used in subsequent generation processes.

[0176] Step 5:

[0177] Dataset analysis (server)

[0178] The server analyzes the stored dataset. In the case of an image dataset, features such as color tone, shape, and pattern are extracted, and in the case of a text dataset, keywords, themes, and writing style are extracted. Various analysis algorithms and libraries (e.g., OpenCV and NLTK) are used for the analysis.

[0179] Step 6:

[0180] Storage of characteristic data (server)

[0181] The server stores the extracted characteristic data in the database again, and this characteristic data is used for subsequent optimization of the generation parameters.

[0182] Step 7:

[0183] Optimizing generation parameters (server)

[0184] The server calculates optimal generation parameters based on the stored characteristic data and emotion data. For example, if the user is "happy," it sets generation parameters including bright colors and positive content. This is done using optimization algorithms and machine learning models.

[0185] Step 8:

[0186] Initialization of the generative AI model (server)

[0187] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[0188] Step 9:

[0189] Execution of the generation process (server)

[0190] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., data preprocessing, model input, and execution of generation).

[0191] Step 10:

[0192] Real-time evaluation of artifacts (server)

[0193] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[0194] Step 11:

[0195] Parameter adjustment and regeneration (server)

[0196] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[0197] Step 12:

[0198] Storing and providing the generated data (server)

[0199] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[0200] Example 2

[0201] 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."

[0202] Conventional generation systems created products based solely on a dataset provided by the user, which meant they were unable to fully reflect the user's emotions and intentions. In particular, it was often difficult to precisely reflect the quality and atmosphere of the product desired by the user. Furthermore, they lacked a mechanism for evaluating the quality of the product in real time and adjusting the generation parameters as needed, making it difficult to provide high-quality products to users.

[0203] 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.

[0204] In this invention, the server includes means for processing a specific dataset received from a user, means for analyzing the dataset and extracting its characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for recognizing user emotion data in real time and adjusting the generation parameters based on the emotion data, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and readjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a high-quality product that is more in line with the user's intentions based on the user's emotions and the characteristics of the dataset.

[0205] A "dataset" is a series of data that a user uploads from their own terminal to a server, and includes image data or text data.

[0206] "Analysis" refers to the process undertaken to extract characteristics of an uploaded dataset, using image analysis algorithms and natural language processing algorithms.

[0207] "Generation parameters" are various setting values ​​that are set by the server to create an optimal product based on the extracted characteristics and emotion data.

[0208] The "emotion engine" is a system that analyzes the user's facial expressions, voice, and text input, and recognizes the user's emotional data in real time.

[0209] A "product" is an artifact such as an image or text that is automatically created using set generation parameters.

[0210] "Quality assessment" is the process of evaluating the quality of the generated product, and is performed quantitatively using a quality assessment algorithm.

[0211] "Providing" refers to a series of processes that the server performs to make the product accessible to the user, specifically including storing the product and generating a download link.

[0212] The present invention combines an emotion engine that recognizes user emotions in a system that automatically generates high-quality products from a specific dataset. This system operates in cooperation with a server, a terminal, and a user, and uses the following hardware and software.

[0213] Specifically, the server is a high-performance computer, and is ideally equipped with an environment capable of processing and analyzing large volumes of data. The server receives a specific data set and analyzes it using OpenCV as an image analysis algorithm and BERT as a natural language processing algorithm. This allows it to extract color, shape, and patterns from image data, and keywords, themes, and writing style from text data.

[0214] Next, the generation parameters are automatically set based on the extracted characteristics. Based on the analyzed data, the server calculates the optimal filters and transformation methods (in the case of image generation), as well as the optimal sentence structure and vocabulary selection (in the case of text generation). In this process, the latest generation AI models such as GPT-4 and GAN are often used.

[0215] When a user uploads a dataset from their device, the device's built-in emotion engine analyzes the user's facial expressions, voice, and text input. The emotion engine recognizes the user's emotional state in real time and provides this information to the server. For example, it outputs the user's emotions, such as whether they are happy or sad, as the analysis result. Based on this emotion data, the server adjusts the generation parameters to create a product that matches the user's emotions.

[0216] Once the generation process is complete, the server evaluates the quality of the initial product. It uses a quality evaluation algorithm to quantify the quality of the product. If the evaluation result does not meet the standards, the server adjusts the generation parameters again and repeats the generation. This process ensures high-quality products.

[0217] Finally, the server stores the product and generates a download link that can be accessed by the user, through which the user can download the product.

[0218] As a specific example of operation, if a user uploads a dataset containing bright landscape images and the emotion engine determines that the user is happy, the server will generate landscape images with bright tones. Conversely, if a user requests quiet and calm landscape images and the emotion engine determines that the user is sad, the server will generate landscape images with warm tones.

[0219] An example prompt is:

[0220] "Based on an image dataset uploaded by a user, generate natural landscape images that match his / her emotions. For example, use bright colors if the user is happy, and warm colors if the user is sad."

[0221] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0222] Step 1:

[0223] A user selects a specific dataset from their device and uploads it to the server. For example, when a user selects an image dataset or a text dataset and clicks the upload button, the device sends the dataset to the server. The input is the dataset selected by the user, and the output is the data uploaded to the server.

[0224] Step 2:

[0225] The server receives datasets uploaded by users. The input is the dataset sent by the user, and the output is the dataset stored in the server. Specifically, the server validates the dataset to ensure that the received data has been stored correctly.

[0226] Step 3:

[0227] The server analyzes the received dataset. The input is the dataset stored on the server, and the output is the characteristics extracted from the dataset. Specifically, for image datasets, the server uses OpenCV to analyze color, shape, and patterns, and for text datasets, it uses BERT to extract keywords, themes, and writing style.

[0228] Step 4:

[0229] The server automatically sets generation parameters based on the extracted characteristics. The input is the analysis result, and the output is the set generation parameters. Specifically, the server determines the filter and conversion method for image generation, and the sentence structure and vocabulary selection for text generation.

[0230] Step 5:

[0231] When a user uploads a dataset, the device's emotion engine analyzes the user's facial expressions, voice, text input, etc. The input is the user's emotional data, and the output is the emotional state recognized in real time. Specifically, the device uses emotion recognition software to analyze the user's emotions and sends emotions such as "happy" or "sad" to the server.

[0232] Step 6:

[0233] The server adjusts the generation parameters based on the emotion data recognized by the emotion engine. The input is the emotion data recognized in real time, and the output is the adjusted generation parameters. Specifically, the server adjusts the color tone and tone according to the user's emotion.

[0234] Step 7:

[0235] The server runs a generation process using the set generation parameters to create products. The input is the adjusted generation parameters and dataset, and the output is the generated product. Specifically, the server uses a generative AI model such as GPT-4 or GAN to create new images and text.

[0236] Step 8:

[0237] The server evaluates the quality of the generated product in real time. The input is the generated product, and the output is the evaluation result. Specifically, the server uses a quality evaluation algorithm to quantify the quality of the product and determine whether it meets the criteria.

[0238] Step 9:

[0239] Based on the evaluation results, the server readjusts the generation parameters as necessary and repeats the generation process. The inputs are the quality evaluation results and the current generation parameters, and the outputs are the readjusted generation parameters and the regenerated product.

[0240] Step 10:

[0241] The server finally stores the product and generates a link for users to download it. The input is the final product, and the output is the download link. Users download the high-quality product through this link. Specifically, the server stores the product in cloud storage and notifies the user of the link.

[0242] (Application example 2)

[0243] 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."

[0244] Conventional systems for creating products do not take into account the user's emotions, making it difficult to provide content that matches the user's needs and emotions. In particular, food delivery services require customized menu suggestions based on the user's current emotions, making it necessary to provide a more personalized experience.

[0245] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for recognizing the user's emotions in real time, means for adjusting the generation parameters based on the recognized user's emotions, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a product that is in line with the user's emotions.

[0246] (Definitions of important words)

[0247] A "specific dataset" is a collection of information provided by a user, and is a series of data that may be in the form of images, text, or the like.

[0248] A "receiving means" is a device or software module for obtaining a particular data set from a user.

[0249] "Analysis Means" means any equipment or software containing algorithms or programs for analyzing and examining received data sets and extracting their characteristics.

[0250] "Characteristics" are characteristics such as color, shape, pattern, keywords, themes, and writing style contained within a dataset.

[0251] "Generation parameters" refer to the specific conditions and settings used when creating a product.

[0252] The "means for setting generation parameters" refers to an algorithm or program that determines generation parameters based on characteristics extracted by analysis.

[0253] "Product" refers to the final output, including images, text, and other formats, that is automatically created based on a particular dataset and generation parameters.

[0254] "Creation means" refers to an algorithm or program for creating a product based on generation parameters, and hardware for executing the algorithm or program.

[0255] The "emotion recognition means" is a device that includes an engine or algorithm for analyzing the user's facial expressions, voice, text, etc., and recognizing the user's emotions in real time.

[0256] "Emotion-based generation parameter adjustment means" refers to an algorithm or program for dynamically changing generation parameters in response to recognized user emotion information.

[0257] A "quality evaluation means" is a device that includes an algorithm or program for evaluating the quality of the created product and readjusting the production parameters based on quantitative criteria.

[0258] "Providing means" refers to an interface or module for delivering the final product to the user.

[0259] As an embodiment of the present invention, a system for optimizing menu suggestions for a food delivery service based on user emotions will be described. This system operates by integrating a server, a user device, an emotion engine, an analysis algorithm, a generative AI model, and a quality evaluation algorithm.

[0260] 1. Overall system configuration

[0261] The system uses the following hardware and software:

[0262] Hardware:

[0263] Smartphone: Accepts data input from the user.

[0264] Camera and microphone: Devices for capturing the user's facial expressions and voice.

[0265] Cloud server: Analyzes datasets, recognizes emotions, runs generative AI models, and evaluates the quality of the products.

[0266] software:

[0267] Emotion AI: Analyzes a user's facial expressions, tone of voice, and text input to collect emotional data in real time. For example, common cloud services include Google Cloud's Face API and Amazon Rekognition.

[0268] Analysis algorithms: These include algorithms that extract characteristics of the dataset and set generation parameters, such as color, shape, keywords, themes, and style.

[0269] Generative AI models (e.g., GPT-4 or Transformer-based models): Generate appropriate menu suggestions based on user sentiment data and food menus.

[0270] Quality assessment algorithm: An algorithm for assessing the quality of the proposed menu and readjusting the generation parameters if necessary.

[0271] 2. Data processing flow

[0272] The server processes the data in the following specific steps:

[0273] 1. Receiving the dataset:

[0274] The user uploads a dataset via a smartphone app, which simultaneously captures the user's facial expressions and voice and sends them to a cloud server.

[0275] 2. Dataset Analysis:

[0276] The cloud server analyzes the received data set and extracts relevant characteristics, such as past order history, current facial expression, and tone of voice, to analyze the user's emotions in real time.

[0277] 3. Set generation parameters:

[0278] Based on the extracted characteristics and the recognized emotion data, the generation parameters are automatically set. For example, if the user is tired, the generation parameters are set to suggest a menu that will help them relax.

[0279] 4. Creating the product:

[0280] Using the set generation parameters, the generative AI model automatically creates appropriate menu suggestions.

[0281] 5. Product quality assessment:

[0282] The quality of the generated menu suggestions is evaluated, and if necessary, the generation parameters are adjusted and regenerated.

[0283] 6. Provision of Products:

[0284] The final menu proposal is then provided to the user, who can then view the proposed menu and place an order via a smartphone app.

[0285] 3. Examples of concrete examples and prompts

[0286] As a concrete example, the following prompt sentence is input into the generative AI model:

[0287] If the emotion engine analyzes a photo of a user smiling and determines that the user is "happy":

[0288] "It's a wonderful day today! I'd like to suggest a fresh and delicious seafood spaghetti dish that's perfect for a day like this."

[0289] If the emotion engine determines that the user is "tired" based on facial expressions and voice that indicate fatigue:

[0290] "Thank you for your hard work. How about a relaxing custom meatloaf dinner? It'll warm your soul and your body."

[0291] This makes it possible to provide customized services that are in tune with the user's emotions.

[0292] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0293] Program processing steps

[0294] Step 1:

[0295] The user launches the smartphone app, captures facial expression and voice data via the camera and microphone, selects the data set for food delivery, and sends it to the cloud server via the app.

[0296] Input: facial expression data, audio data, dataset

[0297] Output: Data sent to the cloud server

[0298] Step 2:

[0299] The server analyzes the received dataset and extracts relevant characteristics, applying analysis algorithms to extract color tones and shapes in the case of images, and keywords and themes in the case of text.

[0300] Input: Dataset

[0301] Output: Extracted features

[0302] Step 3:

[0303] The server uses facial expression and voice data sent from the camera and microphone to enable the emotion engine to recognize the user's emotions in real time, using Google Cloud's Face API and Amazon Rekognition to identify emotions such as joy, surprise, and fatigue.

[0304] Input: facial expression data, voice data

[0305] Output: Emotion data

[0306] Step 4:

[0307] The server sets generation parameters based on the characteristics and emotion data, and determines appropriate filters and conditions for content generation, taking into account past order history and user emotion data.

[0308] Input: extracted features, emotion data

[0309] Output: Generated parameters

[0310] Step 5:

[0311] The server uses the configured generation parameters to generate food menu suggestions using a generative AI model (e.g., GPT-4 or a Transformer-based model).

[0312] Input: Generation parameters

[0313] Output: The food menu suggestions created

[0314] Step 6:

[0315] The server evaluates the quality of the generated food menu suggestions, applies a quality evaluation algorithm to measure the quality of the output, and, if necessary, readjusts the generation parameters and runs the generation process again.

[0316] Input: Created food menu proposal

[0317] Output: Evaluation results, regenerated proposals (if necessary)

[0318] Step 7:

[0319] The server then provides the final food menu suggestions to the user, who can then review the suggestions, make a selection, and place an order through a smartphone app. Optionally, a personalized message tailored to the user's mood can also be sent.

[0320] Input: Final generated food menu proposals

[0321] Output: Food menu suggestions provided to the user, customized messages tailored to their emotions

[0322] Specific operation example

[0323] Step 1:

[0324] The user opens the smartphone app, selects the "Food Delivery" menu, and takes a photo of their face with the camera. Audio data is recorded and sent to the cloud, showing how depressed they are.

[0325] Step 2:

[0326] The server analyzes the received image data and picks out keywords from color, shape, or voice data. For example, if the user says "I'm tired," this will be extracted as an analysis result.

[0327] Step 3:

[0328] The server uses Google Cloud's Face API to recognize the user's facial expression, such as "tired," and also analyzes emotions from the tone of their voice.

[0329] Step 4:

[0330] The server sets generation parameters for suggesting relaxing dishes based on the characteristics and the emotional data of "tired."

[0331] Step 5:

[0332] The server uses a generative AI model to create food menu suggestions that fit the "tired" emotion, such as "Custom Meatloaf Dinner."

[0333] Step 6:

[0334] The server applies a quality assessment algorithm to evaluate the quality of the proposed menu, readjusting the generation parameters based on the results and regenerating the menu if necessary.

[0335] Step 7:

[0336] The server presents the user with the final menu suggestion along with a message saying, "Great work! How about a relaxing custom meatloaf dinner?" The user completes the order through the app.

[0337] 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.

[0338] 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.

[0339] 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.

[0340] [Second embodiment]

[0341] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0342] 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.

[0343] 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).

[0344] 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.

[0345] 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.

[0346] 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).

[0347] 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.

[0348] 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.

[0349] 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.

[0350] 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.

[0351] 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.

[0352] 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."

[0353] As an embodiment of the present invention, a system for automatically generating high-quality products from a specific data set will be described. This system operates in cooperation between a server, a terminal, and a user, and executes the necessary processing at each step.

[0354] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[0355] The server then analyzes the received dataset. It applies analysis algorithms to extract features within the dataset. For image datasets, this could be color, shape, or pattern. For text datasets, this could be keywords, themes, etc. These features are then used to optimize the generation parameters, as described below.

[0356] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0357] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[0358] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[0359] Finally, the server prepares the product for delivery to the user. The server stores the product and generates a download link for the user, who then obtains the high-quality product through the link. This system enables the efficient generation of high-quality products using a specific dataset, and each step is automated, minimizing user effort.

[0360] A concrete example is a process where a user uploads a text dataset of literary works, and a server generates new literary fragments based on that dataset. In this example, the server analyzes keywords and themes, selects optimal sentence structures and vocabulary, generates new text, evaluates and adjusts the quality, and ultimately delivers a high-quality output to the user. Similarly, a similar process can be applied to generate new music using a dataset of music tracks.

[0361] The processing flow will be explained below.

[0362] Step 1:

[0363] Upload Dataset (User)

[0364] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The device then sends the dataset to the server via an HTTP POST request.

[0365] Step 2:

[0366] Data reception and storage (server)

[0367] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[0368] Step 3:

[0369] Dataset analysis (server)

[0370] The server analyzes the stored dataset. In the case of an image dataset, it extracts features such as color, shape, and pattern. In the case of a text dataset, it extracts keywords, themes, and writing style. This is done using various analysis algorithms and libraries (e.g., OpenCV and NLTK).

[0371] Step 4:

[0372] Storage of characteristic data (server)

[0373] The server stores the extracted characteristic data in a database, which is used for subsequent optimization of the generation parameters.

[0374] Step 5:

[0375] Optimizing generation parameters (server)

[0376] The server calculates the optimal generation parameters based on the stored characteristic data. For example, in the case of image generation, it automatically sets the type of filter and parameters to be used, and in the case of text generation, it automatically sets the sentence structure and vocabulary selection. This is done using optimization algorithms and machine learning models.

[0377] Step 6:

[0378] Initialization of the generative AI model (server)

[0379] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[0380] Step 7:

[0381] Execution of the generation process (server)

[0382] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., preprocessing the data, inputting it into the model, and running the generation).

[0383] Step 8:

[0384] Real-time evaluation of artifacts (server)

[0385] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[0386] Step 9:

[0387] Parameter adjustment and regeneration (server)

[0388] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[0389] Step 10:

[0390] Storing and providing the generated data (server)

[0391] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[0392] Example 1

[0393] 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."

[0394] In today's information society, there is a growing demand for systems that allow users to efficiently create high-quality products based on their own datasets. However, conventional systems often require users to analyze large amounts of data, set optimal generation parameters, and evaluate and adjust the quality of the products, which is a manual process that places a heavy burden on users. Another issue is that the quality of the products cannot be maintained consistently at a high level.

[0395] 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.

[0396] In this invention, the server includes means for receiving a specific dataset from a user, means for analyzing the received dataset and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This allows users to easily obtain high-quality products automatically. Furthermore, the server includes means for applying a generation algorithm used to set the generation parameters, means for implementing an analysis algorithm for extracting specific generation parameters, and means for defining product evaluation criteria and making adjustments based on the quality criteria. This automates the analysis, generation, and evaluation processes, ensuring efficient and consistent product quality.

[0397] A "specific dataset" is a collection of data, regardless of type, such as images, text, or audio, that is input or provided by a user.

[0398] The "means for receiving data from a user" refers to the interface that a user uses to provide data and the function that a server uses to receive it.

[0399] "Analysis and characterization means" are algorithms or software used to understand the content of the data and extract important elements or features.

[0400] The "means for automatically setting generation parameters" is a function that automatically determines the setting values ​​for optimal operation of the generation algorithm based on the extracted characteristics.

[0401] The "means for automatically generating a product" is a function for generating new data or a product for a user based on generation parameters. The product is generated using an artificial intelligence model.

[0402] The "means for evaluating quality and adjusting generation parameters as necessary" is a function for evaluating the quality of the generated deliverables and resetting and regenerating them if they do not meet the standards.

[0403] The "means for providing the final product to the user" is a function for delivering the completed product to the user, and is usually an operation including saving the file and generating a download link.

[0404] A "generative algorithm" is a computational method or program used to create new products from data.

[0405] An "analysis algorithm" is a computational method or program used to analyze data and extract characteristics.

[0406] "Metrics" are standards or indicators established to measure the quality of a product.

[0407] The "means for adjusting based on quality standards" is a function for adjusting generation parameters so as to satisfy evaluation standards and performing regeneration.

[0408] The system of this invention is designed to automatically generate high-quality products from a specific dataset. The system operates in cooperation with a server, terminals, and users, and performs the necessary processing at each step.

[0409] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user might upload a folder of images or a text file with hundreds of pages. The hardware used here is the user's device, and the software is a web browser or a specialized upload application.

[0410] The server then analyzes the received dataset by running Python scripts or specific libraries (e.g., Pandas, OpenCV) to extract color tones, shapes, and patterns in the case of image datasets, and keywords and themes in the case of text datasets. Specifically, the server uses Python's PIL library to analyze the color distribution of the images, and an NLP library (e.g., NLTK) to extract keywords from the text.

[0411] Based on the analysis results, the server automatically sets generation parameters. For example, it generates instructions for the text generation model GPT-3 on what sentence structure and vocabulary to use. Based on the generated theme, the server generates prompt sentences such as "Explain in detail new trends in the market." For image generation, instructions include "Apply a filter that emphasizes blue tones."

[0412] Next, the server creates artifacts from the dataset using the configured generation parameters. The server initializes an appropriate generative AI model (e.g., DALL-E or GPT-3) and runs the model with specific generation parameters. For example, the server uses the DALL-E model to generate new images and the GPT-3 model to generate new sentences. Specifically, the server starts the generation process by calling the function "model.generate(prompt)" in Python code.

[0413] The quality of the generated content is evaluated in real time. The server uses a quality evaluation algorithm (e.g., a library that calculates the Mean Opinion Score or BLEU score) to quantify the quality of the generated content and regenerates it if necessary. Specifically, the server executes the function "quality_score = evaluate(generated_content)" against the algorithm.

[0414] Finally, the server prepares the product to be provided to the user. It saves the product and generates a download link that the user can access. The user can download the product via that link. For example, the server creates a URL for the generated file and sends the link to the user's email address. Specifically, the server executes the function "file_url = generate_download_link(generated_file)" and sends the link using "send_email(user_email, file_url)".

[0415] For example, a user can upload a text dataset of literary works and generate a fragment of a new literary work based on that dataset. The server analyzes keywords and themes, selects the most appropriate sentence structure and vocabulary, and generates the new text. An example of a prompt might be, "Write a poem in 15th-century English incorporating elements from Shakespeare's works."

[0416] The system allows users to obtain high-quality products with minimal effort, and utilizes specialized algorithms and techniques at each step to ensure efficient and consistent product quality.

[0417] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0418] Step 1:

[0419] The user selects a dataset and uploads it to the server. The user selects a specific dataset (e.g., a folder of image files, a text file with several hundred pages, etc.) on their own device. To do this, they operate the file selection dialog in their web browser and click the "Upload" button. The input is the dataset on the user's device, and the output is the dataset file sent to the server. Specifically, after selecting the file, "upload" is executed and an HTTP request is sent to the server.

[0420] Step 2:

[0421] The server analyzes the received dataset. When the server receives the dataset, it begins analysis using Python scripts and libraries such as Pandas and OpenCV. It receives the dataset file as input and outputs the analyzed characteristics (color tone, shape, and pattern for image datasets, and keywords and themes for text datasets). Specifically, the server uses OpenCV to analyze images and NLTK to extract keywords from text. Processes such as "image = cv2.imread('uploaded_image.jpg')" and "keywords = nltk.word_tokenize(text)" are executed.

[0422] Step 3:

[0423] The server automatically sets the generation parameters. Based on the analyzed features, the server determines the generation parameters. The input is the analyzed features, and the output is the generation parameters. Specifically, for image generation, this could be a filter or conversion method, and for text generation, it could be sentence structure or vocabulary selection. For example, "emphasize blue tones" or "describe market trends in detail" are generated. The function "parameters = generate_parameters(extracted_features)" is called.

[0424] Step 4:

[0425] The server automatically creates the product. It creates the product from the dataset based on the set generation parameters. The input is the generation parameters and the original dataset, and the output is the newly generated product. The server initializes an appropriate generative AI model (DALL-E, GPT-3, etc.) and executes the generation process. For example, code such as "generated_image = dalle.generate(image, parameters)" and "generated_text = gpt3.generate(prompt)" is executed.

[0426] Step 5:

[0427] The server evaluates the quality of the generated image. It applies a quality evaluation algorithm to evaluate the quality of the generated image. The input is the new image and the output is a quality score. For example, "quality_score = evaluate(generated_image)" or "text_quality_score = evaluate_text(generated_text)" are executed. If the evaluation criteria are not met, the server adjusts the generation parameters to regenerate the image. It calls the function "adjusted_parameters = adjust_parameters(quality_score)".

[0428] Step 6:

[0429] The server provides the product to the user. After the final product is completed, the server saves the product and generates a download link that the user can access. The input is the final product, and the output is the download link. Specifically, a URL is created for the generated file and the link is sent to the user's email address. Code such as "file_url = generate_download_link(generated_file)" and "send_email(user_email, file_url)" is executed. The user clicks the link to download the product.

[0430] (Application example 1)

[0431] 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."

[0432] In the automatic generation of advertising banners, manual design and quality control require time and effort, and there is no guarantee that high-quality results will be produced. In particular, the quality of the results varies depending on the data set uploaded by the user. Furthermore, unless the generation parameters are adjusted in real time, regeneration to improve quality may not be performed smoothly.

[0433] 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.

[0434] In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, means for providing the final product to the user, and means for providing the created advertising banner, thereby enabling the creation of high-quality advertising banners based on data provided by the user.

[0435] A "specific dataset" is data, such as images or text, provided by a user, that serves as the basis for the analysis and generation process used by the AI ​​model.

[0436] The "receiving means" refers to an interface or function for importing a data set provided by a user into a device such as a server.

[0437] "Means for analyzing and extracting characteristics" refers to algorithms or functions that analyze the received data set to find characteristics such as color tone, shape, keywords, etc.

[0438] The "means for automatically setting generation parameters" is a function that calculates and sets optimal parameters to be applied to the generation model based on the extracted characteristics.

[0439] "Means for automatically creating products" refers to a function that uses set generation parameters to create products such as images and text using an AI model.

[0440] The "means for evaluating the quality of the product" refers to an algorithm or function that evaluates the quality of the generated product in real time and quantifies the evaluation results.

[0441] The "means for adjusting the production parameters" is a function for resetting the production parameters and re-executing the production process based on the results of the product quality evaluation.

[0442] The "means for providing the final product to the user" is a function for storing the quality-guaranteed product in storage and generating a link that allows the user to download it.

[0443] The "means for providing generated advertising banners" is a function for displaying or saving high-quality advertising banners generated based on a data set provided by the user and providing them directly to the user.

[0444] The present invention provides a system for automatically generating high-quality advertising banners from a specific data set. This system operates in cooperation with a server, a terminal, and a user, and performs the necessary processing at each step.

[0445] First, a user selects a specific data set (e.g., product images and product description text) from their device and uploads it to the server. The data provided by the user becomes the raw material for creating advertising banners. The device communicates with the server via a high-speed Internet connection.

[0446] The server then analyzes the received dataset. The server applies an analysis algorithm to extract characteristics within the dataset. For image data, this includes color, shape, and patterns. For text data, this includes key keywords and themes. The analyzed characteristics are used to optimize the generation parameters, as described below.

[0447] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0448] The server then automatically creates an advertising banner using the set generation parameters. The server initializes an appropriate generative AI model and executes the generation process. For example, it generates high-quality banner ads based on product images and description text. Examples of AI models used include OpenAI's GPT-3 and DALL-E.

[0449] After the product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality advertising banners.

[0450] Finally, the server provides the generated advertisement banner to the user, stores the result, and generates a download link for the user, through which the user can obtain a high-quality advertisement banner.

[0451] For example, if a user uploads a product image and the description text "This is a great product that solves many issues," the server extracts the "keyword" and identifies the color scheme "vivid," and generates an advertising banner based on the following: "keywords: great product, issues, layout: banner, color scheme: vivid."

[0452] Example prompt sentence:

[0453] Create a banner with keywords: great product, issues, layout: banner, and color scheme: vivid.

[0454] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0455] Step 1:

[0456] The user selects a specific dataset (product images and product description text) from the terminal and uploads it to the server. The input data are image files and text files. The terminal sends these files to the server via a high-speed Internet connection. The output is the image and text data received on the server side.

[0457] Step 2:

[0458] The server analyzes the received dataset, applying analysis algorithms to extract color tones, shapes, patterns, etc. in the case of images, and key keywords and themes in the case of text. The input data are the received image and text data, and the output data are the extracted image features and text keywords.

[0459] Step 3:

[0460] The server automatically sets generation parameters based on the extracted characteristics. For example, it determines filters and conversion methods based on image characteristics, and determines the style and structure of the ad based on extracted keywords. The input is the characteristic data extracted in step 2, and the output is the optimal generation parameter settings.

[0461] Step 4:

[0462] The server automatically creates a product (advertising banner) using the set generation parameters. It initializes a generative AI model (e.g., OpenAI's GPT-3 or DALL-E) and runs the generation process. The inputs are the generation parameters, the original image, and text data, and the output is the created advertising banner.

[0463] Step 5:

[0464] The server evaluates the quality of the created product in real time. It applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. The input is the created advertising banner, and the output is the quality evaluation result. If necessary, the generation parameters are readjusted based on the evaluation result.

[0465] Step 6:

[0466] The server prepares the final product for the user, stores it and generates a link for the user to download it. The input is a high-quality advertising banner, and the output is a download link for the product available to the user.

[0467] In this way, a series of automated processes can be used to generate high-quality advertising banners based on user-provided data.

[0468] 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.

[0469] As an embodiment of the present invention, we will explain a configuration that combines a system that automatically generates high-quality products from a specific data set with an emotion engine that recognizes user emotions. This system operates in cooperation between a server, a terminal, and a user, and performs the necessary processing at each step.

[0470] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[0471] The server then analyzes the received dataset. The server applies analysis algorithms to extract characteristics within the dataset. For image datasets, this includes color, shape, and pattern. For text datasets, this includes keywords, themes, and style. These characteristics are then used to optimize the generation parameters, as described below.

[0472] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0473] Furthermore, the present invention incorporates an emotion engine that recognizes user emotions. When a user uploads a dataset, the emotion engine analyzes the user's facial expressions, voice, text input, etc. to collect emotion data. For example, it can recognize the user's emotions, such as happiness or sadness, in real time.

[0474] Based on the emotion engine's recognition, the server adjusts the generation parameters. For example, if the user expresses sadness, the server selects warmer colors and softer tones as the generation parameters. Therefore, the generated image will be more suitable for the user's emotion.

[0475] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[0476] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[0477] Finally, the server prepares the generated results for delivery to the user. The server stores the results and generates a download link for the user. The user can then obtain a high-quality result via the link. This system enables the efficient generation of high-quality results using a specific dataset and user emotion data, and each step is automated to minimize user effort.

[0478] For example, if the emotion engine determines that the user is happy, it will use that data to generate brightly colored images and positive text. In this way, the system can provide optimal results based on the user's emotions. The emotion engine can also analyze the user's emotions from voice input and adjust the results in real time. In this way, customization based on emotions is possible.

[0479] The processing flow will be explained below.

[0480] Step 1:

[0481] Upload Dataset (User)

[0482] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The user's device then sends the dataset to the server via an HTTP POST request.

[0483] Step 2:

[0484] Data reception and storage (server)

[0485] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[0486] Step 3:

[0487] Emotion data collection (terminal, server)

[0488] The device analyzes the user's emotions, such as facial expressions, voice, and text input, using an emotion engine to obtain emotion data in real time. For example, it analyzes facial expression data from a webcam and voice data from a microphone. The obtained emotion data is sent to a server.

[0489] Step 4:

[0490] Emotion data storage (server)

[0491] The server stores the emotion data received from the device in a database. The emotion data is associated with a session ID and a timestamp and is used in subsequent generation processes.

[0492] Step 5:

[0493] Dataset analysis (server)

[0494] The server analyzes the stored dataset. In the case of an image dataset, features such as color tone, shape, and pattern are extracted, and in the case of a text dataset, keywords, themes, and writing style are extracted. Various analysis algorithms and libraries (e.g., OpenCV and NLTK) are used for the analysis.

[0495] Step 6:

[0496] Storage of characteristic data (server)

[0497] The server stores the extracted characteristic data in the database again, and this characteristic data is used for subsequent optimization of the generation parameters.

[0498] Step 7:

[0499] Optimizing generation parameters (server)

[0500] The server calculates optimal generation parameters based on the stored characteristic data and emotion data. For example, if the user is "happy," it sets generation parameters including bright colors and positive content. This is done using optimization algorithms and machine learning models.

[0501] Step 8:

[0502] Initialization of the generative AI model (server)

[0503] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[0504] Step 9:

[0505] Execution of the generation process (server)

[0506] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., data preprocessing, model input, and execution of generation).

[0507] Step 10:

[0508] Real-time evaluation of artifacts (server)

[0509] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[0510] Step 11:

[0511] Parameter adjustment and regeneration (server)

[0512] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[0513] Step 12:

[0514] Storing and providing the generated data (server)

[0515] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[0516] Example 2

[0517] 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."

[0518] Conventional generation systems created products based solely on a dataset provided by the user, which meant they were unable to fully reflect the user's emotions and intentions. In particular, it was often difficult to precisely reflect the quality and atmosphere of the product desired by the user. Furthermore, they lacked a mechanism for evaluating the quality of the product in real time and adjusting the generation parameters as needed, making it difficult to provide high-quality products to users.

[0519] 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.

[0520] In this invention, the server includes means for processing a specific dataset received from a user, means for analyzing the dataset and extracting its characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for recognizing user emotion data in real time and adjusting the generation parameters based on the emotion data, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and readjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a high-quality product that is more in line with the user's intentions based on the user's emotions and the characteristics of the dataset.

[0521] A "dataset" is a series of data that a user uploads from their own terminal to a server, and includes image data or text data.

[0522] "Analysis" refers to the process undertaken to extract characteristics of an uploaded dataset, using image analysis algorithms and natural language processing algorithms.

[0523] "Generation parameters" are various setting values ​​that are set by the server to create an optimal product based on the extracted characteristics and emotion data.

[0524] The "emotion engine" is a system that analyzes the user's facial expressions, voice, and text input, and recognizes the user's emotional data in real time.

[0525] A "product" is an artifact such as an image or text that is automatically created using set generation parameters.

[0526] "Quality assessment" is the process of evaluating the quality of the generated product, and is performed quantitatively using a quality assessment algorithm.

[0527] "Providing" refers to a series of processes that the server performs to make the product accessible to the user, specifically including storing the product and generating a download link.

[0528] The present invention combines an emotion engine that recognizes user emotions in a system that automatically generates high-quality products from a specific dataset. This system operates in cooperation with a server, a terminal, and a user, and uses the following hardware and software.

[0529] Specifically, the server is a high-performance computer, and is ideally equipped with an environment capable of processing and analyzing large volumes of data. The server receives a specific data set and analyzes it using OpenCV as an image analysis algorithm and BERT as a natural language processing algorithm. This allows it to extract color, shape, and patterns from image data, and keywords, themes, and writing style from text data.

[0530] Next, the generation parameters are automatically set based on the extracted characteristics. Based on the analyzed data, the server calculates the optimal filters and transformation methods (in the case of image generation), as well as the optimal sentence structure and vocabulary selection (in the case of text generation). In this process, the latest generation AI models such as GPT-4 and GAN are often used.

[0531] When a user uploads a dataset from their device, the device's built-in emotion engine analyzes the user's facial expressions, voice, and text input. The emotion engine recognizes the user's emotional state in real time and provides this information to the server. For example, it outputs the user's emotions, such as whether they are happy or sad, as the analysis result. Based on this emotion data, the server adjusts the generation parameters to create a product that matches the user's emotions.

[0532] Once the generation process is complete, the server evaluates the quality of the initial product. It uses a quality evaluation algorithm to quantify the quality of the product. If the evaluation result does not meet the standards, the server adjusts the generation parameters again and repeats the generation. This process ensures high-quality products.

[0533] Finally, the server stores the product and generates a download link that can be accessed by the user, through which the user can download the product.

[0534] As a specific example of operation, if a user uploads a dataset containing bright landscape images and the emotion engine determines that the user is happy, the server will generate landscape images with bright tones. Conversely, if a user requests quiet and calm landscape images and the emotion engine determines that the user is sad, the server will generate landscape images with warm tones.

[0535] An example prompt is:

[0536] "Based on an image dataset uploaded by a user, generate natural landscape images that match his / her emotions. For example, use bright colors if the user is happy, and warm colors if the user is sad."

[0537] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0538] Step 1:

[0539] A user selects a specific dataset from their device and uploads it to the server. For example, when a user selects an image dataset or a text dataset and clicks the upload button, the device sends the dataset to the server. The input is the dataset selected by the user, and the output is the data uploaded to the server.

[0540] Step 2:

[0541] The server receives datasets uploaded by users. The input is the dataset sent by the user, and the output is the dataset stored in the server. Specifically, the server validates the dataset to ensure that the received data has been stored correctly.

[0542] Step 3:

[0543] The server analyzes the received dataset. The input is the dataset stored on the server, and the output is the characteristics extracted from the dataset. Specifically, for image datasets, the server uses OpenCV to analyze color, shape, and patterns, and for text datasets, it uses BERT to extract keywords, themes, and writing style.

[0544] Step 4:

[0545] The server automatically sets generation parameters based on the extracted characteristics. The input is the analysis result, and the output is the set generation parameters. Specifically, the server determines the filter and conversion method for image generation, and the sentence structure and vocabulary selection for text generation.

[0546] Step 5:

[0547] When a user uploads a dataset, the device's emotion engine analyzes the user's facial expressions, voice, text input, etc. The input is the user's emotional data, and the output is the emotional state recognized in real time. Specifically, the device uses emotion recognition software to analyze the user's emotions and sends emotions such as "happy" or "sad" to the server.

[0548] Step 6:

[0549] The server adjusts the generation parameters based on the emotion data recognized by the emotion engine. The input is the emotion data recognized in real time, and the output is the adjusted generation parameters. Specifically, the server adjusts the color tone and tone according to the user's emotion.

[0550] Step 7:

[0551] The server runs a generation process using the set generation parameters to create products. The input is the adjusted generation parameters and dataset, and the output is the generated product. Specifically, the server uses a generative AI model such as GPT-4 or GAN to create new images and text.

[0552] Step 8:

[0553] The server evaluates the quality of the generated product in real time. The input is the generated product, and the output is the evaluation result. Specifically, the server uses a quality evaluation algorithm to quantify the quality of the product and determine whether it meets the criteria.

[0554] Step 9:

[0555] Based on the evaluation results, the server readjusts the generation parameters as necessary and repeats the generation process. The inputs are the quality evaluation results and the current generation parameters, and the outputs are the readjusted generation parameters and the regenerated product.

[0556] Step 10:

[0557] The server finally stores the product and generates a link for users to download it. The input is the final product, and the output is the download link. Users download the high-quality product through this link. Specifically, the server stores the product in cloud storage and notifies the user of the link.

[0558] (Application example 2)

[0559] 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."

[0560] Conventional systems for creating products do not take into account the user's emotions, making it difficult to provide content that matches the user's needs and emotions. In particular, food delivery services require customized menu suggestions based on the user's current emotions, making it necessary to provide a more personalized experience.

[0561] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for recognizing the user's emotions in real time, means for adjusting the generation parameters based on the recognized user's emotions, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a product that is in line with the user's emotions.

[0562] (Definitions of important words)

[0563] A "specific dataset" is a collection of information provided by a user, and is a series of data that may be in the form of images, text, or the like.

[0564] A "receiving means" is a device or software module for obtaining a particular data set from a user.

[0565] "Analysis Means" means any equipment or software containing algorithms or programs for analyzing and examining received data sets and extracting their characteristics.

[0566] "Characteristics" are characteristics such as color, shape, pattern, keywords, themes, and writing style contained within a dataset.

[0567] "Generation parameters" refer to the specific conditions and settings used when creating a product.

[0568] The "means for setting generation parameters" refers to an algorithm or program that determines generation parameters based on characteristics extracted by analysis.

[0569] "Product" refers to the final output, including images, text, and other formats, that is automatically created based on a particular dataset and generation parameters.

[0570] "Creation means" refers to an algorithm or program for creating a product based on generation parameters, and hardware for executing the algorithm or program.

[0571] The "emotion recognition means" is a device that includes an engine or algorithm for analyzing the user's facial expressions, voice, text, etc., and recognizing the user's emotions in real time.

[0572] "Emotion-based generation parameter adjustment means" refers to an algorithm or program for dynamically changing generation parameters in response to recognized user emotion information.

[0573] A "quality evaluation means" is a device that includes an algorithm or program for evaluating the quality of the created product and readjusting the production parameters based on quantitative criteria.

[0574] "Providing means" refers to an interface or module for delivering the final product to the user.

[0575] As an embodiment of the present invention, a system for optimizing menu suggestions for a food delivery service based on user emotions will be described. This system operates by integrating a server, a user device, an emotion engine, an analysis algorithm, a generative AI model, and a quality evaluation algorithm.

[0576] 1. Overall system configuration

[0577] The system uses the following hardware and software:

[0578] Hardware:

[0579] Smartphone: Accepts data input from the user.

[0580] Camera and microphone: Devices for capturing the user's facial expressions and voice.

[0581] Cloud server: Analyzes datasets, recognizes emotions, runs generative AI models, and evaluates the quality of the products.

[0582] software:

[0583] Emotion AI: Analyzes a user's facial expressions, tone of voice, and text input to collect emotional data in real time. For example, common cloud services include Google Cloud's Face API and Amazon Rekognition.

[0584] Analysis algorithms: These include algorithms that extract characteristics of the dataset and set generation parameters, such as color, shape, keywords, themes, and style.

[0585] Generative AI models (e.g., GPT-4 or Transformer-based models): Generate appropriate menu suggestions based on user sentiment data and food menus.

[0586] Quality assessment algorithm: An algorithm for assessing the quality of the proposed menu and readjusting the generation parameters if necessary.

[0587] 2. Data processing flow

[0588] The server processes the data in the following specific steps:

[0589] 1. Receiving the dataset:

[0590] The user uploads a dataset via a smartphone app, which simultaneously captures the user's facial expressions and voice and sends them to a cloud server.

[0591] 2. Dataset Analysis:

[0592] The cloud server analyzes the received data set and extracts relevant characteristics, such as past order history, current facial expression, and tone of voice, to analyze the user's emotions in real time.

[0593] 3. Set generation parameters:

[0594] Based on the extracted characteristics and the recognized emotion data, the generation parameters are automatically set. For example, if the user is tired, the generation parameters are set to suggest a menu that will help them relax.

[0595] 4. Creating the product:

[0596] Using the set generation parameters, the generative AI model automatically creates appropriate menu suggestions.

[0597] 5. Product quality assessment:

[0598] The quality of the generated menu suggestions is evaluated, and if necessary, the generation parameters are adjusted and regenerated.

[0599] 6. Provision of Products:

[0600] The final menu proposal is then provided to the user, who can then view the proposed menu and place an order via a smartphone app.

[0601] 3. Examples of concrete examples and prompts

[0602] As a concrete example, the following prompt sentence is input into the generative AI model:

[0603] If the emotion engine analyzes a photo of a user smiling and determines that the user is "happy":

[0604] "It's a wonderful day today! I'd like to suggest a fresh and delicious seafood spaghetti dish that's perfect for a day like this."

[0605] If the emotion engine determines that the user is "tired" based on facial expressions and voice that indicate fatigue:

[0606] "Thank you for your hard work. How about a relaxing custom meatloaf dinner? It'll warm your soul and your body."

[0607] This makes it possible to provide customized services that are in tune with the user's emotions.

[0608] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0609] Program processing steps

[0610] Step 1:

[0611] The user launches the smartphone app, captures facial expression and voice data via the camera and microphone, selects the data set for food delivery, and sends it to the cloud server via the app.

[0612] Input: facial expression data, audio data, dataset

[0613] Output: Data sent to the cloud server

[0614] Step 2:

[0615] The server analyzes the received dataset and extracts relevant characteristics, applying analysis algorithms to extract color tones and shapes in the case of images, and keywords and themes in the case of text.

[0616] Input: Dataset

[0617] Output: Extracted features

[0618] Step 3:

[0619] The server uses facial expression and voice data sent from the camera and microphone to enable the emotion engine to recognize the user's emotions in real time, using Google Cloud's Face API and Amazon Rekognition to identify emotions such as joy, surprise, and fatigue.

[0620] Input: facial expression data, voice data

[0621] Output: Emotion data

[0622] Step 4:

[0623] The server sets generation parameters based on the characteristics and emotion data, and determines appropriate filters and conditions for content generation, taking into account past order history and user emotion data.

[0624] Input: extracted features, emotion data

[0625] Output: Generated parameters

[0626] Step 5:

[0627] The server uses the configured generation parameters to generate food menu suggestions using a generative AI model (e.g., GPT-4 or a Transformer-based model).

[0628] Input: Generation parameters

[0629] Output: The food menu suggestions created

[0630] Step 6:

[0631] The server evaluates the quality of the generated food menu suggestions, applies a quality evaluation algorithm to measure the quality of the output, and, if necessary, readjusts the generation parameters and runs the generation process again.

[0632] Input: Created food menu proposal

[0633] Output: Evaluation results, regenerated proposals (if necessary)

[0634] Step 7:

[0635] The server then provides the final food menu suggestions to the user, who can then review the suggestions, make a selection, and place an order through a smartphone app. Optionally, a personalized message tailored to the user's mood can also be sent.

[0636] Input: Final generated food menu proposals

[0637] Output: Food menu suggestions provided to the user, customized messages tailored to their emotions

[0638] Specific operation example

[0639] Step 1:

[0640] The user opens the smartphone app, selects the "Food Delivery" menu, and takes a photo of their face with the camera. Audio data is recorded and sent to the cloud, showing how depressed they are.

[0641] Step 2:

[0642] The server analyzes the received image data and picks out keywords from color, shape, or voice data. For example, if the user says "I'm tired," this will be extracted as an analysis result.

[0643] Step 3:

[0644] The server uses Google Cloud's Face API to recognize the user's facial expression, such as "tired," and also analyzes emotions from the tone of their voice.

[0645] Step 4:

[0646] The server sets generation parameters for suggesting relaxing dishes based on the characteristics and the emotional data of "tired."

[0647] Step 5:

[0648] The server uses a generative AI model to create food menu suggestions that fit the "tired" emotion, such as "Custom Meatloaf Dinner."

[0649] Step 6:

[0650] The server applies a quality assessment algorithm to evaluate the quality of the proposed menu, readjusting the generation parameters based on the results and regenerating the menu if necessary.

[0651] Step 7:

[0652] The server presents the user with the final menu suggestion along with a message saying, "Great work! How about a relaxing custom meatloaf dinner?" The user completes the order through the app.

[0653] 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.

[0654] 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.

[0655] 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.

[0656] [Third embodiment]

[0657] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0658] 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.

[0659] 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).

[0660] 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.

[0661] 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.

[0662] 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).

[0663] 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.

[0664] 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.

[0665] 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.

[0666] 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.

[0667] 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.

[0668] 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."

[0669] As an embodiment of the present invention, a system for automatically generating high-quality products from a specific data set will be described. This system operates in cooperation between a server, a terminal, and a user, and executes the necessary processing at each step.

[0670] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[0671] The server then analyzes the received dataset. It applies analysis algorithms to extract features within the dataset. For image datasets, this could be color, shape, or pattern. For text datasets, this could be keywords, themes, etc. These features are then used to optimize the generation parameters, as described below.

[0672] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0673] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[0674] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[0675] Finally, the server prepares the product for delivery to the user. The server stores the product and generates a download link for the user, who then obtains the high-quality product through the link. This system enables the efficient generation of high-quality products using a specific dataset, and each step is automated, minimizing user effort.

[0676] A concrete example is a process where a user uploads a text dataset of literary works, and a server generates new literary fragments based on that dataset. In this example, the server analyzes keywords and themes, selects optimal sentence structures and vocabulary, generates new text, evaluates and adjusts the quality, and ultimately delivers a high-quality output to the user. Similarly, a similar process can be applied to generate new music using a dataset of music tracks.

[0677] The processing flow will be explained below.

[0678] Step 1:

[0679] Upload Dataset (User)

[0680] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The device then sends the dataset to the server via an HTTP POST request.

[0681] Step 2:

[0682] Data reception and storage (server)

[0683] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[0684] Step 3:

[0685] Dataset analysis (server)

[0686] The server analyzes the stored dataset. In the case of an image dataset, it extracts features such as color, shape, and pattern. In the case of a text dataset, it extracts keywords, themes, and writing style. This is done using various analysis algorithms and libraries (e.g., OpenCV and NLTK).

[0687] Step 4:

[0688] Storage of characteristic data (server)

[0689] The server stores the extracted characteristic data in a database, which is used for subsequent optimization of the generation parameters.

[0690] Step 5:

[0691] Optimizing generation parameters (server)

[0692] The server calculates the optimal generation parameters based on the stored characteristic data. For example, in the case of image generation, it automatically sets the type of filter and parameters to be used, and in the case of text generation, it automatically sets the sentence structure and vocabulary selection. This is done using optimization algorithms and machine learning models.

[0693] Step 6:

[0694] Initialization of the generative AI model (server)

[0695] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[0696] Step 7:

[0697] Execution of the generation process (server)

[0698] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., preprocessing the data, inputting it into the model, and running the generation).

[0699] Step 8:

[0700] Real-time evaluation of artifacts (server)

[0701] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[0702] Step 9:

[0703] Parameter adjustment and regeneration (server)

[0704] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[0705] Step 10:

[0706] Storing and providing the generated data (server)

[0707] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[0708] Example 1

[0709] 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."

[0710] In today's information society, there is a growing demand for systems that allow users to efficiently create high-quality products based on their own datasets. However, conventional systems often require users to analyze large amounts of data, set optimal generation parameters, and evaluate and adjust the quality of the products, which is a manual process that places a heavy burden on users. Another issue is that the quality of the products cannot be maintained consistently at a high level.

[0711] 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.

[0712] In this invention, the server includes means for receiving a specific dataset from a user, means for analyzing the received dataset and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This allows users to easily obtain high-quality products automatically. Furthermore, the server includes means for applying a generation algorithm used to set the generation parameters, means for implementing an analysis algorithm for extracting specific generation parameters, and means for defining product evaluation criteria and making adjustments based on the quality criteria. This automates the analysis, generation, and evaluation processes, ensuring efficient and consistent product quality.

[0713] A "specific dataset" is a collection of data, regardless of type, such as images, text, or audio, that is input or provided by a user.

[0714] The "means for receiving data from a user" refers to the interface that a user uses to provide data and the function that a server uses to receive it.

[0715] "Analysis and characterization means" are algorithms or software used to understand the content of the data and extract important elements or features.

[0716] The "means for automatically setting generation parameters" is a function that automatically determines the setting values ​​for optimal operation of the generation algorithm based on the extracted characteristics.

[0717] The "means for automatically generating a product" is a function for generating new data or a product for a user based on generation parameters. The product is generated using an artificial intelligence model.

[0718] The "means for evaluating quality and adjusting generation parameters as necessary" is a function for evaluating the quality of the generated deliverables and resetting and regenerating them if they do not meet the standards.

[0719] The "means for providing the final product to the user" is a function for delivering the completed product to the user, and is usually an operation including saving the file and generating a download link.

[0720] A "generative algorithm" is a computational method or program used to create new products from data.

[0721] An "analysis algorithm" is a computational method or program used to analyze data and extract characteristics.

[0722] "Metrics" are standards or indicators established to measure the quality of a product.

[0723] The "means for adjusting based on quality standards" is a function for adjusting generation parameters so as to satisfy evaluation standards and performing regeneration.

[0724] The system of this invention is designed to automatically generate high-quality products from a specific dataset. The system operates in cooperation with a server, terminals, and users, and performs the necessary processing at each step.

[0725] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user might upload a folder of images or a text file with hundreds of pages. The hardware used here is the user's device, and the software is a web browser or a specialized upload application.

[0726] The server then analyzes the received dataset by running Python scripts or specific libraries (e.g., Pandas, OpenCV) to extract color tones, shapes, and patterns in the case of image datasets, and keywords and themes in the case of text datasets. Specifically, the server uses Python's PIL library to analyze the color distribution of the images, and an NLP library (e.g., NLTK) to extract keywords from the text.

[0727] Based on the analysis results, the server automatically sets generation parameters. For example, it generates instructions for the text generation model GPT-3 on what sentence structure and vocabulary to use. Based on the generated theme, the server generates prompt sentences such as "Explain in detail new trends in the market." For image generation, instructions include "Apply a filter that emphasizes blue tones."

[0728] Next, the server creates artifacts from the dataset using the configured generation parameters. The server initializes an appropriate generative AI model (e.g., DALL-E or GPT-3) and runs the model with specific generation parameters. For example, the server uses the DALL-E model to generate new images and the GPT-3 model to generate new sentences. Specifically, the server starts the generation process by calling the function "model.generate(prompt)" in Python code.

[0729] The quality of the generated content is evaluated in real time. The server uses a quality evaluation algorithm (e.g., a library that calculates the Mean Opinion Score or BLEU score) to quantify the quality of the generated content and regenerates it if necessary. Specifically, the server executes the function "quality_score = evaluate(generated_content)" against the algorithm.

[0730] Finally, the server prepares the product to be provided to the user. It saves the product and generates a download link that the user can access. The user can download the product via that link. For example, the server creates a URL for the generated file and sends the link to the user's email address. Specifically, the server executes the function "file_url = generate_download_link(generated_file)" and sends the link using "send_email(user_email, file_url)".

[0731] For example, a user can upload a text dataset of literary works and generate a fragment of a new literary work based on that dataset. The server analyzes keywords and themes, selects the most appropriate sentence structure and vocabulary, and generates the new text. An example of a prompt might be, "Write a poem in 15th-century English incorporating elements from Shakespeare's works."

[0732] The system allows users to obtain high-quality products with minimal effort, and utilizes specialized algorithms and techniques at each step to ensure efficient and consistent product quality.

[0733] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0734] Step 1:

[0735] The user selects a dataset and uploads it to the server. The user selects a specific dataset (e.g., a folder of image files, a text file with several hundred pages, etc.) on their own device. To do this, they operate the file selection dialog in their web browser and click the "Upload" button. The input is the dataset on the user's device, and the output is the dataset file sent to the server. Specifically, after selecting the file, "upload" is executed and an HTTP request is sent to the server.

[0736] Step 2:

[0737] The server analyzes the received dataset. When the server receives the dataset, it begins analysis using Python scripts and libraries such as Pandas and OpenCV. It receives the dataset file as input and outputs the analyzed characteristics (color tone, shape, and pattern for image datasets, and keywords and themes for text datasets). Specifically, the server uses OpenCV to analyze images and NLTK to extract keywords from text. Processes such as "image = cv2.imread('uploaded_image.jpg')" and "keywords = nltk.word_tokenize(text)" are executed.

[0738] Step 3:

[0739] The server automatically sets the generation parameters. Based on the analyzed features, the server determines the generation parameters. The input is the analyzed features, and the output is the generation parameters. Specifically, for image generation, this could be a filter or conversion method, and for text generation, it could be sentence structure or vocabulary selection. For example, "emphasize blue tones" or "describe market trends in detail" are generated. The function "parameters = generate_parameters(extracted_features)" is called.

[0740] Step 4:

[0741] The server automatically creates the product. It creates the product from the dataset based on the set generation parameters. The input is the generation parameters and the original dataset, and the output is the newly generated product. The server initializes an appropriate generative AI model (DALL-E, GPT-3, etc.) and executes the generation process. For example, code such as "generated_image = dalle.generate(image, parameters)" and "generated_text = gpt3.generate(prompt)" is executed.

[0742] Step 5:

[0743] The server evaluates the quality of the generated image. It applies a quality evaluation algorithm to evaluate the quality of the generated image. The input is the new image and the output is a quality score. For example, "quality_score = evaluate(generated_image)" or "text_quality_score = evaluate_text(generated_text)" are executed. If the evaluation criteria are not met, the server adjusts the generation parameters to regenerate the image. It calls the function "adjusted_parameters = adjust_parameters(quality_score)".

[0744] Step 6:

[0745] The server provides the product to the user. After the final product is completed, the server saves the product and generates a download link that the user can access. The input is the final product, and the output is the download link. Specifically, a URL is created for the generated file and the link is sent to the user's email address. Code such as "file_url = generate_download_link(generated_file)" and "send_email(user_email, file_url)" is executed. The user clicks the link to download the product.

[0746] (Application example 1)

[0747] 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."

[0748] In the automatic generation of advertising banners, manual design and quality control require time and effort, and there is no guarantee that high-quality results will be produced. In particular, the quality of the results varies depending on the data set uploaded by the user. Furthermore, unless the generation parameters are adjusted in real time, regeneration to improve quality may not be performed smoothly.

[0749] 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.

[0750] In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, means for providing the final product to the user, and means for providing the created advertising banner, thereby enabling the creation of high-quality advertising banners based on data provided by the user.

[0751] A "specific dataset" is data, such as images or text, provided by a user, that serves as the basis for the analysis and generation process used by the AI ​​model.

[0752] The "receiving means" refers to an interface or function for importing a data set provided by a user into a device such as a server.

[0753] "Means for analyzing and extracting characteristics" refers to algorithms or functions that analyze the received data set to find characteristics such as color tone, shape, keywords, etc.

[0754] The "means for automatically setting generation parameters" is a function that calculates and sets optimal parameters to be applied to the generation model based on the extracted characteristics.

[0755] "Means for automatically creating products" refers to a function that uses set generation parameters to create products such as images and text using an AI model.

[0756] The "means for evaluating the quality of the product" refers to an algorithm or function that evaluates the quality of the generated product in real time and quantifies the evaluation results.

[0757] The "means for adjusting the production parameters" is a function for resetting the production parameters and re-executing the production process based on the results of the product quality evaluation.

[0758] The "means for providing the final product to the user" is a function for storing the quality-guaranteed product in storage and generating a link that allows the user to download it.

[0759] The "means for providing generated advertising banners" is a function for displaying or saving high-quality advertising banners generated based on a data set provided by the user and providing them directly to the user.

[0760] The present invention provides a system for automatically generating high-quality advertising banners from a specific data set. This system operates in cooperation with a server, a terminal, and a user, and performs the necessary processing at each step.

[0761] First, a user selects a specific data set (e.g., product images and product description text) from their device and uploads it to the server. The data provided by the user becomes the raw material for creating advertising banners. The device communicates with the server via a high-speed Internet connection.

[0762] The server then analyzes the received dataset. The server applies an analysis algorithm to extract characteristics within the dataset. For image data, this includes color, shape, and patterns. For text data, this includes key keywords and themes. The analyzed characteristics are used to optimize the generation parameters, as described below.

[0763] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0764] The server then automatically creates an advertising banner using the set generation parameters. The server initializes an appropriate generative AI model and executes the generation process. For example, it generates high-quality banner ads based on product images and description text. Examples of AI models used include OpenAI's GPT-3 and DALL-E.

[0765] After the product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality advertising banners.

[0766] Finally, the server provides the generated advertisement banner to the user, stores the result, and generates a download link for the user, through which the user can obtain a high-quality advertisement banner.

[0767] For example, if a user uploads a product image and the description text "This is a great product that solves many issues," the server extracts the "keyword" and identifies the color scheme "vivid," and generates an advertising banner based on the following: "keywords: great product, issues, layout: banner, color scheme: vivid."

[0768] Example prompt sentence:

[0769] Create a banner with keywords: great product, issues, layout: banner, and color scheme: vivid.

[0770] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0771] Step 1:

[0772] The user selects a specific dataset (product images and product description text) from the terminal and uploads it to the server. The input data are image files and text files. The terminal sends these files to the server via a high-speed Internet connection. The output is the image and text data received on the server side.

[0773] Step 2:

[0774] The server analyzes the received dataset, applying analysis algorithms to extract color tones, shapes, patterns, etc. in the case of images, and key keywords and themes in the case of text. The input data are the received image and text data, and the output data are the extracted image features and text keywords.

[0775] Step 3:

[0776] The server automatically sets generation parameters based on the extracted characteristics. For example, it determines filters and conversion methods based on image characteristics, and determines the style and structure of the ad based on extracted keywords. The input is the characteristic data extracted in step 2, and the output is the optimal generation parameter settings.

[0777] Step 4:

[0778] The server automatically creates a product (advertising banner) using the set generation parameters. It initializes a generative AI model (e.g., OpenAI's GPT-3 or DALL-E) and runs the generation process. The inputs are the generation parameters, the original image, and text data, and the output is the created advertising banner.

[0779] Step 5:

[0780] The server evaluates the quality of the created product in real time. It applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. The input is the created advertising banner, and the output is the quality evaluation result. If necessary, the generation parameters are readjusted based on the evaluation result.

[0781] Step 6:

[0782] The server prepares the final product for the user, stores it and generates a link for the user to download it. The input is a high-quality advertising banner, and the output is a download link for the product available to the user.

[0783] In this way, a series of automated processes can be used to generate high-quality advertising banners based on user-provided data.

[0784] 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.

[0785] As an embodiment of the present invention, we will explain a configuration that combines a system that automatically generates high-quality products from a specific data set with an emotion engine that recognizes user emotions. This system operates in cooperation between a server, a terminal, and a user, and performs the necessary processing at each step.

[0786] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[0787] The server then analyzes the received dataset. The server applies analysis algorithms to extract characteristics within the dataset. For image datasets, this includes color, shape, and pattern. For text datasets, this includes keywords, themes, and style. These characteristics are then used to optimize the generation parameters, as described below.

[0788] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0789] Furthermore, the present invention incorporates an emotion engine that recognizes user emotions. When a user uploads a dataset, the emotion engine analyzes the user's facial expressions, voice, text input, etc. to collect emotion data. For example, it can recognize the user's emotions, such as happiness or sadness, in real time.

[0790] Based on the emotion engine's recognition, the server adjusts the generation parameters. For example, if the user expresses sadness, the server selects warmer colors and softer tones as the generation parameters. Therefore, the generated image will be more suitable for the user's emotion.

[0791] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[0792] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[0793] Finally, the server prepares the generated results for delivery to the user. The server stores the results and generates a download link for the user. The user can then obtain a high-quality result via the link. This system enables the efficient generation of high-quality results using a specific dataset and user emotion data, and each step is automated to minimize user effort.

[0794] For example, if the emotion engine determines that the user is happy, it will use that data to generate brightly colored images and positive text. In this way, the system can provide optimal results based on the user's emotions. The emotion engine can also analyze the user's emotions from voice input and adjust the results in real time. In this way, customization based on emotions is possible.

[0795] The processing flow will be explained below.

[0796] Step 1:

[0797] Upload Dataset (User)

[0798] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The user's device then sends the dataset to the server via an HTTP POST request.

[0799] Step 2:

[0800] Data reception and storage (server)

[0801] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[0802] Step 3:

[0803] Emotion data collection (terminal, server)

[0804] The device analyzes the user's emotions, such as facial expressions, voice, and text input, using an emotion engine to obtain emotion data in real time. For example, it analyzes facial expression data from a webcam and voice data from a microphone. The obtained emotion data is sent to a server.

[0805] Step 4:

[0806] Emotion data storage (server)

[0807] The server stores the emotion data received from the device in a database. The emotion data is associated with a session ID and a timestamp and is used in subsequent generation processes.

[0808] Step 5:

[0809] Dataset analysis (server)

[0810] The server analyzes the stored dataset. In the case of an image dataset, features such as color tone, shape, and pattern are extracted, and in the case of a text dataset, keywords, themes, and writing style are extracted. Various analysis algorithms and libraries (e.g., OpenCV and NLTK) are used for the analysis.

[0811] Step 6:

[0812] Storage of characteristic data (server)

[0813] The server stores the extracted characteristic data in the database again, and this characteristic data is used for subsequent optimization of the generation parameters.

[0814] Step 7:

[0815] Optimizing generation parameters (server)

[0816] The server calculates optimal generation parameters based on the stored characteristic data and emotion data. For example, if the user is "happy," it sets generation parameters including bright colors and positive content. This is done using optimization algorithms and machine learning models.

[0817] Step 8:

[0818] Initialization of the generative AI model (server)

[0819] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[0820] Step 9:

[0821] Execution of the generation process (server)

[0822] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., data preprocessing, model input, and execution of generation).

[0823] Step 10:

[0824] Real-time evaluation of artifacts (server)

[0825] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[0826] Step 11:

[0827] Parameter adjustment and regeneration (server)

[0828] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[0829] Step 12:

[0830] Storing and providing the generated data (server)

[0831] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[0832] Example 2

[0833] 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."

[0834] Conventional generation systems created products based solely on a dataset provided by the user, which meant they were unable to fully reflect the user's emotions and intentions. In particular, it was often difficult to precisely reflect the quality and atmosphere of the product desired by the user. Furthermore, they lacked a mechanism for evaluating the quality of the product in real time and adjusting the generation parameters as needed, making it difficult to provide high-quality products to users.

[0835] 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.

[0836] In this invention, the server includes means for processing a specific dataset received from a user, means for analyzing the dataset and extracting its characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for recognizing user emotion data in real time and adjusting the generation parameters based on the emotion data, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and readjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a high-quality product that is more in line with the user's intentions based on the user's emotions and the characteristics of the dataset.

[0837] A "dataset" is a series of data that a user uploads from their own terminal to a server, and includes image data or text data.

[0838] "Analysis" refers to the process undertaken to extract characteristics of an uploaded dataset, using image analysis algorithms and natural language processing algorithms.

[0839] "Generation parameters" are various setting values ​​that are set by the server to create an optimal product based on the extracted characteristics and emotion data.

[0840] The "emotion engine" is a system that analyzes the user's facial expressions, voice, and text input, and recognizes the user's emotional data in real time.

[0841] A "product" is an artifact such as an image or text that is automatically created using set generation parameters.

[0842] "Quality assessment" is the process of evaluating the quality of the generated product, and is performed quantitatively using a quality assessment algorithm.

[0843] "Providing" refers to a series of processes that the server performs to make the product accessible to the user, specifically including storing the product and generating a download link.

[0844] The present invention combines an emotion engine that recognizes user emotions in a system that automatically generates high-quality products from a specific dataset. This system operates in cooperation with a server, a terminal, and a user, and uses the following hardware and software.

[0845] Specifically, the server is a high-performance computer, and is ideally equipped with an environment capable of processing and analyzing large volumes of data. The server receives a specific data set and analyzes it using OpenCV as an image analysis algorithm and BERT as a natural language processing algorithm. This allows it to extract color, shape, and patterns from image data, and keywords, themes, and writing style from text data.

[0846] Next, the generation parameters are automatically set based on the extracted characteristics. Based on the analyzed data, the server calculates the optimal filters and transformation methods (in the case of image generation), as well as the optimal sentence structure and vocabulary selection (in the case of text generation). In this process, the latest generation AI models such as GPT-4 and GAN are often used.

[0847] When a user uploads a dataset from their device, the device's built-in emotion engine analyzes the user's facial expressions, voice, and text input. The emotion engine recognizes the user's emotional state in real time and provides this information to the server. For example, it outputs the user's emotions, such as whether they are happy or sad, as the analysis result. Based on this emotion data, the server adjusts the generation parameters to create a product that matches the user's emotions.

[0848] Once the generation process is complete, the server evaluates the quality of the initial product. It uses a quality evaluation algorithm to quantify the quality of the product. If the evaluation result does not meet the standards, the server adjusts the generation parameters again and repeats the generation. This process ensures high-quality products.

[0849] Finally, the server stores the product and generates a download link that can be accessed by the user, through which the user can download the product.

[0850] As a specific example of operation, if a user uploads a dataset containing bright landscape images and the emotion engine determines that the user is happy, the server will generate landscape images with bright tones. Conversely, if a user requests quiet and calm landscape images and the emotion engine determines that the user is sad, the server will generate landscape images with warm tones.

[0851] An example prompt is:

[0852] "Based on an image dataset uploaded by a user, generate natural landscape images that match his / her emotions. For example, use bright colors if the user is happy, and warm colors if the user is sad."

[0853] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0854] Step 1:

[0855] A user selects a specific dataset from their device and uploads it to the server. For example, when a user selects an image dataset or a text dataset and clicks the upload button, the device sends the dataset to the server. The input is the dataset selected by the user, and the output is the data uploaded to the server.

[0856] Step 2:

[0857] The server receives datasets uploaded by users. The input is the dataset sent by the user, and the output is the dataset stored in the server. Specifically, the server validates the dataset to ensure that the received data has been stored correctly.

[0858] Step 3:

[0859] The server analyzes the received dataset. The input is the dataset stored on the server, and the output is the characteristics extracted from the dataset. Specifically, for image datasets, the server uses OpenCV to analyze color, shape, and patterns, and for text datasets, it uses BERT to extract keywords, themes, and writing style.

[0860] Step 4:

[0861] The server automatically sets generation parameters based on the extracted characteristics. The input is the analysis result, and the output is the set generation parameters. Specifically, the server determines the filter and conversion method for image generation, and the sentence structure and vocabulary selection for text generation.

[0862] Step 5:

[0863] When a user uploads a dataset, the device's emotion engine analyzes the user's facial expressions, voice, text input, etc. The input is the user's emotional data, and the output is the emotional state recognized in real time. Specifically, the device uses emotion recognition software to analyze the user's emotions and sends emotions such as "happy" or "sad" to the server.

[0864] Step 6:

[0865] The server adjusts the generation parameters based on the emotion data recognized by the emotion engine. The input is the emotion data recognized in real time, and the output is the adjusted generation parameters. Specifically, the server adjusts the color tone and tone according to the user's emotion.

[0866] Step 7:

[0867] The server runs a generation process using the set generation parameters to create products. The input is the adjusted generation parameters and dataset, and the output is the generated product. Specifically, the server uses a generative AI model such as GPT-4 or GAN to create new images and text.

[0868] Step 8:

[0869] The server evaluates the quality of the generated product in real time. The input is the generated product, and the output is the evaluation result. Specifically, the server uses a quality evaluation algorithm to quantify the quality of the product and determine whether it meets the criteria.

[0870] Step 9:

[0871] Based on the evaluation results, the server readjusts the generation parameters as necessary and repeats the generation process. The inputs are the quality evaluation results and the current generation parameters, and the outputs are the readjusted generation parameters and the regenerated product.

[0872] Step 10:

[0873] The server finally stores the product and generates a link for users to download it. The input is the final product, and the output is the download link. Users download the high-quality product through this link. Specifically, the server stores the product in cloud storage and notifies the user of the link.

[0874] (Application example 2)

[0875] 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."

[0876] Conventional systems for creating products do not take into account the user's emotions, making it difficult to provide content that matches the user's needs and emotions. In particular, food delivery services require customized menu suggestions based on the user's current emotions, making it necessary to provide a more personalized experience.

[0877] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for recognizing the user's emotions in real time, means for adjusting the generation parameters based on the recognized user's emotions, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a product that is in line with the user's emotions.

[0878] (Definitions of important words)

[0879] A "specific dataset" is a collection of information provided by a user, and is a series of data that may be in the form of images, text, or the like.

[0880] A "receiving means" is a device or software module for obtaining a particular data set from a user.

[0881] "Analysis Means" means any equipment or software containing algorithms or programs for analyzing and examining received data sets and extracting their characteristics.

[0882] "Characteristics" are characteristics such as color, shape, pattern, keywords, themes, and writing style contained within a dataset.

[0883] "Generation parameters" refer to the specific conditions and settings used when creating a product.

[0884] The "means for setting generation parameters" refers to an algorithm or program that determines generation parameters based on characteristics extracted by analysis.

[0885] "Product" refers to the final output, including images, text, and other formats, that is automatically created based on a particular dataset and generation parameters.

[0886] "Creation means" refers to an algorithm or program for creating a product based on generation parameters, and hardware for executing the algorithm or program.

[0887] The "emotion recognition means" is a device that includes an engine or algorithm for analyzing the user's facial expressions, voice, text, etc., and recognizing the user's emotions in real time.

[0888] "Emotion-based generation parameter adjustment means" refers to an algorithm or program for dynamically changing generation parameters in response to recognized user emotion information.

[0889] A "quality evaluation means" is a device that includes an algorithm or program for evaluating the quality of the created product and readjusting the production parameters based on quantitative criteria.

[0890] "Providing means" refers to an interface or module for delivering the final product to the user.

[0891] As an embodiment of the present invention, a system for optimizing menu suggestions for a food delivery service based on user emotions will be described. This system operates by integrating a server, a user device, an emotion engine, an analysis algorithm, a generative AI model, and a quality evaluation algorithm.

[0892] 1. Overall system configuration

[0893] The system uses the following hardware and software:

[0894] Hardware:

[0895] Smartphone: Accepts data input from the user.

[0896] Camera and microphone: Devices for capturing the user's facial expressions and voice.

[0897] Cloud server: Analyzes datasets, recognizes emotions, runs generative AI models, and evaluates the quality of the products.

[0898] software:

[0899] Emotion AI: Analyzes a user's facial expressions, tone of voice, and text input to collect emotional data in real time. For example, common cloud services include Google Cloud's Face API and Amazon Rekognition.

[0900] Analysis algorithms: These include algorithms that extract characteristics of the dataset and set generation parameters, such as color, shape, keywords, themes, and style.

[0901] Generative AI models (e.g., GPT-4 or Transformer-based models): Generate appropriate menu suggestions based on user sentiment data and food menus.

[0902] Quality assessment algorithm: An algorithm for assessing the quality of the proposed menu and readjusting the generation parameters if necessary.

[0903] 2. Data processing flow

[0904] The server processes the data in the following specific steps:

[0905] 1. Receiving the dataset:

[0906] The user uploads a dataset via a smartphone app, which simultaneously captures the user's facial expressions and voice and sends them to a cloud server.

[0907] 2. Dataset Analysis:

[0908] The cloud server analyzes the received data set and extracts relevant characteristics, such as past order history, current facial expression, and tone of voice, to analyze the user's emotions in real time.

[0909] 3. Set generation parameters:

[0910] Based on the extracted characteristics and the recognized emotion data, the generation parameters are automatically set. For example, if the user is tired, the generation parameters are set to suggest a menu that will help them relax.

[0911] 4. Creating the product:

[0912] Using the set generation parameters, the generative AI model automatically creates appropriate menu suggestions.

[0913] 5. Product quality assessment:

[0914] The quality of the generated menu suggestions is evaluated, and if necessary, the generation parameters are adjusted and regenerated.

[0915] 6. Provision of Products:

[0916] The final menu proposal is then provided to the user, who can then view the proposed menu and place an order via a smartphone app.

[0917] 3. Examples of concrete examples and prompts

[0918] As a concrete example, the following prompt sentence is input into the generative AI model:

[0919] If the emotion engine analyzes a photo of a user smiling and determines that the user is "happy":

[0920] "It's a wonderful day today! I'd like to suggest a fresh and delicious seafood spaghetti dish that's perfect for a day like this."

[0921] If the emotion engine determines that the user is "tired" based on facial expressions and voice that indicate fatigue:

[0922] "Thank you for your hard work. How about a relaxing custom meatloaf dinner? It'll warm your soul and your body."

[0923] This makes it possible to provide customized services that are in tune with the user's emotions.

[0924] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0925] Program processing steps

[0926] Step 1:

[0927] The user launches the smartphone app, captures facial expression and voice data via the camera and microphone, selects the data set for food delivery, and sends it to the cloud server via the app.

[0928] Input: facial expression data, audio data, dataset

[0929] Output: Data sent to the cloud server

[0930] Step 2:

[0931] The server analyzes the received dataset and extracts relevant characteristics, applying analysis algorithms to extract color tones and shapes in the case of images, and keywords and themes in the case of text.

[0932] Input: Dataset

[0933] Output: Extracted features

[0934] Step 3:

[0935] The server uses facial expression and voice data sent from the camera and microphone to enable the emotion engine to recognize the user's emotions in real time, using Google Cloud's Face API and Amazon Rekognition to identify emotions such as joy, surprise, and fatigue.

[0936] Input: facial expression data, voice data

[0937] Output: Emotion data

[0938] Step 4:

[0939] The server sets generation parameters based on the characteristics and emotion data, and determines appropriate filters and conditions for content generation, taking into account past order history and user emotion data.

[0940] Input: extracted features, emotion data

[0941] Output: Generated parameters

[0942] Step 5:

[0943] The server uses the configured generation parameters to generate food menu suggestions using a generative AI model (e.g., GPT-4 or a Transformer-based model).

[0944] Input: Generation parameters

[0945] Output: The food menu suggestions created

[0946] Step 6:

[0947] The server evaluates the quality of the generated food menu suggestions, applies a quality evaluation algorithm to measure the quality of the output, and, if necessary, readjusts the generation parameters and runs the generation process again.

[0948] Input: Created food menu proposal

[0949] Output: Evaluation results, regenerated proposals (if necessary)

[0950] Step 7:

[0951] The server then provides the final food menu suggestions to the user, who can then review the suggestions, make a selection, and place an order through a smartphone app. Optionally, a personalized message tailored to the user's mood can also be sent.

[0952] Input: Final generated food menu proposals

[0953] Output: Food menu suggestions provided to the user, customized messages tailored to their emotions

[0954] Specific operation example

[0955] Step 1:

[0956] The user opens the smartphone app, selects the "Food Delivery" menu, and takes a photo of their face with the camera. Audio data is recorded and sent to the cloud, showing how depressed they are.

[0957] Step 2:

[0958] The server analyzes the received image data and picks out keywords from color, shape, or voice data. For example, if the user says "I'm tired," this will be extracted as an analysis result.

[0959] Step 3:

[0960] The server uses Google Cloud's Face API to recognize the user's facial expression, such as "tired," and also analyzes emotions from the tone of their voice.

[0961] Step 4:

[0962] The server sets generation parameters for suggesting relaxing dishes based on the characteristics and the emotional data of "tired."

[0963] Step 5:

[0964] The server uses a generative AI model to create food menu suggestions that fit the "tired" emotion, such as "Custom Meatloaf Dinner."

[0965] Step 6:

[0966] The server applies a quality assessment algorithm to evaluate the quality of the proposed menu, readjusting the generation parameters based on the results and regenerating the menu if necessary.

[0967] Step 7:

[0968] The server presents the user with the final menu suggestion along with a message saying, "Great work! How about a relaxing custom meatloaf dinner?" The user completes the order through the app.

[0969] 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.

[0970] 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.

[0971] 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.

[0972] [Fourth embodiment]

[0973] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0974] 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.

[0975] 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).

[0976] 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.

[0977] 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.

[0978] 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).

[0979] 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.

[0980] 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.

[0981] 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.

[0982] 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.

[0983] 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.

[0984] 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.

[0985] 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."

[0986] As an embodiment of the present invention, a system for automatically generating high-quality products from a specific data set will be described. This system operates in cooperation between a server, a terminal, and a user, and executes the necessary processing at each step.

[0987] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[0988] The server then analyzes the received dataset. It applies analysis algorithms to extract features within the dataset. For image datasets, this could be color, shape, or pattern. For text datasets, this could be keywords, themes, etc. These features are then used to optimize the generation parameters, as described below.

[0989] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[0990] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[0991] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[0992] Finally, the server prepares the product for delivery to the user. The server stores the product and generates a download link for the user, who then obtains the high-quality product through the link. This system enables the efficient generation of high-quality products using a specific dataset, and each step is automated, minimizing user effort.

[0993] A concrete example is a process where a user uploads a text dataset of literary works, and a server generates new literary fragments based on that dataset. In this example, the server analyzes keywords and themes, selects optimal sentence structures and vocabulary, generates new text, evaluates and adjusts the quality, and ultimately delivers a high-quality output to the user. Similarly, a similar process can be applied to generate new music using a dataset of music tracks.

[0994] The processing flow will be explained below.

[0995] Step 1:

[0996] Upload Dataset (User)

[0997] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The device then sends the dataset to the server via an HTTP POST request.

[0998] Step 2:

[0999] Data reception and storage (server)

[1000] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[1001] Step 3:

[1002] Dataset analysis (server)

[1003] The server analyzes the stored dataset. In the case of an image dataset, it extracts features such as color, shape, and pattern. In the case of a text dataset, it extracts keywords, themes, and writing style. This is done using various analysis algorithms and libraries (e.g., OpenCV and NLTK).

[1004] Step 4:

[1005] Storage of characteristic data (server)

[1006] The server stores the extracted characteristic data in a database, which is used for subsequent optimization of the generation parameters.

[1007] Step 5:

[1008] Optimizing generation parameters (server)

[1009] The server calculates the optimal generation parameters based on the stored characteristic data. For example, in the case of image generation, it automatically sets the type of filter and parameters to be used, and in the case of text generation, it automatically sets the sentence structure and vocabulary selection. This is done using optimization algorithms and machine learning models.

[1010] Step 6:

[1011] Initialization of the generative AI model (server)

[1012] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[1013] Step 7:

[1014] Execution of the generation process (server)

[1015] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., preprocessing the data, inputting it into the model, and running the generation).

[1016] Step 8:

[1017] Real-time evaluation of artifacts (server)

[1018] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[1019] Step 9:

[1020] Parameter adjustment and regeneration (server)

[1021] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[1022] Step 10:

[1023] Storing and providing the generated data (server)

[1024] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[1025] Example 1

[1026] 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."

[1027] In today's information society, there is a growing demand for systems that allow users to efficiently create high-quality products based on their own datasets. However, conventional systems often require users to analyze large amounts of data, set optimal generation parameters, and evaluate and adjust the quality of the products, which is a manual process that places a heavy burden on users. Another issue is that the quality of the products cannot be maintained consistently at a high level.

[1028] 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.

[1029] In this invention, the server includes means for receiving a specific dataset from a user, means for analyzing the received dataset and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This allows users to easily obtain high-quality products automatically. Furthermore, the server includes means for applying a generation algorithm used to set the generation parameters, means for implementing an analysis algorithm for extracting specific generation parameters, and means for defining product evaluation criteria and making adjustments based on the quality criteria. This automates the analysis, generation, and evaluation processes, ensuring efficient and consistent product quality.

[1030] A "specific dataset" is a collection of data, regardless of type, such as images, text, or audio, that is input or provided by a user.

[1031] The "means for receiving data from a user" refers to the interface that a user uses to provide data and the function that a server uses to receive it.

[1032] "Analysis and characterization means" are algorithms or software used to understand the content of the data and extract important elements or features.

[1033] The "means for automatically setting generation parameters" is a function that automatically determines the setting values ​​for optimal operation of the generation algorithm based on the extracted characteristics.

[1034] The "means for automatically generating a product" is a function for generating new data or a product for a user based on generation parameters. The product is generated using an artificial intelligence model.

[1035] The "means for evaluating quality and adjusting generation parameters as necessary" is a function for evaluating the quality of the generated deliverables and resetting and regenerating them if they do not meet the standards.

[1036] The "means for providing the final product to the user" is a function for delivering the completed product to the user, and is usually an operation including saving the file and generating a download link.

[1037] A "generative algorithm" is a computational method or program used to create new products from data.

[1038] An "analysis algorithm" is a computational method or program used to analyze data and extract characteristics.

[1039] "Metrics" are standards or indicators established to measure the quality of a product.

[1040] The "means for adjusting based on quality standards" is a function for adjusting generation parameters so as to satisfy evaluation standards and performing regeneration.

[1041] The system of this invention is designed to automatically generate high-quality products from a specific dataset. The system operates in cooperation with a server, terminals, and users, and performs the necessary processing at each step.

[1042] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user might upload a folder of images or a text file with hundreds of pages. The hardware used here is the user's device, and the software is a web browser or a specialized upload application.

[1043] The server then analyzes the received dataset by running Python scripts or specific libraries (e.g., Pandas, OpenCV) to extract color tones, shapes, and patterns in the case of image datasets, and keywords and themes in the case of text datasets. Specifically, the server uses Python's PIL library to analyze the color distribution of the images, and an NLP library (e.g., NLTK) to extract keywords from the text.

[1044] Based on the analysis results, the server automatically sets generation parameters. For example, it generates instructions for the text generation model GPT-3 on what sentence structure and vocabulary to use. Based on the generated theme, the server generates prompt sentences such as "Explain in detail new trends in the market." For image generation, instructions include "Apply a filter that emphasizes blue tones."

[1045] Next, the server creates artifacts from the dataset using the configured generation parameters. The server initializes an appropriate generative AI model (e.g., DALL-E or GPT-3) and runs the model with specific generation parameters. For example, the server uses the DALL-E model to generate new images and the GPT-3 model to generate new sentences. Specifically, the server starts the generation process by calling the function "model.generate(prompt)" in Python code.

[1046] The quality of the generated content is evaluated in real time. The server uses a quality evaluation algorithm (e.g., a library that calculates the Mean Opinion Score or BLEU score) to quantify the quality of the generated content and regenerates it if necessary. Specifically, the server executes the function "quality_score = evaluate(generated_content)" against the algorithm.

[1047] Finally, the server prepares the product to be provided to the user. It saves the product and generates a download link that the user can access. The user can download the product via that link. For example, the server creates a URL for the generated file and sends the link to the user's email address. Specifically, the server executes the function "file_url = generate_download_link(generated_file)" and sends the link using "send_email(user_email, file_url)".

[1048] For example, a user can upload a text dataset of literary works and generate a fragment of a new literary work based on that dataset. The server analyzes keywords and themes, selects the most appropriate sentence structure and vocabulary, and generates the new text. An example of a prompt might be, "Write a poem in 15th-century English incorporating elements from Shakespeare's works."

[1049] The system allows users to obtain high-quality products with minimal effort, and utilizes specialized algorithms and techniques at each step to ensure efficient and consistent product quality.

[1050] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1051] Step 1:

[1052] The user selects a dataset and uploads it to the server. The user selects a specific dataset (e.g., a folder of image files, a text file with several hundred pages, etc.) on their own device. To do this, they operate the file selection dialog in their web browser and click the "Upload" button. The input is the dataset on the user's device, and the output is the dataset file sent to the server. Specifically, after selecting the file, "upload" is executed and an HTTP request is sent to the server.

[1053] Step 2:

[1054] The server analyzes the received dataset. When the server receives the dataset, it begins analysis using Python scripts and libraries such as Pandas and OpenCV. It receives the dataset file as input and outputs the analyzed characteristics (color tone, shape, and pattern for image datasets, and keywords and themes for text datasets). Specifically, the server uses OpenCV to analyze images and NLTK to extract keywords from text. Processes such as "image = cv2.imread('uploaded_image.jpg')" and "keywords = nltk.word_tokenize(text)" are executed.

[1055] Step 3:

[1056] The server automatically sets the generation parameters. Based on the analyzed features, the server determines the generation parameters. The input is the analyzed features, and the output is the generation parameters. Specifically, for image generation, this could be a filter or conversion method, and for text generation, it could be sentence structure or vocabulary selection. For example, "emphasize blue tones" or "describe market trends in detail" are generated. The function "parameters = generate_parameters(extracted_features)" is called.

[1057] Step 4:

[1058] The server automatically creates the product. It creates the product from the dataset based on the set generation parameters. The input is the generation parameters and the original dataset, and the output is the newly generated product. The server initializes an appropriate generative AI model (DALL-E, GPT-3, etc.) and executes the generation process. For example, code such as "generated_image = dalle.generate(image, parameters)" and "generated_text = gpt3.generate(prompt)" is executed.

[1059] Step 5:

[1060] The server evaluates the quality of the generated image. It applies a quality evaluation algorithm to evaluate the quality of the generated image. The input is the new image and the output is a quality score. For example, "quality_score = evaluate(generated_image)" or "text_quality_score = evaluate_text(generated_text)" are executed. If the evaluation criteria are not met, the server adjusts the generation parameters to regenerate the image. It calls the function "adjusted_parameters = adjust_parameters(quality_score)".

[1061] Step 6:

[1062] The server provides the product to the user. After the final product is completed, the server saves the product and generates a download link that the user can access. The input is the final product, and the output is the download link. Specifically, a URL is created for the generated file and the link is sent to the user's email address. Code such as "file_url = generate_download_link(generated_file)" and "send_email(user_email, file_url)" is executed. The user clicks the link to download the product.

[1063] (Application example 1)

[1064] 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."

[1065] In the automatic generation of advertising banners, manual design and quality control require time and effort, and there is no guarantee that high-quality results will be produced. In particular, the quality of the results varies depending on the data set uploaded by the user. Furthermore, unless the generation parameters are adjusted in real time, regeneration to improve quality may not be performed smoothly.

[1066] 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.

[1067] In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, means for providing the final product to the user, and means for providing the created advertising banner, thereby enabling the creation of high-quality advertising banners based on data provided by the user.

[1068] A "specific dataset" is data, such as images or text, provided by a user, that serves as the basis for the analysis and generation process used by the AI ​​model.

[1069] The "receiving means" refers to an interface or function for importing a data set provided by a user into a device such as a server.

[1070] "Means for analyzing and extracting characteristics" refers to algorithms or functions that analyze the received data set to find characteristics such as color tone, shape, keywords, etc.

[1071] The "means for automatically setting generation parameters" is a function that calculates and sets optimal parameters to be applied to the generation model based on the extracted characteristics.

[1072] "Means for automatically creating products" refers to a function that uses set generation parameters to create products such as images and text using an AI model.

[1073] The "means for evaluating the quality of the product" refers to an algorithm or function that evaluates the quality of the generated product in real time and quantifies the evaluation results.

[1074] The "means for adjusting the production parameters" is a function for resetting the production parameters and re-executing the production process based on the results of the product quality evaluation.

[1075] The "means for providing the final product to the user" is a function for storing the quality-guaranteed product in storage and generating a link that allows the user to download it.

[1076] The "means for providing generated advertising banners" is a function for displaying or saving high-quality advertising banners generated based on a data set provided by the user and providing them directly to the user.

[1077] The present invention provides a system for automatically generating high-quality advertising banners from a specific data set. This system operates in cooperation with a server, a terminal, and a user, and performs the necessary processing at each step.

[1078] First, a user selects a specific data set (e.g., product images and product description text) from their device and uploads it to the server. The data provided by the user becomes the raw material for creating advertising banners. The device communicates with the server via a high-speed Internet connection.

[1079] The server then analyzes the received dataset. The server applies an analysis algorithm to extract characteristics within the dataset. For image data, this includes color, shape, and patterns. For text data, this includes key keywords and themes. The analyzed characteristics are used to optimize the generation parameters, as described below.

[1080] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[1081] The server then automatically creates an advertising banner using the set generation parameters. The server initializes an appropriate generative AI model and executes the generation process. For example, it generates high-quality banner ads based on product images and description text. Examples of AI models used include OpenAI's GPT-3 and DALL-E.

[1082] After the product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality advertising banners.

[1083] Finally, the server provides the generated advertisement banner to the user, stores the result, and generates a download link for the user, through which the user can obtain a high-quality advertisement banner.

[1084] For example, if a user uploads a product image and the description text "This is a great product that solves many issues," the server extracts the "keyword" and identifies the color scheme "vivid," and generates an advertising banner based on the following: "keywords: great product, issues, layout: banner, color scheme: vivid."

[1085] Example prompt sentence:

[1086] Create a banner with keywords: great product, issues, layout: banner, and color scheme: vivid.

[1087] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1088] Step 1:

[1089] The user selects a specific dataset (product images and product description text) from the terminal and uploads it to the server. The input data are image files and text files. The terminal sends these files to the server via a high-speed Internet connection. The output is the image and text data received on the server side.

[1090] Step 2:

[1091] The server analyzes the received dataset, applying analysis algorithms to extract color tones, shapes, patterns, etc. in the case of images, and key keywords and themes in the case of text. The input data are the received image and text data, and the output data are the extracted image features and text keywords.

[1092] Step 3:

[1093] The server automatically sets generation parameters based on the extracted characteristics. For example, it determines filters and conversion methods based on image characteristics, and determines the style and structure of the ad based on extracted keywords. The input is the characteristic data extracted in step 2, and the output is the optimal generation parameter settings.

[1094] Step 4:

[1095] The server automatically creates a product (advertising banner) using the set generation parameters. It initializes a generative AI model (e.g., OpenAI's GPT-3 or DALL-E) and runs the generation process. The inputs are the generation parameters, the original image, and text data, and the output is the created advertising banner.

[1096] Step 5:

[1097] The server evaluates the quality of the created product in real time. It applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. The input is the created advertising banner, and the output is the quality evaluation result. If necessary, the generation parameters are readjusted based on the evaluation result.

[1098] Step 6:

[1099] The server prepares the final product for the user, stores it and generates a link for the user to download it. The input is a high-quality advertising banner, and the output is a download link for the product available to the user.

[1100] In this way, a series of automated processes can be used to generate high-quality advertising banners based on user-provided data.

[1101] 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.

[1102] As an embodiment of the present invention, we will explain a configuration that combines a system that automatically generates high-quality products from a specific data set with an emotion engine that recognizes user emotions. This system operates in cooperation between a server, a terminal, and a user, and performs the necessary processing at each step.

[1103] First, a user selects a specific dataset from their device and uploads it to the server. For example, a user can select an image dataset or a text dataset and send it to the server. This dataset becomes the raw material for creating a product.

[1104] The server then analyzes the received dataset. The server applies analysis algorithms to extract characteristics within the dataset. For image datasets, this includes color, shape, and pattern. For text datasets, this includes keywords, themes, and style. These characteristics are then used to optimize the generation parameters, as described below.

[1105] Based on the analysis results, the server automatically sets the generation parameters. The server calculates the optimal generation parameters according to the characteristics of the dataset and applies them to the AI ​​model. Specific examples include filters and conversion methods in the case of image generation, and sentence structure and vocabulary selection in the case of text generation.

[1106] Furthermore, the present invention incorporates an emotion engine that recognizes user emotions. When a user uploads a dataset, the emotion engine analyzes the user's facial expressions, voice, text input, etc. to collect emotion data. For example, it can recognize the user's emotions, such as happiness or sadness, in real time.

[1107] Based on the emotion engine's recognition, the server adjusts the generation parameters. For example, if the user expresses sadness, the server selects warmer colors and softer tones as the generation parameters. Therefore, the generated image will be more suitable for the user's emotion.

[1108] The server then uses the configured generation parameters to create artifacts from the dataset, initializes the appropriate generative AI model, and executes the generation process, for example, generating high-quality images from an image dataset, or generating a new piece of literature from a text dataset.

[1109] After a product is created, the server evaluates its quality in real time. The server applies a quality evaluation algorithm to quantify the quality of the product based on the evaluation criteria. Based on the evaluation results, the server adjusts the generation parameters and regenerates the product if necessary. This process ensures high-quality products.

[1110] Finally, the server prepares the generated results for delivery to the user. The server stores the results and generates a download link for the user. The user can then obtain a high-quality result via the link. This system enables the efficient generation of high-quality results using a specific dataset and user emotion data, and each step is automated to minimize user effort.

[1111] For example, if the emotion engine determines that the user is happy, it will use that data to generate brightly colored images and positive text. In this way, the system can provide optimal results based on the user's emotions. The emotion engine can also analyze the user's emotions from voice input and adjust the results in real time. In this way, customization based on emotions is possible.

[1112] The processing flow will be explained below.

[1113] Step 1:

[1114] Upload Dataset (User)

[1115] The user selects the dataset (e.g., image dataset or text dataset) to be used for generative AI on their device. Specifically, they use a file browser to specify the folder containing the dataset and click the upload button. The user's device then sends the dataset to the server via an HTTP POST request.

[1116] Step 2:

[1117] Data reception and storage (server)

[1118] The server receives the dataset sent by the user. The received dataset is saved in temporary storage and then stored in the database. At this time, the dataset's metadata (file name, upload date and time, etc.) is also recorded.

[1119] Step 3:

[1120] Emotion data collection (terminal, server)

[1121] The device analyzes the user's emotions, such as facial expressions, voice, and text input, using an emotion engine to obtain emotion data in real time. For example, it analyzes facial expression data from a webcam and voice data from a microphone. The obtained emotion data is sent to a server.

[1122] Step 4:

[1123] Emotion data storage (server)

[1124] The server stores the emotion data received from the device in a database. The emotion data is associated with a session ID and a timestamp and is used in subsequent generation processes.

[1125] Step 5:

[1126] Dataset analysis (server)

[1127] The server analyzes the stored dataset. In the case of an image dataset, features such as color tone, shape, and pattern are extracted, and in the case of a text dataset, keywords, themes, and writing style are extracted. Various analysis algorithms and libraries (e.g., OpenCV and NLTK) are used for the analysis.

[1128] Step 6:

[1129] Storage of characteristic data (server)

[1130] The server stores the extracted characteristic data in the database again, and this characteristic data is used for subsequent optimization of the generation parameters.

[1131] Step 7:

[1132] Optimizing generation parameters (server)

[1133] The server calculates optimal generation parameters based on the stored characteristic data and emotion data. For example, if the user is "happy," it sets generation parameters including bright colors and positive content. This is done using optimization algorithms and machine learning models.

[1134] Step 8:

[1135] Initialization of the generative AI model (server)

[1136] The server initializes the generative AI model based on the optimized generation parameters. Specifically, it loads the necessary libraries and frameworks and configures the model structure. For example, it uses deep learning frameworks such as TensorFlow and PyTorch.

[1137] Step 9:

[1138] Execution of the generation process (server)

[1139] The server uses the initialized generative AI model to create artifacts from the dataset. In the case of image generation, it generates new images, and in the case of text generation, it generates new sentences. This includes detailed steps for the generation (e.g., data preprocessing, model input, and execution of generation).

[1140] Step 10:

[1141] Real-time evaluation of artifacts (server)

[1142] The server applies quality assessment algorithms to evaluate the generated output in real time, such as the resolution of the generated images, color consistency, grammar check and consistency of the text, etc.

[1143] Step 11:

[1144] Parameter adjustment and regeneration (server)

[1145] The server adjusts the generation parameters based on the quality evaluation results and regenerates the image as necessary. This process is repeated until the quality meets the standard. Specific adjustments include reselecting filters and optimizing weight parameters.

[1146] Step 12:

[1147] Storing and providing the generated data (server)

[1148] The server then stores the final product that is judged to be of high quality in a database or file system, and generates a link that allows the user to download the product and notifies the user. The user clicks on the link to download and use the product.

[1149] Example 2

[1150] 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."

[1151] Conventional generation systems created products based solely on a dataset provided by the user, which meant they were unable to fully reflect the user's emotions and intentions. In particular, it was often difficult to precisely reflect the quality and atmosphere of the product desired by the user. Furthermore, they lacked a mechanism for evaluating the quality of the product in real time and adjusting the generation parameters as needed, making it difficult to provide high-quality products to users.

[1152] 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.

[1153] In this invention, the server includes means for processing a specific dataset received from a user, means for analyzing the dataset and extracting its characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for recognizing user emotion data in real time and adjusting the generation parameters based on the emotion data, means for automatically creating a product using the set generation parameters, means for evaluating the quality of the created product and readjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a high-quality product that is more in line with the user's intentions based on the user's emotions and the characteristics of the dataset.

[1154] A "dataset" is a series of data that a user uploads from their own terminal to a server, and includes image data or text data.

[1155] "Analysis" refers to the process undertaken to extract characteristics of an uploaded dataset, using image analysis algorithms and natural language processing algorithms.

[1156] "Generation parameters" are various setting values ​​that are set by the server to create an optimal product based on the extracted characteristics and emotion data.

[1157] The "emotion engine" is a system that analyzes the user's facial expressions, voice, and text input, and recognizes the user's emotional data in real time.

[1158] A "product" is an artifact such as an image or text that is automatically created using set generation parameters.

[1159] "Quality assessment" is the process of evaluating the quality of the generated product, and is performed quantitatively using a quality assessment algorithm.

[1160] "Providing" refers to a series of processes that the server performs to make the product accessible to the user, specifically including storing the product and generating a download link.

[1161] The present invention combines an emotion engine that recognizes user emotions in a system that automatically generates high-quality products from a specific dataset. This system operates in cooperation with a server, a terminal, and a user, and uses the following hardware and software.

[1162] Specifically, the server is a high-performance computer, and is ideally equipped with an environment capable of processing and analyzing large volumes of data. The server receives a specific data set and analyzes it using OpenCV as an image analysis algorithm and BERT as a natural language processing algorithm. This allows it to extract color, shape, and patterns from image data, and keywords, themes, and writing style from text data.

[1163] Next, the generation parameters are automatically set based on the extracted characteristics. Based on the analyzed data, the server calculates the optimal filters and transformation methods (in the case of image generation), as well as the optimal sentence structure and vocabulary selection (in the case of text generation). In this process, the latest generation AI models such as GPT-4 and GAN are often used.

[1164] When a user uploads a dataset from their device, the device's built-in emotion engine analyzes the user's facial expressions, voice, and text input. The emotion engine recognizes the user's emotional state in real time and provides this information to the server. For example, it outputs the user's emotions, such as whether they are happy or sad, as the analysis result. Based on this emotion data, the server adjusts the generation parameters to create a product that matches the user's emotions.

[1165] Once the generation process is complete, the server evaluates the quality of the initial product. It uses a quality evaluation algorithm to quantify the quality of the product. If the evaluation result does not meet the standards, the server adjusts the generation parameters again and repeats the generation. This process ensures high-quality products.

[1166] Finally, the server stores the product and generates a download link that can be accessed by the user, through which the user can download the product.

[1167] As a specific example of operation, if a user uploads a dataset containing bright landscape images and the emotion engine determines that the user is happy, the server will generate landscape images with bright tones. Conversely, if a user requests quiet and calm landscape images and the emotion engine determines that the user is sad, the server will generate landscape images with warm tones.

[1168] An example prompt is:

[1169] "Based on an image dataset uploaded by a user, generate natural landscape images that match his / her emotions. For example, use bright colors if the user is happy, and warm colors if the user is sad."

[1170] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1171] Step 1:

[1172] A user selects a specific dataset from their device and uploads it to the server. For example, when a user selects an image dataset or a text dataset and clicks the upload button, the device sends the dataset to the server. The input is the dataset selected by the user, and the output is the data uploaded to the server.

[1173] Step 2:

[1174] The server receives datasets uploaded by users. The input is the dataset sent by the user, and the output is the dataset stored in the server. Specifically, the server validates the dataset to ensure that the received data has been stored correctly.

[1175] Step 3:

[1176] The server analyzes the received dataset. The input is the dataset stored on the server, and the output is the characteristics extracted from the dataset. Specifically, for image datasets, the server uses OpenCV to analyze color, shape, and patterns, and for text datasets, it uses BERT to extract keywords, themes, and writing style.

[1177] Step 4:

[1178] The server automatically sets generation parameters based on the extracted characteristics. The input is the analysis result, and the output is the set generation parameters. Specifically, the server determines the filter and conversion method for image generation, and the sentence structure and vocabulary selection for text generation.

[1179] Step 5:

[1180] When a user uploads a dataset, the device's emotion engine analyzes the user's facial expressions, voice, text input, etc. The input is the user's emotional data, and the output is the emotional state recognized in real time. Specifically, the device uses emotion recognition software to analyze the user's emotions and sends emotions such as "happy" or "sad" to the server.

[1181] Step 6:

[1182] The server adjusts the generation parameters based on the emotion data recognized by the emotion engine. The input is the emotion data recognized in real time, and the output is the adjusted generation parameters. Specifically, the server adjusts the color tone and tone according to the user's emotion.

[1183] Step 7:

[1184] The server runs a generation process using the set generation parameters to create products. The input is the adjusted generation parameters and dataset, and the output is the generated product. Specifically, the server uses a generative AI model such as GPT-4 or GAN to create new images and text.

[1185] Step 8:

[1186] The server evaluates the quality of the generated product in real time. The input is the generated product, and the output is the evaluation result. Specifically, the server uses a quality evaluation algorithm to quantify the quality of the product and determine whether it meets the criteria.

[1187] Step 9:

[1188] Based on the evaluation results, the server readjusts the generation parameters as necessary and repeats the generation process. The inputs are the quality evaluation results and the current generation parameters, and the outputs are the readjusted generation parameters and the regenerated product.

[1189] Step 10:

[1190] The server finally stores the product and generates a link for users to download it. The input is the final product, and the output is the download link. Users download the high-quality product through this link. Specifically, the server stores the product in cloud storage and notifies the user of the link.

[1191] (Application example 2)

[1192] 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."

[1193] Conventional systems for creating products do not take into account the user's emotions, making it difficult to provide content that matches the user's needs and emotions. In particular, food delivery services require customized menu suggestions based on the user's current emotions, making it necessary to provide a more personalized experience.

[1194] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving a specific data set from a user, means for analyzing the received data set and extracting characteristics, means for automatically setting generation parameters based on the extracted characteristics, means for automatically creating a product using the set generation parameters, means for recognizing the user's emotions in real time, means for adjusting the generation parameters based on the recognized user's emotions, means for evaluating the quality of the created product and adjusting the generation parameters as necessary, and means for providing the final product to the user. This makes it possible to provide a product that is in line with the user's emotions.

[1195] (Definitions of important words)

[1196] A "specific dataset" is a collection of information provided by a user, and is a series of data that may be in the form of images, text, or the like.

[1197] A "receiving means" is a device or software module for obtaining a particular data set from a user.

[1198] "Analysis Means" means any equipment or software containing algorithms or programs for analyzing and examining received data sets and extracting their characteristics.

[1199] "Characteristics" are characteristics such as color, shape, pattern, keywords, themes, and writing style contained within a dataset.

[1200] "Generation parameters" refer to the specific conditions and settings used when creating a product.

[1201] The "means for setting generation parameters" refers to an algorithm or program that determines generation parameters based on characteristics extracted by analysis.

[1202] "Product" refers to the final output, including images, text, and other formats, that is automatically created based on a particular dataset and generation parameters.

[1203] "Creation means" refers to an algorithm or program for creating a product based on generation parameters, and hardware for executing the algorithm or program.

[1204] The "emotion recognition means" is a device that includes an engine or algorithm for analyzing the user's facial expressions, voice, text, etc., and recognizing the user's emotions in real time.

[1205] "Emotion-based generation parameter adjustment means" refers to an algorithm or program for dynamically changing generation parameters in response to recognized user emotion information.

[1206] A "quality evaluation means" is a device that includes an algorithm or program for evaluating the quality of the created product and readjusting the production parameters based on quantitative criteria.

[1207] "Providing means" refers to an interface or module for delivering the final product to the user.

[1208] As an embodiment of the present invention, a system for optimizing menu suggestions for a food delivery service based on user emotions will be described. This system operates by integrating a server, a user device, an emotion engine, an analysis algorithm, a generative AI model, and a quality evaluation algorithm.

[1209] 1. Overall system configuration

[1210] The system uses the following hardware and software:

[1211] Hardware:

[1212] Smartphone: Accepts data input from the user.

[1213] Camera and microphone: Devices for capturing the user's facial expressions and voice.

[1214] Cloud server: Analyzes datasets, recognizes emotions, runs generative AI models, and evaluates the quality of the products.

[1215] software:

[1216] Emotion AI: Analyzes a user's facial expressions, tone of voice, and text input to collect emotional data in real time. For example, common cloud services include Google Cloud's Face API and Amazon Rekognition.

[1217] Analysis algorithms: These include algorithms that extract characteristics of the dataset and set generation parameters, such as color, shape, keywords, themes, and style.

[1218] Generative AI models (e.g., GPT-4 or Transformer-based models): Generate appropriate menu suggestions based on user sentiment data and food menus.

[1219] Quality assessment algorithm: An algorithm for assessing the quality of the proposed menu and readjusting the generation parameters if necessary.

[1220] 2. Data processing flow

[1221] The server processes the data in the following specific steps:

[1222] 1. Receiving the dataset:

[1223] The user uploads a dataset via a smartphone app, which simultaneously captures the user's facial expressions and voice and sends them to a cloud server.

[1224] 2. Dataset Analysis:

[1225] The cloud server analyzes the received data set and extracts relevant characteristics, such as past order history, current facial expression, and tone of voice, to analyze the user's emotions in real time.

[1226] 3. Set generation parameters:

[1227] Based on the extracted characteristics and the recognized emotion data, the generation parameters are automatically set. For example, if the user is tired, the generation parameters are set to suggest a menu that will help them relax.

[1228] 4. Creating the product:

[1229] Using the set generation parameters, the generative AI model automatically creates appropriate menu suggestions.

[1230] 5. Product quality assessment:

[1231] The quality of the generated menu suggestions is evaluated, and if necessary, the generation parameters are adjusted and regenerated.

[1232] 6. Provision of Products:

[1233] The final menu proposal is then provided to the user, who can then view the proposed menu and place an order via a smartphone app.

[1234] 3. Examples of concrete examples and prompts

[1235] As a concrete example, the following prompt sentence is input into the generative AI model:

[1236] If the emotion engine analyzes a photo of a user smiling and determines that the user is "happy":

[1237] "It's a wonderful day today! I'd like to suggest a fresh and delicious seafood spaghetti dish that's perfect for a day like this."

[1238] If the emotion engine determines that the user is "tired" based on facial expressions and voice that indicate fatigue:

[1239] "Thank you for your hard work. How about a relaxing custom meatloaf dinner? It'll warm your soul and your body."

[1240] This makes it possible to provide customized services that are in tune with the user's emotions.

[1241] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1242] Program processing steps

[1243] Step 1:

[1244] The user launches the smartphone app, captures facial expression and voice data via the camera and microphone, selects the data set for food delivery, and sends it to the cloud server via the app.

[1245] Input: facial expression data, audio data, dataset

[1246] Output: Data sent to the cloud server

[1247] Step 2:

[1248] The server analyzes the received dataset and extracts relevant characteristics, applying analysis algorithms to extract color tones and shapes in the case of images, and keywords and themes in the case of text.

[1249] Input: Dataset

[1250] Output: Extracted features

[1251] Step 3:

[1252] The server uses facial expression and voice data sent from the camera and microphone to enable the emotion engine to recognize the user's emotions in real time, using Google Cloud's Face API and Amazon Rekognition to identify emotions such as joy, surprise, and fatigue.

[1253] Input: facial expression data, voice data

[1254] Output: Emotion data

[1255] Step 4:

[1256] The server sets generation parameters based on the characteristics and emotion data, and determines appropriate filters and conditions for content generation, taking into account past order history and user emotion data.

[1257] Input: extracted features, emotion data

[1258] Output: Generated parameters

[1259] Step 5:

[1260] The server uses the configured generation parameters to generate food menu suggestions using a generative AI model (e.g., GPT-4 or a Transformer-based model).

[1261] Input: Generation parameters

[1262] Output: The food menu suggestions created

[1263] Step 6:

[1264] The server evaluates the quality of the generated food menu suggestions, applies a quality evaluation algorithm to measure the quality of the output, and, if necessary, readjusts the generation parameters and runs the generation process again.

[1265] Input: Created food menu proposal

[1266] Output: Evaluation results, regenerated proposals (if necessary)

[1267] Step 7:

[1268] The server then provides the final food menu suggestions to the user, who can then review the suggestions, make a selection, and place an order through a smartphone app. Optionally, a personalized message tailored to the user's mood can also be sent.

[1269] Input: Final generated food menu proposals

[1270] Output: Food menu suggestions provided to the user, customized messages tailored to their emotions

[1271] Specific operation example

[1272] Step 1:

[1273] The user opens the smartphone app, selects the "Food Delivery" menu, and takes a photo of their face with the camera. Audio data is recorded and sent to the cloud, showing how depressed they are.

[1274] Step 2:

[1275] The server analyzes the received image data and picks out keywords from color, shape, or voice data. For example, if the user says "I'm tired," this will be extracted as an analysis result.

[1276] Step 3:

[1277] The server uses Google Cloud's Face API to recognize the user's facial expression, such as "tired," and also analyzes emotions from the tone of their voice.

[1278] Step 4:

[1279] The server sets generation parameters for suggesting relaxing dishes based on the characteristics and the emotional data of "tired."

[1280] Step 5:

[1281] The server uses a generative AI model to create food menu suggestions that fit the "tired" emotion, such as "Custom Meatloaf Dinner."

[1282] Step 6:

[1283] The server applies a quality assessment algorithm to evaluate the quality of the proposed menu, readjusting the generation parameters based on the results and regenerating the menu if necessary.

[1284] Step 7:

[1285] The server presents the user with the final menu suggestion along with a message saying, "Great work! How about a relaxing custom meatloaf dinner?" The user completes the order through the app.

[1286] 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.

[1287] 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.

[1288] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1289] 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.

[1290] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1291] 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.

[1292] 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).

[1293] 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.

[1294] 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."

[1295] 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.

[1296] 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).

[1297] 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.

[1298] 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.

[1299] 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.

[1300] 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.

[1301] 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.

[1302] 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.

[1303] 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.

[1304] 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.

[1305] 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.

[1306] 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.

[1307] The following is further disclosed regarding the above embodiment.

[1308] (Claim 1)

[1309] means for receiving a particular data set from a user;

[1310] means for analyzing the received data set and extracting characteristics;

[1311] means for automatically setting generation parameters based on the extracted characteristics;

[1312] means for automatically creating a product using set generation parameters;

[1313] means for assessing the quality of the product produced and adjusting production parameters as necessary;

[1314] means for providing the final product to a user;

[1315] A system including:

[1316] (Claim 2)

[1317] 10. The system of claim 1, comprising an optimization algorithm for setting optimal generation parameters based on characteristics of a dataset received from a user.

[1318] (Claim 3)

[1319] The system according to claim 1, further comprising means for adjusting production parameters in real time based on the evaluation results of the produced product.

[1320] "Example 1"

[1321] (Claim 1)

[1322] means for receiving a particular data set from a user;

[1323] means for analyzing the received data set and extracting characteristics;

[1324] means for automatically setting generation parameters based on the extracted characteristics;

[1325] means for automatically creating a product using set generation parameters;

[1326] means for assessing the quality of the product produced and adjusting production parameters as necessary;

[1327] means for providing the final product to a user;

[1328] A system comprising: means for applying a generation algorithm used to set generation parameters;

[1329] means for implementing an analysis algorithm for extracting specific generation parameters;

[1330] a means for defining product evaluation criteria and making adjustments based on quality criteria;

[1331] The system further comprises:

[1332] (Claim 2)

[1333] 10. The system of claim 1, comprising an optimization algorithm for setting optimal generative parameters based on characteristics of a dataset received from a user, and in particular means for initializing and applying a generative AI model to the analysis results.

[1334] (Claim 3)

[1335] 2. The system of claim 1, further comprising means for adjusting production parameters in real time based on the results of evaluation of the produced product. 3. The system of claim 1, further comprising means for automatically evaluating the quality of the product and triggering any necessary regeneration processes.

[1336] "Application Example 1"

[1337] New Claims:

[1338] (Claim 1)

[1339] means for receiving a particular data set from a user;

[1340] means for analyzing the received data set and extracting characteristics;

[1341] means for automatically setting generation parameters based on the extracted characteristics;

[1342] means for automatically creating a product using set generation parameters;

[1343] means for assessing the quality of the product produced and adjusting production parameters as necessary;

[1344] means for providing the final product to a user;

[1345] means for providing the generated advertising banner;

[1346] A system including:

[1347] (Claim 2)

[1348] 10. The system of claim 1, comprising an optimization algorithm for setting optimal generation parameters based on characteristics of a dataset received from a user.

[1349] (Claim 3)

[1350] The system according to claim 1, further comprising means for adjusting production parameters in real time based on the evaluation results of the produced product.

[1351] "Example 2: Combining Emotion Engines"

[1352] (Claim 1)

[1353] means for processing a particular data set received from a user;

[1354] a means for analyzing the dataset and extracting its characteristics;

[1355] means for automatically setting generation parameters based on the extracted characteristics;

[1356] means for recognizing user emotion data in real time and adjusting generation parameters based on the emotion data;

[1357] means for automatically creating a product using set creation parameters;

[1358] means for assessing the quality of the product produced and readjusting production parameters as necessary;

[1359] means for providing the final product to a user;

[1360] A system including:

[1361] (Claim 2)

[1362] 10. The system of claim 1, comprising an optimization algorithm for setting optimal generation parameters based on both the dataset and the emotion data received from the user.

[1363] (Claim 3)

[1364] 2. The system according to claim 1, further comprising: means for adjusting generation parameters in real time based on evaluation results of the created product; and means for redetecting emotion data in real time and further adjusting the generation parameters.

[1365] "Application example 2 when combining emotion engines"

[1366] New Claims

[1367] (Claim 1)

[1368] means for receiving a particular data set from a user;

[1369] means for analyzing the received data set and extracting characteristics;

[1370] means for automatically setting generation parameters based on the extracted characteristics;

[1371] means for automatically creating a product using set generation parameters;

[1372] means for recognizing user emotions in real time;

[1373] means for adjusting generation parameters based on the recognized user emotion;

[1374] means for assessing the quality of the product produced and adjusting production parameters as necessary;

[1375] means for providing the final product to a user;

[1376] A system including:

[1377] (Claim 2)

[1378] 10. The system of claim 1, further comprising an optimization algorithm for setting optimal generation parameters based on user sentiment and past data sets.

[1379] (Claim 3)

[1380] 10. The system of claim 1, further comprising means for providing customized suggestions based on sentiment analysis and past usage history. [Explanation of symbols]

[1381] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving a particular data set from a user; means for analyzing the received data set and extracting characteristics; means for automatically setting generation parameters based on the extracted characteristics; means for automatically creating a product using set generation parameters; means for assessing the quality of the product produced and adjusting production parameters as necessary; means for providing the final product to a user; A system including:

2. The system of claim 1 , further comprising an optimization algorithm for setting optimal generation parameters based on characteristics of a dataset received from a user.

3. The system according to claim 1, further comprising means for adjusting production parameters in real time based on the evaluation results of the produced product.

Images