system

Abstract

A system is provided.SOLUTION: Means for receiving, from a plurality of clothing images selected by a user on the e-commerce website, identifiers and image URLs of the selected clothing images, means for acquiring the plurality of clothing images based on the identifiers and the image URLs and analyzing features of clothing in each image, means for executing a generation AI model for synthesizing a generated dressing image using the acquired clothing images, means for storing the generated dressing image and generating a URL for display to a user, and means for transmitting the URL to the user terminal; Means for enabling to be displayed at the user terminal.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] On conventional e-commerce sites, users cannot see how clothes will look when actually worn, which leads to frequent failed purchases and returns. Another problem is that it is difficult for users to see in advance how the overall look will look when various clothing items are combined, which reduces their motivation to purchase. There is a need to solve this problem and increase the success rate of purchases and reduce the hassle of returns by providing users with a concrete image of how the clothes will look before they buy. [Means for solving the problem]

[0005] The present invention provides a means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on an e-commerce site, acquiring the multiple clothing images based on the identifier and image URL, and analyzing the characteristics of the clothing in each image. It also provides a means for executing a generative AI model using the analyzed characteristic information and synthesizing a generated outfit image using the acquired clothing images. Furthermore, it provides a means for saving the generated outfit image and generating a URL for displaying it to the user, and a means for transmitting the generated URL to the user's terminal so that the generated URL can be displayed on the user's terminal, thereby solving the above-mentioned problems.

[0006] "User" refers to a person who uses a service that selects clothing on an e-commerce site and generates try-on images based on the selection.

[0007] "E-commerce site" refers to an online platform that sells clothing and other products over the internet.

[0008] "Clothing Images" refers to image data containing visual information about clothing sold on an e-commerce site.

[0009] "Identifier" refers to information (such as an ID) that uniquely identifies a specific clothing image.

[0010] "Image URL" refers to the web address used to access a specific clothing image located on the Internet.

[0011] "Getting" refers to the process in which the server downloads clothing image data based on the specified identifier and image URL.

[0012] "Analysis" refers to the process of extracting the contours and characteristics of clothing from the acquired clothing images and using that information in the next process.

[0013] "Generative AI model" refers to a model that uses artificial intelligence technology (e.g., GAN) to synthesize generated outfit images.

[0014] "Synthesis" refers to the process of generating a new outfit image from multiple clothing images.

[0015] "Saving" refers to the process of storing the generated outfit image as a file.

[0016] "URL" refers to the web address for accessing the generated outfit image.

[0017] "Sending" refers to the process of delivering the URL of the generated outfit image to the user's terminal.

[0018] "Display" refers to a process that allows the generated outfit image to be visually confirmed on the user terminal. [Brief explanation of the drawings]

[0019] [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

[0020] 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.

[0021] First, the terms used in the following description will be explained.

[0022] 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).

[0023] 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.

[0024] 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.

[0025] 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.

[0026] 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."

[0027] [First embodiment]

[0028] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0029] 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.

[0030] 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).

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] 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.

[0037] 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.

[0038] 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.

[0039] 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."

[0040] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[0041] User selection of clothing

[0042] First, a user accesses an e-commerce site and selects the clothing item they want from various categories (tops, pants, shoes, etc.). If the user selects, for example, a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[0043] Server receives selection information and acquires images

[0044] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[0045] Image analysis and feature extraction

[0046] The server analyzes the acquired image data and extracts key features of the clothing (e.g., collar, sleeves, trouser legs, etc.). The feature information obtained from this analysis plays an important role in the subsequent synthesis process.

[0047] Image synthesis using generative AI models

[0048] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[0049] Save composite image and generate URL

[0050] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[0051] Sending and displaying composite images

[0052] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[0053] Specific examples

[0054] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0055] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0056] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0057] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0058] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0059] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0060] 6. The server saves the generated try-on image as https: / / example.com / generated_outfit123456.png.

[0061] 7. The server sends this URL to the user's device.

[0062] 8. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user can check the new outfit image.

[0063] This series of processes allows users to get a concrete image of how the clothing will look before purchasing, reducing the risk of making a wrong purchase or having to return the item.

[0064] The processing flow will be explained below.

[0065] Step 1:

[0066] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[0067] Step 2:

[0068] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0069] Step 3:

[0070] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0071] Step 4:

[0072] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[0073] Step 5:

[0074] The server runs a generative AI model, such as a generative adversarial network (GAN), to synthesize clothing images based on the analyzed feature information. This process generates a try-on image combining the user's chosen clothing.

[0075] Step 6:

[0076] The server saves the generated try-on image and generates a URL that can be accessed by the user. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit123456.png.

[0077] Step 7:

[0078] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[0079] Step 8:

[0080] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look in real life and decide whether to purchase it.

[0081] Through this series of processes, users can get a concrete idea of ​​how the clothing will look before purchasing, reducing the risk of making a mistake in purchasing or the hassle of having to return the item.

[0082] Example 1

[0083] 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."

[0084] In conventional e-commerce systems, when users purchase clothing, it is difficult for them to see how it will look when worn without actually trying it on. Even if users select the clothing combinations they want, the combinations are not displayed in a natural arrangement, which reduces post-purchase satisfaction and increases the hassle of returning items. Furthermore, there was a lack of a way to effectively manage and save each user's outfit images, making it difficult to meet individual needs.

[0085] 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.

[0086] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on the e-commerce site; means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for saving the generated outfit image and generating a URL for displaying it to the user; means for transmitting the URL to the user terminal and enabling display on the user terminal; and means for inputting a desired clothing combination by the user and generating a prompt message for the generative AI model to generate a try-on image based on the input. This allows users to see a specific and natural look before purchasing, reducing the hassle of purchasing mistakes and returns. Furthermore, it is possible to effectively manage and store the outfit images generated for each user, thereby providing services tailored to individual needs.

[0087] "User" refers to an individual or group that selects clothing on an e-commerce site and requests the generation of try-on images.

[0088] "E-commerce site" refers to a website that sells clothing and other products over the Internet.

[0089] "Identifier" refers to an ID or tag that uniquely identifies a specific clothing image.

[0090] "Image URL" refers to a Uniform Resource Locator (URL) for accessing a clothing image stored on the Internet.

[0091] "Server" refers to a computer system that receives and processes requests from users.

[0092] "Features" refer to key parts or characteristics of the clothing in the image, such as the collar, sleeves, and hem.

[0093] "Generative AI model" refers to an artificial intelligence model that generates try-on images based on the characteristics of clothing selected by the user. An example is a generative adversarial network (GAN).

[0094] A "try-on image" refers to an image created by combining multiple pieces of clothing selected by the user, creating a composite image that looks as if the clothing is actually being worn.

[0095] "Saving" refers to recording the generated try-on images in a database or storage device.

[0096] "URL generation" refers to creating a uniform resource locator (URL) for accessing the saved try-on images.

[0097] "Prompt sentence" refers to an instruction sentence provided as input to a generative AI model to generate try-on images.

[0098] "User terminal" refers to a device such as a computer, smartphone, or tablet that a user uses to access an e-commerce site, make selections, and check images of items being tried on.

[0099] The present invention provides a system that receives the identifiers and image URLs of selected clothing images from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[0100] User selection of clothing

[0101] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user selects the clothing item they want from various categories (tops, pants, shoes, etc.). For example, if a user selects a "red top" and "blue jeans," this selection information is sent from the user's device to the server.

[0102] Server receives selection information and acquires images

[0103] The server retrieves clothing images from the specified URL based on the identifier (ID) and image URL received from the user's device. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the image data will be downloaded from these URLs. Specifically, the server retrieves images using a client library such as curl or requests.

[0104] Image analysis and feature extraction

[0105] The server analyzes the acquired image data using image analysis software such as OpenCV or TENSORFLOW (registered trademark) to extract the main features of the clothing (collar, sleeves, trouser hems, etc.) The feature information obtained from this analysis plays an important role in the subsequent image synthesis process.

[0106] Image synthesis using generative AI models

[0107] The server uses a generative AI model such as a generative adversarial network (GAN) to synthesize clothing images based on the extracted feature information. An image of what the clothing selected by the user would look like when actually worn is generated. For example, an outfit image of a red top and blue jeans naturally combined can be created. To facilitate this process, a prompt is used as input to the generative AI model. An example of a prompt is "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate an image of the clothing being tried on."

[0108] Save composite image and generate URL

[0109] The generated try-on image is saved to a storage device by the server. The server generates a URL so that the saved try-on image can be accessed. For example, the URL of the generated try-on image is configured as https: / / example.com / generated_outfit.png.

[0110] Sending and displaying composite images

[0111] The server sends the generated URL to the user's device. The user can open this URL on their device to view the new outfit image. For example, the received URL https: / / example.com / generated_outfit.png can be displayed in a web browser or dedicated application.

[0112] Specific examples

[0113] For example, if a user selects a denim jacket and black jeans, the following happens:

[0114] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0115] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0116] 3. The server accesses the sent URL and downloads the image.

[0117] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0118] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0119] 6. The server saves the generated try-on image in storage and generates a URL.

[0120] 7. The server sends the generated URL to the user's device.

[0121] 8. The device displays the received URL, and the user can check the new outfit image.

[0122] This series of processes allows users to get a concrete image of how the garment will look before purchasing, reducing the risk of buying something wrong or having to return it.

[0123] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0124] Step 1: User selects clothing

[0125] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user navigates through categories (tops, pants, shoes, etc.) on the site and selects the desired clothing item. The input is the identifier (ID) and image URL of the selected clothing item. For example, if the user selects "red top" and "blue jeans," the output is the identifier and image URL of the selected clothing item. Specifically, when the user confirms their selection, the selection information is sent from the user's device to the server.

[0126] Step 2: Server receives selection information

[0127] The server receives the identifier (ID) and image URL sent from the user's device. The input is the identifier and image URL sent from the user's device, and the output is to store this information as received data. Specifically, the server receives the HTTPS request and saves the selected information in a database.

[0128] Step 3: Acquiring image data

[0129] The server downloads the specified image data based on the URL received. The input is an identifier and an image URL, and the output is the retrieved image data. Specifically, the server accesses the image URL using a client library such as curl or requests and saves the image data in the server's storage. For example, download the image data of a red top from the URL https: / / example.com / red_top.png.

[0130] Step 4: Image analysis and feature extraction

[0131] The server analyzes the acquired image data and extracts the key features of the clothing. The input is the acquired image data, and the output is the extracted feature information. Specifically, the server uses image analysis software such as OpenCV or TensorFlow to identify features such as the collar, sleeves, and trouser hems of the clothing. For example, the analysis results in identifying the position of the collar and the length of the sleeves of a red top.

[0132] Step 5: Image synthesis using generative AI models

[0133] The server uses the analyzed feature information as input and generates a try-on image using a generative AI model (e.g., a generative adversarial network (GAN)). The input is the extracted feature information, and the output is the generated try-on image. Specifically, the server converts the feature information into a prompt and inputs it into the generative AI model. For example, the server generates a prompt such as, "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate a try-on image." Based on this, the generative AI model generates an image that combines a red top and blue jeans.

[0134] Step 6: Save the composite image and generate a URL

[0135] The server saves the generated try-on image to a storage device and generates an accessible URL. The input is the generated try-on image, and the output is the URL of the saved try-on image. Specifically, the server saves the try-on image as a file and generates a unique URL for that file. For example, the generated URL might be https: / / example.com / generated_outfit.png.

[0136] Step 7: Send and view the composite image

[0137] The server sends the generated URL to the user's device, allowing the user to view the new outfit image on the device. The input is the generated URL, and the output is the try-on image displayed on the user's device. Specifically, the server returns a URL as an API response, and the user's device receives and displays this URL. For example, the user opens the URL https: / / example.com / generated_outfit.png in a browser, and the new outfit image is displayed.

[0138] (Application example 1)

[0139] 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."

[0140] On conventional e-commerce sites, users could not try on clothes before purchasing, making it difficult to confirm whether the item actually fit them. As a result, users had to go through the trouble of returning or exchanging the item after purchase if the size or design did not fit. In addition, it was difficult for users to visualize what the selected clothing would look like in real life, which led to a decrease in purchasing motivation.

[0141] 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.

[0142] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site; means for acquiring the plurality of clothing images based on the identifier and the image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for the user to upload a full-body image of themselves and analyze the acquired clothing images and the user's full-body image; means for executing a generative AI model that generates a try-on image in which the selected clothing naturally fits the user's full-body image based on the analyzed characteristics; means for saving the generated try-on image and generating a URL for displaying it to the user; and means for transmitting the URL to the user terminal and enabling display on the user terminal. This allows the user to specifically check the try-on image based on their full-body image, allowing them to have a detailed image before purchasing, which increases their motivation to purchase and reduces the hassle of returns and exchanges due to mismatched sizes or designs.

[0143] A "user" is a consumer who uses an e-commerce site to select and purchase products.

[0144] An "e-commerce site" is a website that sells and buys goods and services over the Internet.

[0145] "Clothing Images" are digital images of clothing displayed on e-commerce sites.

[0146] An "identifier" is information for uniquely identifying a specific clothing image.

[0147] An "image URL" is an address (URL) for accessing a specific image file on the Internet.

[0148] A "full-body image" is a photograph showing the user's entire body.

[0149] "Feature analysis" is a technique for extracting specific elements or patterns from an image.

[0150] A "generative AI model" is an algorithm that uses artificial intelligence techniques to generate new images or data, such as generative adversarial networks (GANs).

[0151] The "try-on image" is a composite image of the user's whole body and a clothing image, simulating the appearance of the clothing when the user actually tries it on.

[0152] "Saving" means to hold the generated try-on image as data.

[0153] "URL generation" means creating a new URL to access the saved try-on image.

[0154] A "user terminal" is a device used by a user to access an e-commerce site, and includes a smartphone, tablet, computer, etc.

[0155] "Transmission" means sending data or information from the server to the user terminal.

[0156] "Display" refers to the visual presentation of information or images on the screen of a user terminal.

[0157] The present invention is a system that generates try-on images based on multiple clothing images selected by a user on an e-commerce site. The following describes how to implement this system step by step.

[0158] System Overview

[0159] The system consists of a server that runs on the backend of an e-commerce site and a terminal used by users. Users upload a full-body image of themselves and select clothing images on the site. The server analyzes this data and uses a generative AI model to generate try-on images and provide them to the user.

[0160] Hardware and software used

[0161] Smartphones, tablets, and computers: devices used by users to access

[0162] Server: Receives, stores, analyzes, and synthesizes data. Specifically, it uses cloud services such as AWS (registered trademark) and Google (registered trademark) Cloud Platform.

[0163] Generative AI models: Use generative adversarial networks (GANs) using PyTorch and TensorFlow.

[0164] Image analysis software: For example, OpenCV or Scikit-Image can be used to extract image features.

[0165] Data Flow and Processing

[0166] 1. User Action:

[0167] Users access an e-commerce site, select clothing items, and upload a full-body image of themselves.

[0168] 2. Data reception and processing by the server:

[0169] The server receives the identifier and image URL of the garment selected by the user, as well as a full-body image of the user.

[0170] 3. Image Acquisition and Feature Analysis:

[0171] The server retrieves the clothing image based on the identifier and the image URL, and analyzes the characteristics of the clothing and the user's full-body image using image analysis software.

[0172] 4. Generate fitting images:

[0173] Using a generative AI model (e.g., GAN using PyTorch or TensorFlow), a fitting image is generated that naturally combines the selected clothing with the user's full-body image based on the analyzed features.

[0174] 5. Save the generated image and generate a URL:

[0175] The generated try-on images are stored on the server, and a URL is generated that the user can access.

[0176] 6. User Notification and Display:

[0177] The server sends the generated URL to the user's device, and the user can open the URL to check the try-on image.

[0178] Examples of concrete examples and prompts

[0179] Specific examples

[0180] For example, if a user selects a "red top" and "blue jeans" and uploads a full-body image of themselves, the server will process the following:

[0181] 1. The server retrieves the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png for the selected clothing.

[0182] 2. The server receives the user's full-body image.

[0183] 3. The server analyzes the clothing image and the user image and extracts key features.

[0184] 4. Generate try-on images using a generative AI model.

[0185] 5. Save the try-on image and generate the URL https: / / example.com / generated_tryon.png.

[0186] 6. The user opens the URL on their device and checks the fitting image.

[0187] Prompt Sentence Examples

[0188] "Choose a red top and blue jeans and generate a natural try-on image based on the user's preferences. We use a GAN model to generate images that combine these clothes with the user's full-body photo."

[0189] This allows users to see a detailed try-on image before purchasing, which increases purchasing motivation and reduces the hassle of returning or exchanging items due to mismatched sizes or designs.

[0190] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0191] Step 1:

[0192] A user accesses an e-commerce site, selects the clothing item they want, and then takes or selects a full-body image from their smartphone or computer and uploads it.

[0193] Input: Clothing selection information (identifier, image URL), full-body image.

[0194] Output: Clothing selection data and full-body image data.

[0195] Specific operation: The user selects clothing through the interface of the e-commerce site and uploads a full-body image of themselves, which is then stored in cloud storage.

[0196] Step 2:

[0197] The server receives the user's clothing selection information (identifier and image URL) and a full-body image.

[0198] Input: Clothing selection data and full-body image data sent by the user.

[0199] Output: Selection data and whole body image stored on the server.

[0200] Specific operation: The server processes the received data and downloads the clothing image based on the identifier and image URL. The full-body image is stored in the database.

[0201] Step 3:

[0202] The server retrieves and analyzes the selected clothing image based on the identifier and the image URL.

[0203] Input: Clothing image URL.

[0204] Output: Feature data of the analyzed clothing image.

[0205] Specific operation: The server downloads the clothing image using the image URL and analyzes the features of the clothing image (e.g., shape, color, design) using OpenCV and Scikit-Image.

[0206] Step 4:

[0207] The server analyzes the user's full-body image and extracts key features of the image.

[0208] Input: A full-body image of the user.

[0209] Output: Feature data of the analyzed whole-body image.

[0210] Specific operation: The server analyzes the whole-body image using an image processing library such as OpenCV and extracts specific features such as contours and facial position.

[0211] Step 5:

[0212] The server uses the analyzed clothing images and feature data from the full-body image as input and generates a fitting image using a generative AI model (GAN).

[0213] Input: Feature data of analyzed clothing images, feature data of whole-body images.

[0214] Output: The generated try-on image.

[0215] Specific operation: The server uses PyTorch and TensorFlow to execute a generative AI model and generate a fitting image in which the selected clothing fits naturally to the user's full-body image.

[0216] Step 6:

[0217] The server saves the generated try-on image and generates a URL that the user can access.

[0218] Input: Generated try-on images.

[0219] Output: URL of the try-on image.

[0220] Specific operation: The fitting images are saved in a database and a URL for accessing them is generated. These URLs are managed in the database.

[0221] Step 7:

[0222] The server sends the generated URL to the user's terminal, and the user checks the fitting image on their own terminal.

[0223] Input: The URL of the generated try-on image.

[0224] Output: Try-on image displayed on the user's device.

[0225] Specific operation: The server sends the generated URL to the user's device, and the user opens the link to view the fitting image. The user interface is displayed in a browser or a dedicated app.

[0226] This allows the user to see a concrete image of the item before purchasing.

[0227] 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.

[0228] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display, and further combines it with an emotion engine to provide more personalized try-on images and recommended clothing images. Specific implementation of this system will be described below.

[0229] User selection of clothing

[0230] First, a user accesses an e-commerce site and selects the items they want from various categories (tops, pants, shoes, etc.). For example, if the user selects "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[0231] Server receives selection information and acquires images

[0232] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[0233] Image analysis and feature extraction

[0234] The server analyzes the acquired image data and extracts the key features of each garment (e.g., collar, sleeve, trouser leg, etc.). The feature information obtained through this analysis plays an important role in the subsequent synthesis process.

[0235] Image synthesis using generative AI models

[0236] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[0237] Use of emotion engine

[0238] Furthermore, the system uses an emotion engine to recognize the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at the try-on images to understand their emotional state. Based on this state, it adjusts the generation of try-on images and clothing recommendations.

[0239] Save composite image and generate URL

[0240] The generated try-on image is saved by the server and a URL is generated so that users can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[0241] Sending and displaying composite images

[0242] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[0243] Display of recommended clothing

[0244] Furthermore, based on the emotional information recognized by the emotion engine, the system provides recommended clothing images according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[0245] Specific examples

[0246] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0247] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0248] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0249] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0250] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0251] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0252] 6. The emotion engine analyzes the user's reaction and adds similar items to the recommendation list based on the generated try-on images.

[0253] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[0254] 8. The server sends this URL to the user's device.

[0255] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[0256] This process allows users to get a concrete image of how the clothes will look before they buy, and also provides personalized recommendations based on their emotions, which increases the success rate of purchases and reduces the hassle of returns.

[0257] The processing flow will be explained below.

[0258] Step 1:

[0259] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[0260] Step 2:

[0261] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0262] Step 3:

[0263] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0264] Step 4:

[0265] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[0266] Step 5:

[0267] The server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates a try-on image of the user's chosen clothing.

[0268] Step 6:

[0269] The emotion engine recognizes the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at images of items to try on, and understands their emotional state.

[0270] Step 7:

[0271] Based on the user's emotions recognized by the emotion engine, the generated try-on images are adjusted to create a more suitable outfit for the user. For example, if the user expresses positive emotions, the system generates images that emphasize that style.

[0272] Step 8:

[0273] The server saves the generated try-on image and generates a URL that users can access. For example, the URL for the generated try-on image might be https: / / example.com / generated_outfit.png.

[0274] Step 9:

[0275] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[0276] Step 10:

[0277] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look when worn and decide whether to purchase it.

[0278] Step 11:

[0279] Based on the emotional information recognized by the emotion engine, recommended clothing images are provided according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[0280] This specific processing step allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions, thereby increasing the success rate of purchases and reducing the hassle of returns.

[0281] Example 2

[0282] 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."

[0283] On modern e-commerce sites, users have limited options for trying on clothing before purchasing. This can lead to an increase in returns and exchanges, as the product may not meet expectations after purchase. Furthermore, the lack of personalized recommendations based on users' emotions and preferences makes it difficult to increase user satisfaction.

[0284] 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.

[0285] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site, means for acquiring the plurality of clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for saving the generated outfit image and generating a URL for displaying it to the user, means for transmitting the URL to the user terminal and enabling display on the user terminal, and means for recognizing the user's reactions and emotions using an emotion engine and adjusting the generation of try-on images or clothing recommendations based on the user's emotional state. This allows the user to see specific try-on images before purchasing and receive personalized recommendations based on their emotions.

[0286] A "user" is someone who uses this system to select clothing on an e-commerce site and check how it looks when worn.

[0287] An "e-commerce site" is a website for selling and purchasing goods over the Internet.

[0288] A "clothing image" is image data that shows a visual representation of the clothing selected by the user.

[0289] An "identifier" is an ID or code that uniquely identifies a particular clothing image.

[0290] An "image URL" is an address that indicates where a specific clothing image is stored on the Internet.

[0291] The "server" is a computer system that receives information sent by the user, acquires and analyzes images, and synthesizes images for trying on.

[0292] "Image analysis" is a process for extracting feature information from acquired clothing images.

[0293] "Features" are information that indicates specific parts, shapes, patterns, etc. in a clothing image.

[0294] A "generative AI model" is an artificial intelligence algorithm that synthesizes new try-on images based on acquired feature information.

[0295] A "try-on image" is a visual image generated by combining multiple pieces of clothing selected by the user.

[0296] A "URL" is an internet address used to access a particular web page or resource.

[0297] An "emotion engine" is an artificial intelligence system that recognizes users' reactions and emotions and analyzes that data.

[0298] "Reactions" refer to the user's actions, facial expressions, and interactions such as clicking and scrolling when looking at the try-on images.

[0299] "Recommendation" is the process of selecting relevant clothing items based on a user's preferences.

[0300] This system receives the identifier and image URL of a clothing image selected by a user from multiple clothing images selected on an e-commerce site, and generates try-on images through acquisition, analysis, synthesis, and display. Furthermore, by combining an emotion engine, it provides more personalized try-on images and recommended clothing images.

[0301] Hardware and software used

[0302] The server is a computer system with high-performance computing capabilities for receiving data, acquiring images, analyzing them, storing data, and running the generated AI model. For specific implementation, it is recommended to use deep learning libraries such as TensorFlow or PyTorch, and OpenCV for image analysis.

[0303] The user's terminal is typically a smartphone, tablet, or personal computer that accesses the e-commerce site via a web browser.

[0304] An emotion engine requires an algorithm to analyze user reactions and emotions and predict purchasing behavior. Specifically, an analytical model using machine learning is considered.

[0305] Program processing

[0306] A user accesses an e-commerce site and selects the clothing item they want. For example, if the user selects a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server along with an identifier and an image URL.

[0307] The server retrieves clothing images from the specified URLs based on the identifier (ID) and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads images from these URLs.

[0308] The acquired images are analyzed by the server to extract key features using image processing libraries such as OpenCV. After the feature information is extracted, the server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize a try-on image based on the feature information.

[0309] The generated try-on image is saved by the server and a URL that can be accessed by the user is generated. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png, and the server generates this URL and saves it in the database.

[0310] The server then sends the generated URL to the user's device. When the user opens the URL, an image of the new outfit is displayed. The user can then review the image and decide whether to actually purchase it.

[0311] Furthermore, the server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, the server can understand the user's emotional state from their facial expressions when looking at the try-on images, as well as their actions such as clicking and scrolling, and adjust the generation of try-on images or clothing recommendations based on this.

[0312] Specific examples

[0313] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0314] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0315] 2. The device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0316] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0317] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0318] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0319] 6. The emotion engine analyzes the user's reactions and adds similar items to the recommendation list based on the generated try-on images.

[0320] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[0321] 8. The server sends this URL to the user's device.

[0322] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[0323] Using this system, users can easily get a concrete try-on image before purchasing and receive personalized recommendations based on their emotions.

[0324] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0325] Step 1:

[0326] A user accesses an e-commerce site and selects the clothing they want. For example, they select a "red top" and "blue jeans," and then the information (identifier and image URL) is sent from the terminal to the server. The input is the user's selection information, and the output is the clothing identifier and image URL sent to the server.

[0327] Step 2:

[0328] The server retrieves clothing images from the specified URLs based on the identifier and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads image files from these URLs. The input is the identifier and image URL sent in the previous step, and the output is the downloaded clothing image.

[0329] Step 3:

[0330] The server analyzes the acquired image data and extracts key clothing features. This analysis process uses image processing libraries such as OpenCV. For example, it identifies features such as the collar and sleeves of a red top, or the hem and waist of blue jeans. The input is the downloaded clothing image, and the output is the analyzed feature information.

[0331] Step 4:

[0332] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. For example, it generates a try-on image of a red top paired with blue jeans. This process uses deep learning libraries such as TensorFlow and PyTorch. The input is the analyzed feature information, and the output is the generated try-on image.

[0333] Step 5:

[0334] The server saves the generated try-on image and generates a URL to make the image accessible. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit.png. The input is the generated try-on image, and the output is the URL of the generated try-on image.

[0335] Step 6:

[0336] The server sends the generated URL to the user's device. When the user accesses this URL on their device, the generated try-on image is displayed. For example, if the user accesses https: / / example.com / generated_outfit.png in their web browser, they can see the synthesized try-on image. The input is the generated URL, and the output is the try-on image displayed on the user's device.

[0337] Step 7:

[0338] The server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, it collects and analyzes behavioral data such as facial expressions, clicks, and scrolls when the user looks at images of items to try on. The input is the user's reaction data, and the output is analyzed emotional information.

[0339] Step 8:

[0340] The server generates try-on images and adjusts clothing recommendations based on the analyzed emotional information. For example, if the user responds positively, clothing of a similar style is added to the recommendation list. The input is the analyzed emotional information, and the output is the adjusted try-on images and recommendation list.

[0341] This system allows users to see specific try-on images and receive personalized recommendations based on their emotions.

[0342] (Application example 2)

[0343] 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."

[0344] Conventional e-commerce sites lack an effective way to provide users with try-on images of the clothing they select. This leads to users purchasing clothing without visual uncertainty, resulting in a high return rate. Furthermore, the lack of personalized clothing recommendations based on users' emotions and preferences limits the user experience. This creates a need for a system that can facilitate purchasing decisions.

[0345] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0346] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on an e-commerce site, means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for recognizing the user's emotional state using an emotion engine and adjusting the generation of try-on images and clothing recommendations based on that state, means for saving the generated outfit images and generating a URL for displaying them to the user, and means for transmitting the URL to a user terminal so that the URL can be displayed on the user terminal. This allows the user to visually check the try-on images and receive personalized recommendations based on the emotion analysis results.

[0347] "User" refers to a customer or consumer who uses the e-commerce site.

[0348] "E-commerce site" means a website where purchases of goods and services are made over the Internet.

[0349] "Garment image" refers to a digital image that shows the visual characteristics of a garment.

[0350] "Identifier" refers to an ID or code that uniquely identifies each clothing image.

[0351] "Image URL" refers to the web address where the clothing image is stored on the Internet.

[0352] "Acquisition" refers to the act of downloading clothing images and related information from an e-commerce site or server.

[0353] "Analysis" refers to the procedure of detecting and extracting features and patterns within clothing images.

[0354] "Generative AI model" refers to an artificial intelligence (AI) algorithm for generating new try-on images from image data.

[0355] "Synthesis" refers to the act of combining multiple clothing images to create a new try-on image.

[0356] "Try-on image" refers to a visual display of a user virtually trying on the clothing combination they have selected.

[0357] An "emotion engine" refers to technology for detecting and analyzing a user's emotions and reactions.

[0358] "Saving" refers to the act of recording the generated try-on image as data on a server or the like.

[0359] "URL generation" refers to the process of creating a web address that allows users to access the generated try-on images.

[0360] "Sending" refers to the act of transferring the URL of the generated try-on image to the user's device.

[0361] "Display" refers to the act of visually presenting try-on images and recommended clothing images generated on the user terminal.

[0362] "Personalization" refers to the act of providing individually tailored services and recommendations based on a user's preferences and emotions.

[0363] "Recommended Clothing Images" refers to additional clothing images that are recommended based on the user's reactions or emotional state.

[0364] The present invention is a system that generates try-on images from multiple clothing images selected by a user on an e-commerce site and uses an emotion engine to provide the user with personalized try-on images and clothing recommendations. A specific embodiment of the present invention will be described below.

[0365] First, a user visits an e-commerce site and selects the desired clothing item from various categories (tops, pants, shoes, etc.). The identifier and image URL of the selected clothing item are sent from the user's device to the server. This information is sent along with the identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, both of these URLs are sent to the server.

[0366] Next, the server retrieves clothing images from the specified URL based on the received identifier and image URL. The server downloads these image data, for example, image data of a red top and blue jeans. After this, the server analyzes the acquired image data and extracts the main features of each garment (collar, sleeves, hem, etc.). The analyzed feature information plays an important role in the subsequent synthesis process.

[0367] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize each clothing image based on the analyzed feature information. This generates an image of how the user would actually try on the clothing they selected. For example, an image of a natural combination of a red top and blue jeans can be generated.

[0368] Furthermore, an emotion engine is used to analyze the user's reactions and emotions. The system analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when viewing the try-on images to understand their emotional state. Based on these results, the system adjusts the generation of try-on images and clothing recommendations.

[0369] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL for the generated try-on image is created as https: / / example.com / generated_outfit.png. The server sends this URL to the user's device, and the user can check the new outfit image on their device.

[0370] Based on the emotional information recognized by the emotion engine, recommended clothing images are also provided according to the user's preferences. For example, if the user has a positive reaction to a fitting image, clothing of a similar style will be displayed in the recommendation list.

[0371] As a concrete example, consider a user who selects a new denim jacket and black jeans. When the user selects "denim jacket" and "black jeans" on an e-commerce website, their respective IDs, jacket123 and jeans456, and their image URLs, https: / / example.com / jacket123.png and https: / / example.com / jeans456.png, are sent to the server. The server downloads the image data of the denim jacket and black jeans from these URLs, analyzes them, and synthesizes them using a generative AI model to generate new try-on images. The emotion engine analyzes the user's reaction and adds similar items to a recommendation list based on the generated try-on images. The server saves the URLs containing the generated try-on images and recommendation list and sends them to the user's device. The user can view the new outfit images and recommendation list.

[0372] An example of a prompt sentence would be a specific instruction input to the generative AI model, such as "Generate an image of someone trying on a denim jacket and black jeans."

[0373] This allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions.

[0374] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0375] Step 1:

[0376] A user accesses an e-commerce site and selects a desired clothing item. The user then sends data including the identifier and image URL of the selected clothing item from the terminal to the server. The input includes the identifiers and image URLs of multiple clothing images, and by sending these to the server, information on the clothing selected by the user is collected.

[0377] Step 2:

[0378] The server retrieves clothing images from the specified URLs based on the received identifier and image URL. The server accesses each URL and downloads clothing image data. The identifier and image URL are input, and the required clothing images are prepared by retrieving image data based on these.

[0379] Step 3:

[0380] The server analyzes the acquired clothing images and extracts the key features of each garment (collar, sleeves, hem, etc.). The server uses an image analysis algorithm to identify the features of each clothing image. The downloaded image data is used as input and analysis is performed to obtain feature information.

[0381] Step 4:

[0382] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. It then generates an image of what it would look like if the user tried on the selected clothing. In this process, a specific prompt (e.g., "Generate an image of me trying on a denim jacket and black jeans") is input to the model. The input is the analyzed feature information and the prompt, and the generated try-on image is output.

[0383] Step 5:

[0384] The emotion engine analyzes the user's reactions and emotions when viewing images of clothing to try on. The server collects the user's facial expressions and actions (clicks, scrolls, etc.) through cameras and interaction data to recognize the user's emotional state. The input is the user's reaction data, and the output is their emotional state.

[0385] Step 6:

[0386] The server saves the generated try-on images and generates a URL to display them to the user. The server saves the generated outfit image data and creates a corresponding access URL. The input is the generated try-on image, and the output is the save destination URL.

[0387] Step 7:

[0388] The server sends the URL to the user's device, enabling it to be displayed on the user's device. The user uses the received URL to check the new outfit image. The input is the destination URL, and the output is the display on the user's device.

[0389] Step 8:

[0390] Based on the emotion information recognized by the emotion engine, the server, which provides recommended clothing images according to the user's preferences, creates a recommendation list and generates a URL for the recommended products so that the user can access them. The user can refer to the recommendation list to check related items. The input is emotion information and a recommendation algorithm, and the output is a recommendation list and its URL.

[0391] The above are the specific processing steps for processing from the user terminal to the server, generating try-on images, analyzing emotions, and the recommendation system.

[0392] 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.

[0393] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0394] 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.

[0395] [Second embodiment]

[0396] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0397] 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.

[0398] 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).

[0399] 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.

[0400] 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.

[0401] 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).

[0402] 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.

[0403] 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.

[0404] 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.

[0405] 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.

[0406] In the smart glasses 214, 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.

[0407] 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."

[0408] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[0409] User selection of clothing

[0410] First, a user accesses an e-commerce site and selects the clothing item they want from various categories (tops, pants, shoes, etc.). If the user selects, for example, a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[0411] Server receives selection information and acquires images

[0412] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[0413] Image analysis and feature extraction

[0414] The server analyzes the acquired image data and extracts key features of the clothing (e.g., collar, sleeves, trouser legs, etc.). The feature information obtained from this analysis plays an important role in the subsequent synthesis process.

[0415] Image synthesis using generative AI models

[0416] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[0417] Save composite image and generate URL

[0418] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[0419] Sending and displaying composite images

[0420] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[0421] Specific examples

[0422] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0423] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0424] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0425] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0426] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0427] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0428] 6. The server saves the generated try-on image as https: / / example.com / generated_outfit123456.png.

[0429] 7. The server sends this URL to the user's device.

[0430] 8. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user can check the new outfit image.

[0431] This series of processes allows users to get a concrete image of how the clothing will look before purchasing, reducing the risk of making a wrong purchase or having to return the item.

[0432] The processing flow will be explained below.

[0433] Step 1:

[0434] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[0435] Step 2:

[0436] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0437] Step 3:

[0438] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0439] Step 4:

[0440] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[0441] Step 5:

[0442] The server runs a generative AI model, such as a generative adversarial network (GAN), to synthesize clothing images based on the analyzed feature information. This process generates a try-on image combining the user's chosen clothing.

[0443] Step 6:

[0444] The server saves the generated try-on image and generates a URL that can be accessed by the user. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit123456.png.

[0445] Step 7:

[0446] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[0447] Step 8:

[0448] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look in real life and decide whether to purchase it.

[0449] Through this series of processes, users can get a concrete idea of ​​how the clothing will look before purchasing, reducing the risk of making a mistake in purchasing or the hassle of having to return the item.

[0450] Example 1

[0451] 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."

[0452] In conventional e-commerce systems, when users purchase clothing, it is difficult for them to see how it will look when worn without actually trying it on. Even if users select the clothing combinations they want, the combinations are not displayed in a natural arrangement, which reduces post-purchase satisfaction and increases the hassle of returning items. Furthermore, there was a lack of a way to effectively manage and save each user's outfit images, making it difficult to meet individual needs.

[0453] 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.

[0454] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on the e-commerce site; means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for saving the generated outfit image and generating a URL for displaying it to the user; means for transmitting the URL to the user terminal and enabling display on the user terminal; and means for inputting a desired clothing combination by the user and generating a prompt message for the generative AI model to generate a try-on image based on the input. This allows users to see a specific and natural look before purchasing, reducing the hassle of purchasing mistakes and returns. Furthermore, it is possible to effectively manage and store the outfit images generated for each user, thereby providing services tailored to individual needs.

[0455] "User" refers to an individual or group that selects clothing on an e-commerce site and requests the generation of try-on images.

[0456] "E-commerce site" refers to a website that sells clothing and other products over the Internet.

[0457] "Identifier" refers to an ID or tag that uniquely identifies a specific clothing image.

[0458] "Image URL" refers to a Uniform Resource Locator (URL) for accessing a clothing image stored on the Internet.

[0459] "Server" refers to a computer system that receives and processes requests from users.

[0460] "Features" refer to key parts or characteristics of the clothing in the image, such as the collar, sleeves, and hem.

[0461] "Generative AI model" refers to an artificial intelligence model that generates try-on images based on the characteristics of clothing selected by the user. An example is a generative adversarial network (GAN).

[0462] A "try-on image" refers to an image created by combining multiple pieces of clothing selected by the user, creating a composite image that looks as if the clothing is actually being worn.

[0463] "Saving" refers to recording the generated try-on images in a database or storage device.

[0464] "URL generation" refers to creating a uniform resource locator (URL) for accessing the saved try-on images.

[0465] "Prompt sentence" refers to an instruction sentence provided as input to a generative AI model to generate try-on images.

[0466] "User terminal" refers to a device such as a computer, smartphone, or tablet that a user uses to access an e-commerce site, make selections, and check images of items being tried on.

[0467] The present invention provides a system that receives the identifiers and image URLs of selected clothing images from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[0468] User selection of clothing

[0469] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user selects the clothing item they want from various categories (tops, pants, shoes, etc.). For example, if a user selects a "red top" and "blue jeans," this selection information is sent from the user's device to the server.

[0470] Server receives selection information and acquires images

[0471] The server retrieves clothing images from the specified URL based on the identifier (ID) and image URL received from the user's device. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the image data will be downloaded from these URLs. Specifically, the server retrieves images using a client library such as curl or requests.

[0472] Image analysis and feature extraction

[0473] The server analyzes the acquired image data using image analysis software such as OpenCV and TensorFlow to extract key features of the clothing (collar, sleeves, trouser hems, etc.) The feature information obtained from this analysis plays an important role in the subsequent image synthesis process.

[0474] Image synthesis using generative AI models

[0475] The server uses a generative AI model such as a generative adversarial network (GAN) to synthesize clothing images based on the extracted feature information. An image of what the clothing selected by the user would look like when actually worn is generated. For example, an outfit image of a red top and blue jeans naturally combined can be created. To facilitate this process, a prompt is used as input to the generative AI model. An example of a prompt is "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate an image of the clothing being tried on."

[0476] Save composite image and generate URL

[0477] The generated try-on image is saved to a storage device by the server. The server generates a URL so that the saved try-on image can be accessed. For example, the URL of the generated try-on image is configured as https: / / example.com / generated_outfit.png.

[0478] Sending and displaying composite images

[0479] The server sends the generated URL to the user's device. The user can open this URL on their device to view the new outfit image. For example, the received URL https: / / example.com / generated_outfit.png can be displayed in a web browser or dedicated application.

[0480] Specific examples

[0481] For example, if a user selects a denim jacket and black jeans, the following happens:

[0482] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0483] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0484] 3. The server accesses the sent URL and downloads the image.

[0485] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0486] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0487] 6. The server saves the generated try-on image in storage and generates a URL.

[0488] 7. The server sends the generated URL to the user's device.

[0489] 8. The device displays the received URL, and the user can check the new outfit image.

[0490] This series of processes allows users to get a concrete image of how the garment will look before purchasing, reducing the risk of buying something wrong or having to return it.

[0491] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0492] Step 1: User selects clothing

[0493] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user navigates through categories (tops, pants, shoes, etc.) on the site and selects the desired clothing item. The input is the identifier (ID) and image URL of the selected clothing item. For example, if the user selects "red top" and "blue jeans," the output is the identifier and image URL of the selected clothing item. Specifically, when the user confirms their selection, the selection information is sent from the user's device to the server.

[0494] Step 2: Server receives selection information

[0495] The server receives the identifier (ID) and image URL sent from the user's device. The input is the identifier and image URL sent from the user's device, and the output is to store this information as received data. Specifically, the server receives the HTTPS request and saves the selected information in a database.

[0496] Step 3: Acquiring image data

[0497] The server downloads the specified image data based on the URL received. The input is an identifier and an image URL, and the output is the retrieved image data. Specifically, the server accesses the image URL using a client library such as curl or requests and saves the image data in the server's storage. For example, download the image data of a red top from the URL https: / / example.com / red_top.png.

[0498] Step 4: Image analysis and feature extraction

[0499] The server analyzes the acquired image data and extracts the key features of the clothing. The input is the acquired image data, and the output is the extracted feature information. Specifically, the server uses image analysis software such as OpenCV or TensorFlow to identify features such as the collar, sleeves, and trouser hems of the clothing. For example, the analysis results in identifying the position of the collar and the length of the sleeves of a red top.

[0500] Step 5: Image synthesis using generative AI models

[0501] The server uses the analyzed feature information as input and generates a try-on image using a generative AI model (e.g., a generative adversarial network (GAN)). The input is the extracted feature information, and the output is the generated try-on image. Specifically, the server converts the feature information into a prompt and inputs it into the generative AI model. For example, the server generates a prompt such as, "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate a try-on image." Based on this, the generative AI model generates an image that combines a red top and blue jeans.

[0502] Step 6: Save the composite image and generate a URL

[0503] The server saves the generated try-on image to a storage device and generates an accessible URL. The input is the generated try-on image, and the output is the URL of the saved try-on image. Specifically, the server saves the try-on image as a file and generates a unique URL for that file. For example, the generated URL might be https: / / example.com / generated_outfit.png.

[0504] Step 7: Send and view the composite image

[0505] The server sends the generated URL to the user's device, allowing the user to view the new outfit image on the device. The input is the generated URL, and the output is the try-on image displayed on the user's device. Specifically, the server returns a URL as an API response, and the user's device receives and displays this URL. For example, the user opens the URL https: / / example.com / generated_outfit.png in a browser, and the new outfit image is displayed.

[0506] (Application example 1)

[0507] 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."

[0508] On conventional e-commerce sites, users could not try on clothes before purchasing, making it difficult to confirm whether the item actually fit them. As a result, users had to go through the trouble of returning or exchanging the item after purchase if the size or design did not fit. In addition, it was difficult for users to visualize what the selected clothing would look like in real life, which led to a decrease in purchasing motivation.

[0509] 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.

[0510] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site; means for acquiring the plurality of clothing images based on the identifier and the image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for the user to upload a full-body image of themselves and analyze the acquired clothing images and the user's full-body image; means for executing a generative AI model that generates a try-on image in which the selected clothing naturally fits the user's full-body image based on the analyzed characteristics; means for saving the generated try-on image and generating a URL for displaying it to the user; and means for transmitting the URL to the user terminal and enabling display on the user terminal. This allows the user to specifically check the try-on image based on their full-body image, allowing them to have a detailed image before purchasing, which increases their motivation to purchase and reduces the hassle of returns and exchanges due to mismatched sizes or designs.

[0511] A "user" is a consumer who uses an e-commerce site to select and purchase products.

[0512] An "e-commerce site" is a website that sells and buys goods and services over the Internet.

[0513] "Clothing Images" are digital images of clothing displayed on e-commerce sites.

[0514] An "identifier" is information for uniquely identifying a specific clothing image.

[0515] An "image URL" is an address (URL) for accessing a specific image file on the Internet.

[0516] A "full-body image" is a photograph showing the user's entire body.

[0517] "Feature analysis" is a technique for extracting specific elements or patterns from an image.

[0518] A "generative AI model" is an algorithm that uses artificial intelligence techniques to generate new images or data, such as generative adversarial networks (GANs).

[0519] The "try-on image" is a composite image of the user's whole body and a clothing image, simulating the appearance of the clothing when the user actually tries it on.

[0520] "Saving" means to hold the generated try-on image as data.

[0521] "URL generation" means creating a new URL to access the saved try-on image.

[0522] A "user terminal" is a device used by a user to access an e-commerce site, and includes a smartphone, tablet, computer, etc.

[0523] "Transmission" means sending data or information from the server to the user terminal.

[0524] "Display" refers to the visual presentation of information or images on the screen of a user terminal.

[0525] The present invention is a system that generates try-on images based on multiple clothing images selected by a user on an e-commerce site. The following describes how to implement this system step by step.

[0526] System Overview

[0527] The system consists of a server that runs on the backend of an e-commerce site and a terminal used by users. Users upload a full-body image of themselves and select clothing images on the site. The server analyzes this data and uses a generative AI model to generate try-on images and provide them to the user.

[0528] Hardware and software used

[0529] Smartphones, tablets, and computers: devices used by users to access

[0530] Server: Receives, stores, analyzes, and synthesizes data. Specifically, it uses cloud services such as AWS and Google Cloud Platform.

[0531] Generative AI models: Use generative adversarial networks (GANs) using PyTorch and TensorFlow.

[0532] Image analysis software: For example, OpenCV or Scikit-Image can be used to extract image features.

[0533] Data Flow and Processing

[0534] 1. User Action:

[0535] Users access an e-commerce site, select clothing items, and upload a full-body image of themselves.

[0536] 2. Data reception and processing by the server:

[0537] The server receives the identifier and image URL of the garment selected by the user, as well as a full-body image of the user.

[0538] 3. Image Acquisition and Feature Analysis:

[0539] The server retrieves the clothing image based on the identifier and the image URL, and analyzes the characteristics of the clothing and the user's full-body image using image analysis software.

[0540] 4. Generate fitting images:

[0541] Using a generative AI model (e.g., GAN using PyTorch or TensorFlow), a fitting image is generated that naturally combines the selected clothing with the user's full-body image based on the analyzed features.

[0542] 5. Save the generated image and generate a URL:

[0543] The generated try-on images are stored on the server, and a URL is generated that the user can access.

[0544] 6. User Notification and Display:

[0545] The server sends the generated URL to the user's device, and the user can open the URL to check the try-on image.

[0546] Examples of concrete examples and prompts

[0547] Specific examples

[0548] For example, if a user selects a "red top" and "blue jeans" and uploads a full-body image of themselves, the server will process the following:

[0549] 1. The server retrieves the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png for the selected clothing.

[0550] 2. The server receives the user's full-body image.

[0551] 3. The server analyzes the clothing image and the user image and extracts key features.

[0552] 4. Generate try-on images using a generative AI model.

[0553] 5. Save the try-on image and generate the URL https: / / example.com / generated_tryon.png.

[0554] 6. The user opens the URL on their device and checks the fitting image.

[0555] Prompt Sentence Examples

[0556] "Choose a red top and blue jeans and generate a natural try-on image based on the user's preferences. We use a GAN model to generate images that combine these clothes with the user's full-body photo."

[0557] This allows users to see a detailed try-on image before purchasing, which increases purchasing motivation and reduces the hassle of returning or exchanging items due to mismatched sizes or designs.

[0558] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0559] Step 1:

[0560] A user accesses an e-commerce site, selects the clothing item they want, and then takes or selects a full-body image from their smartphone or computer and uploads it.

[0561] Input: Clothing selection information (identifier, image URL), full-body image.

[0562] Output: Clothing selection data and full-body image data.

[0563] Specific operation: The user selects clothing through the interface of the e-commerce site and uploads a full-body image of themselves, which is then stored in cloud storage.

[0564] Step 2:

[0565] The server receives the user's clothing selection information (identifier and image URL) and a full-body image.

[0566] Input: Clothing selection data and full-body image data sent by the user.

[0567] Output: Selection data and whole body image stored on the server.

[0568] Specific operation: The server processes the received data and downloads the clothing image based on the identifier and image URL. The full-body image is stored in the database.

[0569] Step 3:

[0570] The server retrieves and analyzes the selected clothing image based on the identifier and the image URL.

[0571] Input: Clothing image URL.

[0572] Output: Feature data of the analyzed clothing image.

[0573] Specific operation: The server downloads the clothing image using the image URL and analyzes the features of the clothing image (e.g., shape, color, design) using OpenCV and Scikit-Image.

[0574] Step 4:

[0575] The server analyzes the user's full-body image and extracts key features of the image.

[0576] Input: A full-body image of the user.

[0577] Output: Feature data of the analyzed whole-body image.

[0578] Specific operation: The server analyzes the whole-body image using an image processing library such as OpenCV and extracts specific features such as contours and facial position.

[0579] Step 5:

[0580] The server uses the analyzed clothing images and feature data from the full-body image as input and generates a fitting image using a generative AI model (GAN).

[0581] Input: Feature data of analyzed clothing images, feature data of whole-body images.

[0582] Output: The generated try-on image.

[0583] Specific operation: The server uses PyTorch and TensorFlow to execute a generative AI model and generate a fitting image in which the selected clothing fits naturally to the user's full-body image.

[0584] Step 6:

[0585] The server saves the generated try-on image and generates a URL that the user can access.

[0586] Input: Generated try-on images.

[0587] Output: URL of the try-on image.

[0588] Specific operation: The fitting images are saved in a database and a URL for accessing them is generated. These URLs are managed in the database.

[0589] Step 7:

[0590] The server sends the generated URL to the user's terminal, and the user checks the fitting image on their own terminal.

[0591] Input: The URL of the generated try-on image.

[0592] Output: Try-on image displayed on the user's device.

[0593] Specific operation: The server sends the generated URL to the user's device, and the user opens the link to view the fitting image. The user interface is displayed in a browser or a dedicated app.

[0594] This allows the user to see a concrete image of the item before purchasing.

[0595] 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.

[0596] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display, and further combines it with an emotion engine to provide more personalized try-on images and recommended clothing images. Specific implementation of this system will be described below.

[0597] User selection of clothing

[0598] First, a user accesses an e-commerce site and selects the items they want from various categories (tops, pants, shoes, etc.). For example, if the user selects "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[0599] Server receives selection information and acquires images

[0600] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[0601] Image analysis and feature extraction

[0602] The server analyzes the acquired image data and extracts the key features of each garment (e.g., collar, sleeve, trouser leg, etc.). The feature information obtained through this analysis plays an important role in the subsequent synthesis process.

[0603] Image synthesis using generative AI models

[0604] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[0605] Use of emotion engine

[0606] Furthermore, the system uses an emotion engine to recognize the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at the try-on images to understand their emotional state. Based on this state, it adjusts the generation of try-on images and clothing recommendations.

[0607] Save composite image and generate URL

[0608] The generated try-on image is saved by the server and a URL is generated so that users can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[0609] Sending and displaying composite images

[0610] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[0611] Display of recommended clothing

[0612] Furthermore, based on the emotional information recognized by the emotion engine, the system provides recommended clothing images according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[0613] Specific examples

[0614] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0615] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0616] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0617] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0618] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0619] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0620] 6. The emotion engine analyzes the user's reaction and adds similar items to the recommendation list based on the generated try-on images.

[0621] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[0622] 8. The server sends this URL to the user's device.

[0623] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[0624] This process allows users to get a concrete image of how the clothes will look before they buy, and also provides personalized recommendations based on their emotions, which increases the success rate of purchases and reduces the hassle of returns.

[0625] The processing flow will be explained below.

[0626] Step 1:

[0627] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[0628] Step 2:

[0629] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0630] Step 3:

[0631] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0632] Step 4:

[0633] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[0634] Step 5:

[0635] The server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates a try-on image of the user's chosen clothing.

[0636] Step 6:

[0637] The emotion engine recognizes the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at images of items to try on, and understands their emotional state.

[0638] Step 7:

[0639] Based on the user's emotions recognized by the emotion engine, the generated try-on images are adjusted to create a more suitable outfit for the user. For example, if the user expresses positive emotions, the system generates images that emphasize that style.

[0640] Step 8:

[0641] The server saves the generated try-on image and generates a URL that users can access. For example, the URL for the generated try-on image might be https: / / example.com / generated_outfit.png.

[0642] Step 9:

[0643] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[0644] Step 10:

[0645] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look when worn and decide whether to purchase it.

[0646] Step 11:

[0647] Based on the emotional information recognized by the emotion engine, recommended clothing images are provided according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[0648] This specific processing step allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions, thereby increasing the success rate of purchases and reducing the hassle of returns.

[0649] Example 2

[0650] 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."

[0651] On modern e-commerce sites, users have limited options for trying on clothing before purchasing. This can lead to an increase in returns and exchanges, as the product may not meet expectations after purchase. Furthermore, the lack of personalized recommendations based on users' emotions and preferences makes it difficult to increase user satisfaction.

[0652] 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.

[0653] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site, means for acquiring the plurality of clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for saving the generated outfit image and generating a URL for displaying it to the user, means for transmitting the URL to the user terminal and enabling display on the user terminal, and means for recognizing the user's reactions and emotions using an emotion engine and adjusting the generation of try-on images or clothing recommendations based on the user's emotional state. This allows the user to see specific try-on images before purchasing and receive personalized recommendations based on their emotions.

[0654] A "user" is someone who uses this system to select clothing on an e-commerce site and check how it looks when worn.

[0655] An "e-commerce site" is a website for selling and purchasing goods over the Internet.

[0656] A "clothing image" is image data that shows a visual representation of the clothing selected by the user.

[0657] An "identifier" is an ID or code that uniquely identifies a particular clothing image.

[0658] An "image URL" is an address that indicates where a specific clothing image is stored on the Internet.

[0659] The "server" is a computer system that receives information sent by the user, acquires and analyzes images, and synthesizes images for trying on.

[0660] "Image analysis" is a process for extracting feature information from acquired clothing images.

[0661] "Features" are information that indicates specific parts, shapes, patterns, etc. in a clothing image.

[0662] A "generative AI model" is an artificial intelligence algorithm that synthesizes new try-on images based on acquired feature information.

[0663] A "try-on image" is a visual image generated by combining multiple pieces of clothing selected by the user.

[0664] A "URL" is an internet address used to access a particular web page or resource.

[0665] An "emotion engine" is an artificial intelligence system that recognizes users' reactions and emotions and analyzes that data.

[0666] "Reactions" refer to the user's actions, facial expressions, and interactions such as clicking and scrolling when looking at the try-on images.

[0667] "Recommendation" is the process of selecting relevant clothing items based on a user's preferences.

[0668] This system receives the identifier and image URL of a clothing image selected by a user from multiple clothing images selected on an e-commerce site, and generates try-on images through acquisition, analysis, synthesis, and display. Furthermore, by combining an emotion engine, it provides more personalized try-on images and recommended clothing images.

[0669] Hardware and software used

[0670] The server is a computer system with high-performance computing capabilities for receiving data, acquiring images, analyzing them, storing data, and running the generated AI model. For specific implementation, it is recommended to use deep learning libraries such as TensorFlow or PyTorch, and OpenCV for image analysis.

[0671] The user's terminal is typically a smartphone, tablet, or personal computer that accesses the e-commerce site via a web browser.

[0672] An emotion engine requires an algorithm to analyze user reactions and emotions and predict purchasing behavior. Specifically, an analytical model using machine learning is considered.

[0673] Program processing

[0674] A user accesses an e-commerce site and selects the clothing item they want. For example, if the user selects a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server along with an identifier and an image URL.

[0675] The server retrieves clothing images from the specified URLs based on the identifier (ID) and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads images from these URLs.

[0676] The acquired images are analyzed by the server to extract key features using image processing libraries such as OpenCV. After the feature information is extracted, the server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize a try-on image based on the feature information.

[0677] The generated try-on image is saved by the server and a URL that can be accessed by the user is generated. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png, and the server generates this URL and saves it in the database.

[0678] The server then sends the generated URL to the user's device. When the user opens the URL, an image of the new outfit is displayed. The user can then review the image and decide whether to actually purchase it.

[0679] Furthermore, the server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, the server can understand the user's emotional state from their facial expressions when looking at the try-on images, as well as their actions such as clicking and scrolling, and adjust the generation of try-on images or clothing recommendations based on this.

[0680] Specific examples

[0681] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0682] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0683] 2. The device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0684] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0685] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0686] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0687] 6. The emotion engine analyzes the user's reactions and adds similar items to the recommendation list based on the generated try-on images.

[0688] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[0689] 8. The server sends this URL to the user's device.

[0690] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[0691] Using this system, users can easily get a concrete try-on image before purchasing and receive personalized recommendations based on their emotions.

[0692] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0693] Step 1:

[0694] A user accesses an e-commerce site and selects the clothing they want. For example, they select a "red top" and "blue jeans," and then the information (identifier and image URL) is sent from the terminal to the server. The input is the user's selection information, and the output is the clothing identifier and image URL sent to the server.

[0695] Step 2:

[0696] The server retrieves clothing images from the specified URLs based on the identifier and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads image files from these URLs. The input is the identifier and image URL sent in the previous step, and the output is the downloaded clothing image.

[0697] Step 3:

[0698] The server analyzes the acquired image data and extracts key clothing features. This analysis process uses image processing libraries such as OpenCV. For example, it identifies features such as the collar and sleeves of a red top, or the hem and waist of blue jeans. The input is the downloaded clothing image, and the output is the analyzed feature information.

[0699] Step 4:

[0700] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. For example, it generates a try-on image of a red top paired with blue jeans. This process uses deep learning libraries such as TensorFlow and PyTorch. The input is the analyzed feature information, and the output is the generated try-on image.

[0701] Step 5:

[0702] The server saves the generated try-on image and generates a URL to make the image accessible. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit.png. The input is the generated try-on image, and the output is the URL of the generated try-on image.

[0703] Step 6:

[0704] The server sends the generated URL to the user's device. When the user accesses this URL on their device, the generated try-on image is displayed. For example, if the user accesses https: / / example.com / generated_outfit.png in their web browser, they can see the synthesized try-on image. The input is the generated URL, and the output is the try-on image displayed on the user's device.

[0705] Step 7:

[0706] The server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, it collects and analyzes behavioral data such as facial expressions, clicks, and scrolls when the user looks at images of items to try on. The input is the user's reaction data, and the output is analyzed emotional information.

[0707] Step 8:

[0708] The server generates try-on images and adjusts clothing recommendations based on the analyzed emotional information. For example, if the user responds positively, clothing of a similar style is added to the recommendation list. The input is the analyzed emotional information, and the output is the adjusted try-on images and recommendation list.

[0709] This system allows users to see specific try-on images and receive personalized recommendations based on their emotions.

[0710] (Application example 2)

[0711] 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."

[0712] Conventional e-commerce sites lack an effective way to provide users with try-on images of the clothing they select. This leads to users purchasing clothing without visual uncertainty, resulting in a high return rate. Furthermore, the lack of personalized clothing recommendations based on users' emotions and preferences limits the user experience. This creates a need for a system that can facilitate purchasing decisions.

[0713] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0714] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on an e-commerce site, means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for recognizing the user's emotional state using an emotion engine and adjusting the generation of try-on images and clothing recommendations based on that state, means for saving the generated outfit images and generating a URL for displaying them to the user, and means for transmitting the URL to a user terminal so that the URL can be displayed on the user terminal. This allows the user to visually check the try-on images and receive personalized recommendations based on the emotion analysis results.

[0715] "User" refers to a customer or consumer who uses the e-commerce site.

[0716] "E-commerce site" means a website where purchases of goods and services are made over the Internet.

[0717] "Garment image" refers to a digital image that shows the visual characteristics of a garment.

[0718] "Identifier" refers to an ID or code that uniquely identifies each clothing image.

[0719] "Image URL" refers to the web address where the clothing image is stored on the Internet.

[0720] "Acquisition" refers to the act of downloading clothing images and related information from an e-commerce site or server.

[0721] "Analysis" refers to the procedure of detecting and extracting features and patterns within clothing images.

[0722] "Generative AI model" refers to an artificial intelligence (AI) algorithm for generating new try-on images from image data.

[0723] "Synthesis" refers to the act of combining multiple clothing images to create a new try-on image.

[0724] "Try-on image" refers to a visual display of a user virtually trying on the clothing combination they have selected.

[0725] An "emotion engine" refers to technology for detecting and analyzing a user's emotions and reactions.

[0726] "Saving" refers to the act of recording the generated try-on image as data on a server or the like.

[0727] "URL generation" refers to the process of creating a web address that allows users to access the generated try-on images.

[0728] "Sending" refers to the act of transferring the URL of the generated try-on image to the user's device.

[0729] "Display" refers to the act of visually presenting try-on images and recommended clothing images generated on the user terminal.

[0730] "Personalization" refers to the act of providing individually tailored services and recommendations based on a user's preferences and emotions.

[0731] "Recommended Clothing Images" refers to additional clothing images that are recommended based on the user's reactions or emotional state.

[0732] The present invention is a system that generates try-on images from multiple clothing images selected by a user on an e-commerce site and uses an emotion engine to provide the user with personalized try-on images and clothing recommendations. A specific embodiment of the present invention will be described below.

[0733] First, a user visits an e-commerce site and selects the desired clothing item from various categories (tops, pants, shoes, etc.). The identifier and image URL of the selected clothing item are sent from the user's device to the server. This information is sent along with the identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, both of these URLs are sent to the server.

[0734] Next, the server retrieves clothing images from the specified URL based on the received identifier and image URL. The server downloads these image data, for example, image data of a red top and blue jeans. After this, the server analyzes the acquired image data and extracts the main features of each garment (collar, sleeves, hem, etc.). The analyzed feature information plays an important role in the subsequent synthesis process.

[0735] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize each clothing image based on the analyzed feature information. This generates an image of how the user would actually try on the clothing they selected. For example, an image of a natural combination of a red top and blue jeans can be generated.

[0736] Furthermore, an emotion engine is used to analyze the user's reactions and emotions. The system analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when viewing the try-on images to understand their emotional state. Based on these results, the system adjusts the generation of try-on images and clothing recommendations.

[0737] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL for the generated try-on image is created as https: / / example.com / generated_outfit.png. The server sends this URL to the user's device, and the user can check the new outfit image on their device.

[0738] Based on the emotional information recognized by the emotion engine, recommended clothing images are also provided according to the user's preferences. For example, if the user has a positive reaction to a fitting image, clothing of a similar style will be displayed in the recommendation list.

[0739] As a concrete example, consider a user who selects a new denim jacket and black jeans. When the user selects "denim jacket" and "black jeans" on an e-commerce website, their respective IDs, jacket123 and jeans456, and their image URLs, https: / / example.com / jacket123.png and https: / / example.com / jeans456.png, are sent to the server. The server downloads the image data of the denim jacket and black jeans from these URLs, analyzes them, and synthesizes them using a generative AI model to generate new try-on images. The emotion engine analyzes the user's reaction and adds similar items to a recommendation list based on the generated try-on images. The server saves the URLs containing the generated try-on images and recommendation list and sends them to the user's device. The user can view the new outfit images and recommendation list.

[0740] An example of a prompt sentence would be a specific instruction input to the generative AI model, such as "Generate an image of someone trying on a denim jacket and black jeans."

[0741] This allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions.

[0742] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0743] Step 1:

[0744] A user accesses an e-commerce site and selects a desired clothing item. The user then sends data including the identifier and image URL of the selected clothing item from the terminal to the server. The input includes the identifiers and image URLs of multiple clothing images, and by sending these to the server, information on the clothing selected by the user is collected.

[0745] Step 2:

[0746] The server retrieves clothing images from the specified URLs based on the received identifier and image URL. The server accesses each URL and downloads clothing image data. The identifier and image URL are input, and the required clothing images are prepared by retrieving image data based on these.

[0747] Step 3:

[0748] The server analyzes the acquired clothing images and extracts the key features of each garment (collar, sleeves, hem, etc.). The server uses an image analysis algorithm to identify the features of each clothing image. The downloaded image data is used as input and analysis is performed to obtain feature information.

[0749] Step 4:

[0750] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. It then generates an image of what it would look like if the user tried on the selected clothing. In this process, a specific prompt (e.g., "Generate an image of me trying on a denim jacket and black jeans") is input to the model. The input is the analyzed feature information and the prompt, and the generated try-on image is output.

[0751] Step 5:

[0752] The emotion engine analyzes the user's reactions and emotions when viewing images of clothing to try on. The server collects the user's facial expressions and actions (clicks, scrolls, etc.) through cameras and interaction data to recognize the user's emotional state. The input is the user's reaction data, and the output is their emotional state.

[0753] Step 6:

[0754] The server saves the generated try-on images and generates a URL to display them to the user. The server saves the generated outfit image data and creates a corresponding access URL. The input is the generated try-on image, and the output is the save destination URL.

[0755] Step 7:

[0756] The server sends the URL to the user's device, enabling it to be displayed on the user's device. The user uses the received URL to check the new outfit image. The input is the destination URL, and the output is the display on the user's device.

[0757] Step 8:

[0758] Based on the emotion information recognized by the emotion engine, the server, which provides recommended clothing images according to the user's preferences, creates a recommendation list and generates a URL for the recommended products so that the user can access them. The user can refer to the recommendation list to check related items. The input is emotion information and a recommendation algorithm, and the output is a recommendation list and its URL.

[0759] The above are the specific processing steps for processing from the user terminal to the server, generating try-on images, analyzing emotions, and the recommendation system.

[0760] 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.

[0761] 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.

[0762] 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.

[0763] [Third embodiment]

[0764] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0765] 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.

[0766] 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).

[0767] 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.

[0768] 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.

[0769] 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).

[0770] 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.

[0771] 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.

[0772] 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.

[0773] 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.

[0774] 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.

[0775] 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."

[0776] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[0777] User selection of clothing

[0778] First, a user accesses an e-commerce site and selects the clothing item they want from various categories (tops, pants, shoes, etc.). If the user selects, for example, a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[0779] Server receives selection information and acquires images

[0780] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[0781] Image analysis and feature extraction

[0782] The server analyzes the acquired image data and extracts key features of the clothing (e.g., collar, sleeves, trouser legs, etc.). The feature information obtained from this analysis plays an important role in the subsequent synthesis process.

[0783] Image synthesis using generative AI models

[0784] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[0785] Save composite image and generate URL

[0786] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[0787] Sending and displaying composite images

[0788] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[0789] Specific examples

[0790] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0791] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0792] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0793] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0794] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0795] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0796] 6. The server saves the generated try-on image as https: / / example.com / generated_outfit123456.png.

[0797] 7. The server sends this URL to the user's device.

[0798] 8. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user can check the new outfit image.

[0799] This series of processes allows users to get a concrete image of how the clothing will look before purchasing, reducing the risk of making a wrong purchase or having to return the item.

[0800] The processing flow will be explained below.

[0801] Step 1:

[0802] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[0803] Step 2:

[0804] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0805] Step 3:

[0806] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0807] Step 4:

[0808] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[0809] Step 5:

[0810] The server runs a generative AI model, such as a generative adversarial network (GAN), to synthesize clothing images based on the analyzed feature information. This process generates a try-on image combining the user's chosen clothing.

[0811] Step 6:

[0812] The server saves the generated try-on image and generates a URL that can be accessed by the user. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit123456.png.

[0813] Step 7:

[0814] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[0815] Step 8:

[0816] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look in real life and decide whether to purchase it.

[0817] Through this series of processes, users can get a concrete idea of ​​how the clothing will look before purchasing, reducing the risk of making a mistake in purchasing or the hassle of having to return the item.

[0818] Example 1

[0819] 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."

[0820] In conventional e-commerce systems, when users purchase clothing, it is difficult for them to see how it will look when worn without actually trying it on. Even if users select the clothing combinations they want, the combinations are not displayed in a natural arrangement, which reduces post-purchase satisfaction and increases the hassle of returning items. Furthermore, there was a lack of a way to effectively manage and save each user's outfit images, making it difficult to meet individual needs.

[0821] 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.

[0822] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on the e-commerce site; means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for saving the generated outfit image and generating a URL for displaying it to the user; means for transmitting the URL to the user terminal and enabling display on the user terminal; and means for inputting a desired clothing combination by the user and generating a prompt message for the generative AI model to generate a try-on image based on the input. This allows users to see a specific and natural look before purchasing, reducing the hassle of purchasing mistakes and returns. Furthermore, it is possible to effectively manage and store the outfit images generated for each user, thereby providing services tailored to individual needs.

[0823] "User" refers to an individual or group that selects clothing on an e-commerce site and requests the generation of try-on images.

[0824] "E-commerce site" refers to a website that sells clothing and other products over the Internet.

[0825] "Identifier" refers to an ID or tag that uniquely identifies a specific clothing image.

[0826] "Image URL" refers to a Uniform Resource Locator (URL) for accessing a clothing image stored on the Internet.

[0827] "Server" refers to a computer system that receives and processes requests from users.

[0828] "Features" refer to key parts or characteristics of the clothing in the image, such as the collar, sleeves, and hem.

[0829] "Generative AI model" refers to an artificial intelligence model that generates try-on images based on the characteristics of clothing selected by the user. An example is a generative adversarial network (GAN).

[0830] A "try-on image" refers to an image created by combining multiple pieces of clothing selected by the user, creating a composite image that looks as if the clothing is actually being worn.

[0831] "Saving" refers to recording the generated try-on images in a database or storage device.

[0832] "URL generation" refers to creating a uniform resource locator (URL) for accessing the saved try-on images.

[0833] "Prompt sentence" refers to an instruction sentence provided as input to a generative AI model to generate try-on images.

[0834] "User terminal" refers to a device such as a computer, smartphone, or tablet that a user uses to access an e-commerce site, make selections, and check images of items being tried on.

[0835] The present invention provides a system that receives the identifiers and image URLs of selected clothing images from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[0836] User selection of clothing

[0837] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user selects the clothing item they want from various categories (tops, pants, shoes, etc.). For example, if a user selects a "red top" and "blue jeans," this selection information is sent from the user's device to the server.

[0838] Server receives selection information and acquires images

[0839] The server retrieves clothing images from the specified URL based on the identifier (ID) and image URL received from the user's device. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the image data will be downloaded from these URLs. Specifically, the server retrieves images using a client library such as curl or requests.

[0840] Image analysis and feature extraction

[0841] The server analyzes the acquired image data using image analysis software such as OpenCV and TensorFlow to extract key features of the clothing (collar, sleeves, trouser hems, etc.) The feature information obtained from this analysis plays an important role in the subsequent image synthesis process.

[0842] Image synthesis using generative AI models

[0843] The server uses a generative AI model such as a generative adversarial network (GAN) to synthesize clothing images based on the extracted feature information. An image of what the clothing selected by the user would look like when actually worn is generated. For example, an outfit image of a red top and blue jeans naturally combined can be created. To facilitate this process, a prompt is used as input to the generative AI model. An example of a prompt is "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate an image of the clothing being tried on."

[0844] Save composite image and generate URL

[0845] The generated try-on image is saved to a storage device by the server. The server generates a URL so that the saved try-on image can be accessed. For example, the URL of the generated try-on image is configured as https: / / example.com / generated_outfit.png.

[0846] Sending and displaying composite images

[0847] The server sends the generated URL to the user's device. The user can open this URL on their device to view the new outfit image. For example, the received URL https: / / example.com / generated_outfit.png can be displayed in a web browser or dedicated application.

[0848] Specific examples

[0849] For example, if a user selects a denim jacket and black jeans, the following happens:

[0850] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0851] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0852] 3. The server accesses the sent URL and downloads the image.

[0853] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0854] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0855] 6. The server saves the generated try-on image in storage and generates a URL.

[0856] 7. The server sends the generated URL to the user's device.

[0857] 8. The device displays the received URL, and the user can check the new outfit image.

[0858] This series of processes allows users to get a concrete image of how the garment will look before purchasing, reducing the risk of buying something wrong or having to return it.

[0859] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0860] Step 1: User selects clothing

[0861] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user navigates through categories (tops, pants, shoes, etc.) on the site and selects the desired clothing item. The input is the identifier (ID) and image URL of the selected clothing item. For example, if the user selects "red top" and "blue jeans," the output is the identifier and image URL of the selected clothing item. Specifically, when the user confirms their selection, the selection information is sent from the user's device to the server.

[0862] Step 2: Server receives selection information

[0863] The server receives the identifier (ID) and image URL sent from the user's device. The input is the identifier and image URL sent from the user's device, and the output is to store this information as received data. Specifically, the server receives the HTTPS request and saves the selected information in a database.

[0864] Step 3: Acquiring image data

[0865] The server downloads the specified image data based on the URL received. The input is an identifier and an image URL, and the output is the retrieved image data. Specifically, the server accesses the image URL using a client library such as curl or requests and saves the image data in the server's storage. For example, download the image data of a red top from the URL https: / / example.com / red_top.png.

[0866] Step 4: Image analysis and feature extraction

[0867] The server analyzes the acquired image data and extracts the key features of the clothing. The input is the acquired image data, and the output is the extracted feature information. Specifically, the server uses image analysis software such as OpenCV or TensorFlow to identify features such as the collar, sleeves, and trouser hems of the clothing. For example, the analysis results in identifying the position of the collar and the length of the sleeves of a red top.

[0868] Step 5: Image synthesis using generative AI models

[0869] The server uses the analyzed feature information as input and generates a try-on image using a generative AI model (e.g., a generative adversarial network (GAN)). The input is the extracted feature information, and the output is the generated try-on image. Specifically, the server converts the feature information into a prompt and inputs it into the generative AI model. For example, the server generates a prompt such as, "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate a try-on image." Based on this, the generative AI model generates an image that combines a red top and blue jeans.

[0870] Step 6: Save the composite image and generate a URL

[0871] The server saves the generated try-on image to a storage device and generates an accessible URL. The input is the generated try-on image, and the output is the URL of the saved try-on image. Specifically, the server saves the try-on image as a file and generates a unique URL for that file. For example, the generated URL might be https: / / example.com / generated_outfit.png.

[0872] Step 7: Send and view the composite image

[0873] The server sends the generated URL to the user's device, allowing the user to view the new outfit image on the device. The input is the generated URL, and the output is the try-on image displayed on the user's device. Specifically, the server returns a URL as an API response, and the user's device receives and displays this URL. For example, the user opens the URL https: / / example.com / generated_outfit.png in a browser, and the new outfit image is displayed.

[0874] (Application example 1)

[0875] 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."

[0876] On conventional e-commerce sites, users could not try on clothes before purchasing, making it difficult to confirm whether the item actually fit them. As a result, users had to go through the trouble of returning or exchanging the item after purchase if the size or design did not fit. In addition, it was difficult for users to visualize what the selected clothing would look like in real life, which led to a decrease in purchasing motivation.

[0877] 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.

[0878] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site; means for acquiring the plurality of clothing images based on the identifier and the image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for the user to upload a full-body image of themselves and analyze the acquired clothing images and the user's full-body image; means for executing a generative AI model that generates a try-on image in which the selected clothing naturally fits the user's full-body image based on the analyzed characteristics; means for saving the generated try-on image and generating a URL for displaying it to the user; and means for transmitting the URL to the user terminal and enabling display on the user terminal. This allows the user to specifically check the try-on image based on their full-body image, allowing them to have a detailed image before purchasing, which increases their motivation to purchase and reduces the hassle of returns and exchanges due to mismatched sizes or designs.

[0879] A "user" is a consumer who uses an e-commerce site to select and purchase products.

[0880] An "e-commerce site" is a website that sells and buys goods and services over the Internet.

[0881] "Clothing Images" are digital images of clothing displayed on e-commerce sites.

[0882] An "identifier" is information for uniquely identifying a specific clothing image.

[0883] An "image URL" is an address (URL) for accessing a specific image file on the Internet.

[0884] A "full-body image" is a photograph showing the user's entire body.

[0885] "Feature analysis" is a technique for extracting specific elements or patterns from an image.

[0886] A "generative AI model" is an algorithm that uses artificial intelligence techniques to generate new images or data, such as generative adversarial networks (GANs).

[0887] The "try-on image" is a composite image of the user's whole body and a clothing image, simulating the appearance of the clothing when the user actually tries it on.

[0888] "Saving" means to hold the generated try-on image as data.

[0889] "URL generation" means creating a new URL to access the saved try-on image.

[0890] A "user terminal" is a device used by a user to access an e-commerce site, and includes a smartphone, tablet, computer, etc.

[0891] "Transmission" means sending data or information from the server to the user terminal.

[0892] "Display" refers to the visual presentation of information or images on the screen of a user terminal.

[0893] The present invention is a system that generates try-on images based on multiple clothing images selected by a user on an e-commerce site. The following describes how to implement this system step by step.

[0894] System Overview

[0895] The system consists of a server that runs on the backend of an e-commerce site and a terminal used by users. Users upload a full-body image of themselves and select clothing images on the site. The server analyzes this data and uses a generative AI model to generate try-on images and provide them to the user.

[0896] Hardware and software used

[0897] Smartphones, tablets, and computers: devices used by users to access

[0898] Server: Receives, stores, analyzes, and synthesizes data. Specifically, it uses cloud services such as AWS and Google Cloud Platform.

[0899] Generative AI models: Use generative adversarial networks (GANs) using PyTorch and TensorFlow.

[0900] Image analysis software: For example, OpenCV or Scikit-Image can be used to extract image features.

[0901] Data Flow and Processing

[0902] 1. User Action:

[0903] Users access an e-commerce site, select clothing items, and upload a full-body image of themselves.

[0904] 2. Data reception and processing by the server:

[0905] The server receives the identifier and image URL of the garment selected by the user, as well as a full-body image of the user.

[0906] 3. Image Acquisition and Feature Analysis:

[0907] The server retrieves the clothing image based on the identifier and the image URL, and analyzes the characteristics of the clothing and the user's full-body image using image analysis software.

[0908] 4. Generate fitting images:

[0909] Using a generative AI model (e.g., GAN using PyTorch or TensorFlow), a fitting image is generated that naturally combines the selected clothing with the user's full-body image based on the analyzed features.

[0910] 5. Save the generated image and generate a URL:

[0911] The generated try-on images are stored on the server, and a URL is generated that the user can access.

[0912] 6. User Notification and Display:

[0913] The server sends the generated URL to the user's device, and the user can open the URL to check the try-on image.

[0914] Examples of concrete examples and prompts

[0915] Specific examples

[0916] For example, if a user selects a "red top" and "blue jeans" and uploads a full-body image of themselves, the server will process the following:

[0917] 1. The server retrieves the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png for the selected clothing.

[0918] 2. The server receives the user's full-body image.

[0919] 3. The server analyzes the clothing image and the user image and extracts key features.

[0920] 4. Generate try-on images using a generative AI model.

[0921] 5. Save the try-on image and generate the URL https: / / example.com / generated_tryon.png.

[0922] 6. The user opens the URL on their device and checks the fitting image.

[0923] Prompt Sentence Examples

[0924] "Choose a red top and blue jeans and generate a natural try-on image based on the user's preferences. We use a GAN model to generate images that combine these clothes with the user's full-body photo."

[0925] This allows users to see a detailed try-on image before purchasing, which increases purchasing motivation and reduces the hassle of returning or exchanging items due to mismatched sizes or designs.

[0926] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0927] Step 1:

[0928] A user accesses an e-commerce site, selects the clothing item they want, and then takes or selects a full-body image from their smartphone or computer and uploads it.

[0929] Input: Clothing selection information (identifier, image URL), full-body image.

[0930] Output: Clothing selection data and full-body image data.

[0931] Specific operation: The user selects clothing through the interface of the e-commerce site and uploads a full-body image of themselves, which is then stored in cloud storage.

[0932] Step 2:

[0933] The server receives the user's clothing selection information (identifier and image URL) and a full-body image.

[0934] Input: Clothing selection data and full-body image data sent by the user.

[0935] Output: Selection data and whole body image stored on the server.

[0936] Specific operation: The server processes the received data and downloads the clothing image based on the identifier and image URL. The full-body image is stored in the database.

[0937] Step 3:

[0938] The server retrieves and analyzes the selected clothing image based on the identifier and the image URL.

[0939] Input: Clothing image URL.

[0940] Output: Feature data of the analyzed clothing image.

[0941] Specific operation: The server downloads the clothing image using the image URL and analyzes the features of the clothing image (e.g., shape, color, design) using OpenCV and Scikit-Image.

[0942] Step 4:

[0943] The server analyzes the user's full-body image and extracts key features of the image.

[0944] Input: A full-body image of the user.

[0945] Output: Feature data of the analyzed whole-body image.

[0946] Specific operation: The server analyzes the whole-body image using an image processing library such as OpenCV and extracts specific features such as contours and facial position.

[0947] Step 5:

[0948] The server uses the analyzed clothing images and feature data from the full-body image as input and generates a fitting image using a generative AI model (GAN).

[0949] Input: Feature data of analyzed clothing images, feature data of whole-body images.

[0950] Output: The generated try-on image.

[0951] Specific operation: The server uses PyTorch and TensorFlow to execute a generative AI model and generate a fitting image in which the selected clothing fits naturally to the user's full-body image.

[0952] Step 6:

[0953] The server saves the generated try-on image and generates a URL that the user can access.

[0954] Input: Generated try-on images.

[0955] Output: URL of the try-on image.

[0956] Specific operation: The fitting images are saved in a database and a URL for accessing them is generated. These URLs are managed in the database.

[0957] Step 7:

[0958] The server sends the generated URL to the user's terminal, and the user checks the fitting image on their own terminal.

[0959] Input: The URL of the generated try-on image.

[0960] Output: Try-on image displayed on the user's device.

[0961] Specific operation: The server sends the generated URL to the user's device, and the user opens the link to view the fitting image. The user interface is displayed in a browser or a dedicated app.

[0962] This allows the user to see a concrete image of the item before purchasing.

[0963] 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.

[0964] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display, and further combines it with an emotion engine to provide more personalized try-on images and recommended clothing images. Specific implementation of this system will be described below.

[0965] User selection of clothing

[0966] First, a user accesses an e-commerce site and selects the items they want from various categories (tops, pants, shoes, etc.). For example, if the user selects "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[0967] Server receives selection information and acquires images

[0968] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[0969] Image analysis and feature extraction

[0970] The server analyzes the acquired image data and extracts the key features of each garment (e.g., collar, sleeve, trouser leg, etc.). The feature information obtained through this analysis plays an important role in the subsequent synthesis process.

[0971] Image synthesis using generative AI models

[0972] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[0973] Use of emotion engine

[0974] Furthermore, the system uses an emotion engine to recognize the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at the try-on images to understand their emotional state. Based on this state, it adjusts the generation of try-on images and clothing recommendations.

[0975] Save composite image and generate URL

[0976] The generated try-on image is saved by the server and a URL is generated so that users can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[0977] Sending and displaying composite images

[0978] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[0979] Display of recommended clothing

[0980] Furthermore, based on the emotional information recognized by the emotion engine, the system provides recommended clothing images according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[0981] Specific examples

[0982] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[0983] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[0984] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[0985] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[0986] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[0987] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[0988] 6. The emotion engine analyzes the user's reaction and adds similar items to the recommendation list based on the generated try-on images.

[0989] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[0990] 8. The server sends this URL to the user's device.

[0991] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[0992] This process allows users to get a concrete image of how the clothes will look before they buy, and also provides personalized recommendations based on their emotions, which increases the success rate of purchases and reduces the hassle of returns.

[0993] The processing flow will be explained below.

[0994] Step 1:

[0995] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[0996] Step 2:

[0997] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[0998] Step 3:

[0999] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[1000] Step 4:

[1001] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[1002] Step 5:

[1003] The server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates a try-on image of the user's chosen clothing.

[1004] Step 6:

[1005] The emotion engine recognizes the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at images of items to try on, and understands their emotional state.

[1006] Step 7:

[1007] Based on the user's emotions recognized by the emotion engine, the generated try-on images are adjusted to create a more suitable outfit for the user. For example, if the user expresses positive emotions, the system generates images that emphasize that style.

[1008] Step 8:

[1009] The server saves the generated try-on image and generates a URL that users can access. For example, the URL for the generated try-on image might be https: / / example.com / generated_outfit.png.

[1010] Step 9:

[1011] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[1012] Step 10:

[1013] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look when worn and decide whether to purchase it.

[1014] Step 11:

[1015] Based on the emotional information recognized by the emotion engine, recommended clothing images are provided according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[1016] This specific processing step allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions, thereby increasing the success rate of purchases and reducing the hassle of returns.

[1017] Example 2

[1018] 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."

[1019] On modern e-commerce sites, users have limited options for trying on clothing before purchasing. This can lead to an increase in returns and exchanges, as the product may not meet expectations after purchase. Furthermore, the lack of personalized recommendations based on users' emotions and preferences makes it difficult to increase user satisfaction.

[1020] 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.

[1021] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site, means for acquiring the plurality of clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for saving the generated outfit image and generating a URL for displaying it to the user, means for transmitting the URL to the user terminal and enabling display on the user terminal, and means for recognizing the user's reactions and emotions using an emotion engine and adjusting the generation of try-on images or clothing recommendations based on the user's emotional state. This allows the user to see specific try-on images before purchasing and receive personalized recommendations based on their emotions.

[1022] A "user" is someone who uses this system to select clothing on an e-commerce site and check how it looks when worn.

[1023] An "e-commerce site" is a website for selling and purchasing goods over the Internet.

[1024] A "clothing image" is image data that shows a visual representation of the clothing selected by the user.

[1025] An "identifier" is an ID or code that uniquely identifies a particular clothing image.

[1026] An "image URL" is an address that indicates where a specific clothing image is stored on the Internet.

[1027] The "server" is a computer system that receives information sent by the user, acquires and analyzes images, and synthesizes images for trying on.

[1028] "Image analysis" is a process for extracting feature information from acquired clothing images.

[1029] "Features" are information that indicates specific parts, shapes, patterns, etc. in a clothing image.

[1030] A "generative AI model" is an artificial intelligence algorithm that synthesizes new try-on images based on acquired feature information.

[1031] A "try-on image" is a visual image generated by combining multiple pieces of clothing selected by the user.

[1032] A "URL" is an internet address used to access a particular web page or resource.

[1033] An "emotion engine" is an artificial intelligence system that recognizes users' reactions and emotions and analyzes that data.

[1034] "Reactions" refer to the user's actions, facial expressions, and interactions such as clicking and scrolling when looking at the try-on images.

[1035] "Recommendation" is the process of selecting relevant clothing items based on a user's preferences.

[1036] This system receives the identifier and image URL of a clothing image selected by a user from multiple clothing images selected on an e-commerce site, and generates try-on images through acquisition, analysis, synthesis, and display. Furthermore, by combining an emotion engine, it provides more personalized try-on images and recommended clothing images.

[1037] Hardware and software used

[1038] The server is a computer system with high-performance computing capabilities for receiving data, acquiring images, analyzing them, storing data, and running the generated AI model. For specific implementation, it is recommended to use deep learning libraries such as TensorFlow or PyTorch, and OpenCV for image analysis.

[1039] The user's terminal is typically a smartphone, tablet, or personal computer that accesses the e-commerce site via a web browser.

[1040] An emotion engine requires an algorithm to analyze user reactions and emotions and predict purchasing behavior. Specifically, an analytical model using machine learning is considered.

[1041] Program processing

[1042] A user accesses an e-commerce site and selects the clothing item they want. For example, if the user selects a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server along with an identifier and an image URL.

[1043] The server retrieves clothing images from the specified URLs based on the identifier (ID) and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads images from these URLs.

[1044] The acquired images are analyzed by the server to extract key features using image processing libraries such as OpenCV. After the feature information is extracted, the server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize a try-on image based on the feature information.

[1045] The generated try-on image is saved by the server and a URL that can be accessed by the user is generated. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png, and the server generates this URL and saves it in the database.

[1046] The server then sends the generated URL to the user's device. When the user opens the URL, an image of the new outfit is displayed. The user can then review the image and decide whether to actually purchase it.

[1047] Furthermore, the server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, the server can understand the user's emotional state from their facial expressions when looking at the try-on images, as well as their actions such as clicking and scrolling, and adjust the generation of try-on images or clothing recommendations based on this.

[1048] Specific examples

[1049] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[1050] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[1051] 2. The device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[1052] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[1053] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[1054] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[1055] 6. The emotion engine analyzes the user's reactions and adds similar items to the recommendation list based on the generated try-on images.

[1056] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[1057] 8. The server sends this URL to the user's device.

[1058] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[1059] Using this system, users can easily get a concrete try-on image before purchasing and receive personalized recommendations based on their emotions.

[1060] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1061] Step 1:

[1062] A user accesses an e-commerce site and selects the clothing they want. For example, they select a "red top" and "blue jeans," and then the information (identifier and image URL) is sent from the terminal to the server. The input is the user's selection information, and the output is the clothing identifier and image URL sent to the server.

[1063] Step 2:

[1064] The server retrieves clothing images from the specified URLs based on the identifier and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads image files from these URLs. The input is the identifier and image URL sent in the previous step, and the output is the downloaded clothing image.

[1065] Step 3:

[1066] The server analyzes the acquired image data and extracts key clothing features. This analysis process uses image processing libraries such as OpenCV. For example, it identifies features such as the collar and sleeves of a red top, or the hem and waist of blue jeans. The input is the downloaded clothing image, and the output is the analyzed feature information.

[1067] Step 4:

[1068] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. For example, it generates a try-on image of a red top paired with blue jeans. This process uses deep learning libraries such as TensorFlow and PyTorch. The input is the analyzed feature information, and the output is the generated try-on image.

[1069] Step 5:

[1070] The server saves the generated try-on image and generates a URL to make the image accessible. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit.png. The input is the generated try-on image, and the output is the URL of the generated try-on image.

[1071] Step 6:

[1072] The server sends the generated URL to the user's device. When the user accesses this URL on their device, the generated try-on image is displayed. For example, if the user accesses https: / / example.com / generated_outfit.png in their web browser, they can see the synthesized try-on image. The input is the generated URL, and the output is the try-on image displayed on the user's device.

[1073] Step 7:

[1074] The server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, it collects and analyzes behavioral data such as facial expressions, clicks, and scrolls when the user looks at images of items to try on. The input is the user's reaction data, and the output is analyzed emotional information.

[1075] Step 8:

[1076] The server generates try-on images and adjusts clothing recommendations based on the analyzed emotional information. For example, if the user responds positively, clothing of a similar style is added to the recommendation list. The input is the analyzed emotional information, and the output is the adjusted try-on images and recommendation list.

[1077] This system allows users to see specific try-on images and receive personalized recommendations based on their emotions.

[1078] (Application example 2)

[1079] 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."

[1080] Conventional e-commerce sites lack an effective way to provide users with try-on images of the clothing they select. This leads to users purchasing clothing without visual uncertainty, resulting in a high return rate. Furthermore, the lack of personalized clothing recommendations based on users' emotions and preferences limits the user experience. This creates a need for a system that can facilitate purchasing decisions.

[1081] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1082] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on an e-commerce site, means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for recognizing the user's emotional state using an emotion engine and adjusting the generation of try-on images and clothing recommendations based on that state, means for saving the generated outfit images and generating a URL for displaying them to the user, and means for transmitting the URL to a user terminal so that the URL can be displayed on the user terminal. This allows the user to visually check the try-on images and receive personalized recommendations based on the emotion analysis results.

[1083] "User" refers to a customer or consumer who uses the e-commerce site.

[1084] "E-commerce site" means a website where purchases of goods and services are made over the Internet.

[1085] "Garment image" refers to a digital image that shows the visual characteristics of a garment.

[1086] "Identifier" refers to an ID or code that uniquely identifies each clothing image.

[1087] "Image URL" refers to the web address where the clothing image is stored on the Internet.

[1088] "Acquisition" refers to the act of downloading clothing images and related information from an e-commerce site or server.

[1089] "Analysis" refers to the procedure of detecting and extracting features and patterns within clothing images.

[1090] "Generative AI model" refers to an artificial intelligence (AI) algorithm for generating new try-on images from image data.

[1091] "Synthesis" refers to the act of combining multiple clothing images to create a new try-on image.

[1092] "Try-on image" refers to a visual display of a user virtually trying on the clothing combination they have selected.

[1093] An "emotion engine" refers to technology for detecting and analyzing a user's emotions and reactions.

[1094] "Saving" refers to the act of recording the generated try-on image as data on a server or the like.

[1095] "URL generation" refers to the process of creating a web address that allows users to access the generated try-on images.

[1096] "Sending" refers to the act of transferring the URL of the generated try-on image to the user's device.

[1097] "Display" refers to the act of visually presenting try-on images and recommended clothing images generated on the user terminal.

[1098] "Personalization" refers to the act of providing individually tailored services and recommendations based on a user's preferences and emotions.

[1099] "Recommended Clothing Images" refers to additional clothing images that are recommended based on the user's reactions or emotional state.

[1100] The present invention is a system that generates try-on images from multiple clothing images selected by a user on an e-commerce site and uses an emotion engine to provide the user with personalized try-on images and clothing recommendations. A specific embodiment of the present invention will be described below.

[1101] First, a user visits an e-commerce site and selects the desired clothing item from various categories (tops, pants, shoes, etc.). The identifier and image URL of the selected clothing item are sent from the user's device to the server. This information is sent along with the identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, both of these URLs are sent to the server.

[1102] Next, the server retrieves clothing images from the specified URL based on the received identifier and image URL. The server downloads these image data, for example, image data of a red top and blue jeans. After this, the server analyzes the acquired image data and extracts the main features of each garment (collar, sleeves, hem, etc.). The analyzed feature information plays an important role in the subsequent synthesis process.

[1103] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize each clothing image based on the analyzed feature information. This generates an image of how the user would actually try on the clothing they selected. For example, an image of a natural combination of a red top and blue jeans can be generated.

[1104] Furthermore, an emotion engine is used to analyze the user's reactions and emotions. The system analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when viewing the try-on images to understand their emotional state. Based on these results, the system adjusts the generation of try-on images and clothing recommendations.

[1105] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL for the generated try-on image is created as https: / / example.com / generated_outfit.png. The server sends this URL to the user's device, and the user can check the new outfit image on their device.

[1106] Based on the emotional information recognized by the emotion engine, recommended clothing images are also provided according to the user's preferences. For example, if the user has a positive reaction to a fitting image, clothing of a similar style will be displayed in the recommendation list.

[1107] As a concrete example, consider a user who selects a new denim jacket and black jeans. When the user selects "denim jacket" and "black jeans" on an e-commerce website, their respective IDs, jacket123 and jeans456, and their image URLs, https: / / example.com / jacket123.png and https: / / example.com / jeans456.png, are sent to the server. The server downloads the image data of the denim jacket and black jeans from these URLs, analyzes them, and synthesizes them using a generative AI model to generate new try-on images. The emotion engine analyzes the user's reaction and adds similar items to a recommendation list based on the generated try-on images. The server saves the URLs containing the generated try-on images and recommendation list and sends them to the user's device. The user can view the new outfit images and recommendation list.

[1108] An example of a prompt sentence would be a specific instruction input to the generative AI model, such as "Generate an image of someone trying on a denim jacket and black jeans."

[1109] This allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions.

[1110] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1111] Step 1:

[1112] A user accesses an e-commerce site and selects a desired clothing item. The user then sends data including the identifier and image URL of the selected clothing item from the terminal to the server. The input includes the identifiers and image URLs of multiple clothing images, and by sending these to the server, information on the clothing selected by the user is collected.

[1113] Step 2:

[1114] The server retrieves clothing images from the specified URLs based on the received identifier and image URL. The server accesses each URL and downloads clothing image data. The identifier and image URL are input, and the required clothing images are prepared by retrieving image data based on these.

[1115] Step 3:

[1116] The server analyzes the acquired clothing images and extracts the key features of each garment (collar, sleeves, hem, etc.). The server uses an image analysis algorithm to identify the features of each clothing image. The downloaded image data is used as input and analysis is performed to obtain feature information.

[1117] Step 4:

[1118] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. It then generates an image of what it would look like if the user tried on the selected clothing. In this process, a specific prompt (e.g., "Generate an image of me trying on a denim jacket and black jeans") is input to the model. The input is the analyzed feature information and the prompt, and the generated try-on image is output.

[1119] Step 5:

[1120] The emotion engine analyzes the user's reactions and emotions when viewing images of clothing to try on. The server collects the user's facial expressions and actions (clicks, scrolls, etc.) through cameras and interaction data to recognize the user's emotional state. The input is the user's reaction data, and the output is their emotional state.

[1121] Step 6:

[1122] The server saves the generated try-on images and generates a URL to display them to the user. The server saves the generated outfit image data and creates a corresponding access URL. The input is the generated try-on image, and the output is the save destination URL.

[1123] Step 7:

[1124] The server sends the URL to the user's device, enabling it to be displayed on the user's device. The user uses the received URL to check the new outfit image. The input is the destination URL, and the output is the display on the user's device.

[1125] Step 8:

[1126] Based on the emotion information recognized by the emotion engine, the server, which provides recommended clothing images according to the user's preferences, creates a recommendation list and generates a URL for the recommended products so that the user can access them. The user can refer to the recommendation list to check related items. The input is emotion information and a recommendation algorithm, and the output is a recommendation list and its URL.

[1127] The above are the specific processing steps for processing from the user terminal to the server, generating try-on images, analyzing emotions, and the recommendation system.

[1128] 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.

[1129] 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.

[1130] 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.

[1131] [Fourth embodiment]

[1132] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1133] 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.

[1134] 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).

[1135] 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.

[1136] 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.

[1137] 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).

[1138] 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.

[1139] 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.

[1140] 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.

[1141] 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.

[1142] 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.

[1143] 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.

[1144] 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."

[1145] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[1146] User selection of clothing

[1147] First, a user accesses an e-commerce site and selects the clothing item they want from various categories (tops, pants, shoes, etc.). If the user selects, for example, a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[1148] Server receives selection information and acquires images

[1149] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[1150] Image analysis and feature extraction

[1151] The server analyzes the acquired image data and extracts key features of the clothing (e.g., collar, sleeves, trouser legs, etc.). The feature information obtained from this analysis plays an important role in the subsequent synthesis process.

[1152] Image synthesis using generative AI models

[1153] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[1154] Save composite image and generate URL

[1155] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[1156] Sending and displaying composite images

[1157] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[1158] Specific examples

[1159] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[1160] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[1161] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[1162] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[1163] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[1164] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[1165] 6. The server saves the generated try-on image as https: / / example.com / generated_outfit123456.png.

[1166] 7. The server sends this URL to the user's device.

[1167] 8. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user can check the new outfit image.

[1168] This series of processes allows users to get a concrete image of how the clothing will look before purchasing, reducing the risk of making a wrong purchase or having to return the item.

[1169] The processing flow will be explained below.

[1170] Step 1:

[1171] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[1172] Step 2:

[1173] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[1174] Step 3:

[1175] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[1176] Step 4:

[1177] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[1178] Step 5:

[1179] The server runs a generative AI model, such as a generative adversarial network (GAN), to synthesize clothing images based on the analyzed feature information. This process generates a try-on image combining the user's chosen clothing.

[1180] Step 6:

[1181] The server saves the generated try-on image and generates a URL that can be accessed by the user. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit123456.png.

[1182] Step 7:

[1183] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[1184] Step 8:

[1185] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look in real life and decide whether to purchase it.

[1186] Through this series of processes, users can get a concrete idea of ​​how the clothing will look before purchasing, reducing the risk of making a mistake in purchasing or the hassle of having to return the item.

[1187] Example 1

[1188] 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."

[1189] In conventional e-commerce systems, when users purchase clothing, it is difficult for them to see how it will look when worn without actually trying it on. Even if users select the clothing combinations they want, the combinations are not displayed in a natural arrangement, which reduces post-purchase satisfaction and increases the hassle of returning items. Furthermore, there was a lack of a way to effectively manage and save each user's outfit images, making it difficult to meet individual needs.

[1190] 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.

[1191] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on the e-commerce site; means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for saving the generated outfit image and generating a URL for displaying it to the user; means for transmitting the URL to the user terminal and enabling display on the user terminal; and means for inputting a desired clothing combination by the user and generating a prompt message for the generative AI model to generate a try-on image based on the input. This allows users to see a specific and natural look before purchasing, reducing the hassle of purchasing mistakes and returns. Furthermore, it is possible to effectively manage and store the outfit images generated for each user, thereby providing services tailored to individual needs.

[1192] "User" refers to an individual or group that selects clothing on an e-commerce site and requests the generation of try-on images.

[1193] "E-commerce site" refers to a website that sells clothing and other products over the Internet.

[1194] "Identifier" refers to an ID or tag that uniquely identifies a specific clothing image.

[1195] "Image URL" refers to a Uniform Resource Locator (URL) for accessing a clothing image stored on the Internet.

[1196] "Server" refers to a computer system that receives and processes requests from users.

[1197] "Features" refer to key parts or characteristics of the clothing in the image, such as the collar, sleeves, and hem.

[1198] "Generative AI model" refers to an artificial intelligence model that generates try-on images based on the characteristics of clothing selected by the user. An example is a generative adversarial network (GAN).

[1199] A "try-on image" refers to an image created by combining multiple pieces of clothing selected by the user, creating a composite image that looks as if the clothing is actually being worn.

[1200] "Saving" refers to recording the generated try-on images in a database or storage device.

[1201] "URL generation" refers to creating a uniform resource locator (URL) for accessing the saved try-on images.

[1202] "Prompt sentence" refers to an instruction sentence provided as input to a generative AI model to generate try-on images.

[1203] "User terminal" refers to a device such as a computer, smartphone, or tablet that a user uses to access an e-commerce site, make selections, and check images of items being tried on.

[1204] The present invention provides a system that receives the identifiers and image URLs of selected clothing images from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display. Specific implementation of this system will be described below.

[1205] User selection of clothing

[1206] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user selects the clothing item they want from various categories (tops, pants, shoes, etc.). For example, if a user selects a "red top" and "blue jeans," this selection information is sent from the user's device to the server.

[1207] Server receives selection information and acquires images

[1208] The server retrieves clothing images from the specified URL based on the identifier (ID) and image URL received from the user's device. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the image data will be downloaded from these URLs. Specifically, the server retrieves images using a client library such as curl or requests.

[1209] Image analysis and feature extraction

[1210] The server analyzes the acquired image data using image analysis software such as OpenCV and TensorFlow to extract key features of the clothing (collar, sleeves, trouser hems, etc.) The feature information obtained from this analysis plays an important role in the subsequent image synthesis process.

[1211] Image synthesis using generative AI models

[1212] The server uses a generative AI model such as a generative adversarial network (GAN) to synthesize clothing images based on the extracted feature information. An image of what the clothing selected by the user would look like when actually worn is generated. For example, an outfit image of a red top and blue jeans naturally combined can be created. To facilitate this process, a prompt is used as input to the generative AI model. An example of a prompt is "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate an image of the clothing being tried on."

[1213] Save composite image and generate URL

[1214] The generated try-on image is saved to a storage device by the server. The server generates a URL so that the saved try-on image can be accessed. For example, the URL of the generated try-on image is configured as https: / / example.com / generated_outfit.png.

[1215] Sending and displaying composite images

[1216] The server sends the generated URL to the user's device. The user can open this URL on their device to view the new outfit image. For example, the received URL https: / / example.com / generated_outfit.png can be displayed in a web browser or dedicated application.

[1217] Specific examples

[1218] For example, if a user selects a denim jacket and black jeans, the following happens:

[1219] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[1220] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[1221] 3. The server accesses the sent URL and downloads the image.

[1222] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[1223] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[1224] 6. The server saves the generated try-on image in storage and generates a URL.

[1225] 7. The server sends the generated URL to the user's device.

[1226] 8. The device displays the received URL, and the user can check the new outfit image.

[1227] This series of processes allows users to get a concrete image of how the garment will look before purchasing, reducing the risk of buying something wrong or having to return it.

[1228] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1229] Step 1: User selects clothing

[1230] A user accesses an e-commerce site using a device such as a PC, smartphone, or tablet. The user navigates through categories (tops, pants, shoes, etc.) on the site and selects the desired clothing item. The input is the identifier (ID) and image URL of the selected clothing item. For example, if the user selects "red top" and "blue jeans," the output is the identifier and image URL of the selected clothing item. Specifically, when the user confirms their selection, the selection information is sent from the user's device to the server.

[1231] Step 2: Server receives selection information

[1232] The server receives the identifier (ID) and image URL sent from the user's device. The input is the identifier and image URL sent from the user's device, and the output is to store this information as received data. Specifically, the server receives the HTTPS request and saves the selected information in a database.

[1233] Step 3: Acquiring image data

[1234] The server downloads the specified image data based on the URL received. The input is an identifier and an image URL, and the output is the retrieved image data. Specifically, the server accesses the image URL using a client library such as curl or requests and saves the image data in the server's storage. For example, download the image data of a red top from the URL https: / / example.com / red_top.png.

[1235] Step 4: Image analysis and feature extraction

[1236] The server analyzes the acquired image data and extracts the key features of the clothing. The input is the acquired image data, and the output is the extracted feature information. Specifically, the server uses image analysis software such as OpenCV or TensorFlow to identify features such as the collar, sleeves, and trouser hems of the clothing. For example, the analysis results in identifying the position of the collar and the length of the sleeves of a red top.

[1237] Step 5: Image synthesis using generative AI models

[1238] The server uses the analyzed feature information as input and generates a try-on image using a generative AI model (e.g., a generative adversarial network (GAN)). The input is the extracted feature information, and the output is the generated try-on image. Specifically, the server converts the feature information into a prompt and inputs it into the generative AI model. For example, the server generates a prompt such as, "Using the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png of the clothing selected by the user, please use the generative AI model to generate a try-on image." Based on this, the generative AI model generates an image that combines a red top and blue jeans.

[1239] Step 6: Save the composite image and generate a URL

[1240] The server saves the generated try-on image to a storage device and generates an accessible URL. The input is the generated try-on image, and the output is the URL of the saved try-on image. Specifically, the server saves the try-on image as a file and generates a unique URL for that file. For example, the generated URL might be https: / / example.com / generated_outfit.png.

[1241] Step 7: Send and view the composite image

[1242] The server sends the generated URL to the user's device, allowing the user to view the new outfit image on the device. The input is the generated URL, and the output is the try-on image displayed on the user's device. Specifically, the server returns a URL as an API response, and the user's device receives and displays this URL. For example, the user opens the URL https: / / example.com / generated_outfit.png in a browser, and the new outfit image is displayed.

[1243] (Application example 1)

[1244] 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."

[1245] On conventional e-commerce sites, users could not try on clothes before purchasing, making it difficult to confirm whether the item actually fit them. As a result, users had to go through the trouble of returning or exchanging the item after purchase if the size or design did not fit. In addition, it was difficult for users to visualize what the selected clothing would look like in real life, which led to a decrease in purchasing motivation.

[1246] 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.

[1247] In this invention, the server includes: means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site; means for acquiring the plurality of clothing images based on the identifier and the image URL and analyzing the characteristics of the clothing in each image; means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing images; means for the user to upload a full-body image of themselves and analyze the acquired clothing images and the user's full-body image; means for executing a generative AI model that generates a try-on image in which the selected clothing naturally fits the user's full-body image based on the analyzed characteristics; means for saving the generated try-on image and generating a URL for displaying it to the user; and means for transmitting the URL to the user terminal and enabling display on the user terminal. This allows the user to specifically check the try-on image based on their full-body image, allowing them to have a detailed image before purchasing, which increases their motivation to purchase and reduces the hassle of returns and exchanges due to mismatched sizes or designs.

[1248] A "user" is a consumer who uses an e-commerce site to select and purchase products.

[1249] An "e-commerce site" is a website that sells and buys goods and services over the Internet.

[1250] "Clothing Images" are digital images of clothing displayed on e-commerce sites.

[1251] An "identifier" is information for uniquely identifying a specific clothing image.

[1252] An "image URL" is an address (URL) for accessing a specific image file on the Internet.

[1253] A "full-body image" is a photograph showing the user's entire body.

[1254] "Feature analysis" is a technique for extracting specific elements or patterns from an image.

[1255] A "generative AI model" is an algorithm that uses artificial intelligence techniques to generate new images or data, such as generative adversarial networks (GANs).

[1256] The "try-on image" is a composite image of the user's whole body and a clothing image, simulating the appearance of the clothing when the user actually tries it on.

[1257] "Saving" means to hold the generated try-on image as data.

[1258] "URL generation" means creating a new URL to access the saved try-on image.

[1259] A "user terminal" is a device used by a user to access an e-commerce site, and includes a smartphone, tablet, computer, etc.

[1260] "Transmission" means sending data or information from the server to the user terminal.

[1261] "Display" refers to the visual presentation of information or images on the screen of a user terminal.

[1262] The present invention is a system that generates try-on images based on multiple clothing images selected by a user on an e-commerce site. The following describes how to implement this system step by step.

[1263] System Overview

[1264] The system consists of a server that runs on the backend of an e-commerce site and a terminal used by users. Users upload a full-body image of themselves and select clothing images on the site. The server analyzes this data and uses a generative AI model to generate try-on images and provide them to the user.

[1265] Hardware and software used

[1266] Smartphones, tablets, and computers: devices used by users to access

[1267] Server: Receives, stores, analyzes, and synthesizes data. Specifically, it uses cloud services such as AWS and Google Cloud Platform.

[1268] Generative AI models: Use generative adversarial networks (GANs) using PyTorch and TensorFlow.

[1269] Image analysis software: For example, OpenCV or Scikit-Image can be used to extract image features.

[1270] Data Flow and Processing

[1271] 1. User Action:

[1272] Users access an e-commerce site, select clothing items, and upload a full-body image of themselves.

[1273] 2. Data reception and processing by the server:

[1274] The server receives the identifier and image URL of the garment selected by the user, as well as a full-body image of the user.

[1275] 3. Image Acquisition and Feature Analysis:

[1276] The server retrieves the clothing image based on the identifier and the image URL, and analyzes the characteristics of the clothing and the user's full-body image using image analysis software.

[1277] 4. Generate fitting images:

[1278] Using a generative AI model (e.g., GAN using PyTorch or TensorFlow), a fitting image is generated that naturally combines the selected clothing with the user's full-body image based on the analyzed features.

[1279] 5. Save the generated image and generate a URL:

[1280] The generated try-on images are stored on the server, and a URL is generated that the user can access.

[1281] 6. User Notification and Display:

[1282] The server sends the generated URL to the user's device, and the user can open the URL to check the try-on image.

[1283] Examples of concrete examples and prompts

[1284] Specific examples

[1285] For example, if a user selects a "red top" and "blue jeans" and uploads a full-body image of themselves, the server will process the following:

[1286] 1. The server retrieves the image URLs https: / / example.com / red_top.png and https: / / example.com / blue_jeans.png for the selected clothing.

[1287] 2. The server receives the user's full-body image.

[1288] 3. The server analyzes the clothing image and the user image and extracts key features.

[1289] 4. Generate try-on images using a generative AI model.

[1290] 5. Save the try-on image and generate the URL https: / / example.com / generated_tryon.png.

[1291] 6. The user opens the URL on their device and checks the fitting image.

[1292] Prompt Sentence Examples

[1293] "Choose a red top and blue jeans and generate a natural try-on image based on the user's preferences. We use a GAN model to generate images that combine these clothes with the user's full-body photo."

[1294] This allows users to see a detailed try-on image before purchasing, which increases purchasing motivation and reduces the hassle of returning or exchanging items due to mismatched sizes or designs.

[1295] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1296] Step 1:

[1297] A user accesses an e-commerce site, selects the clothing item they want, and then takes or selects a full-body image from their smartphone or computer and uploads it.

[1298] Input: Clothing selection information (identifier, image URL), full-body image.

[1299] Output: Clothing selection data and full-body image data.

[1300] Specific operation: The user selects clothing through the interface of the e-commerce site and uploads a full-body image of themselves, which is then stored in cloud storage.

[1301] Step 2:

[1302] The server receives the user's clothing selection information (identifier and image URL) and a full-body image.

[1303] Input: Clothing selection data and full-body image data sent by the user.

[1304] Output: Selection data and whole body image stored on the server.

[1305] Specific operation: The server processes the received data and downloads the clothing image based on the identifier and image URL. The full-body image is stored in the database.

[1306] Step 3:

[1307] The server retrieves and analyzes the selected clothing image based on the identifier and the image URL.

[1308] Input: Clothing image URL.

[1309] Output: Feature data of the analyzed clothing image.

[1310] Specific operation: The server downloads the clothing image using the image URL and analyzes the features of the clothing image (e.g., shape, color, design) using OpenCV and Scikit-Image.

[1311] Step 4:

[1312] The server analyzes the user's full-body image and extracts key features of the image.

[1313] Input: A full-body image of the user.

[1314] Output: Feature data of the analyzed whole-body image.

[1315] Specific operation: The server analyzes the whole-body image using an image processing library such as OpenCV and extracts specific features such as contours and facial position.

[1316] Step 5:

[1317] The server uses the analyzed clothing images and feature data from the full-body image as input and generates a fitting image using a generative AI model (GAN).

[1318] Input: Feature data of analyzed clothing images, feature data of whole-body images.

[1319] Output: The generated try-on image.

[1320] Specific operation: The server uses PyTorch and TensorFlow to execute a generative AI model and generate a fitting image in which the selected clothing fits naturally to the user's full-body image.

[1321] Step 6:

[1322] The server saves the generated try-on image and generates a URL that the user can access.

[1323] Input: Generated try-on images.

[1324] Output: URL of the try-on image.

[1325] Specific operation: The fitting images are saved in a database and a URL for accessing them is generated. These URLs are managed in the database.

[1326] Step 7:

[1327] The server sends the generated URL to the user's terminal, and the user checks the fitting image on their own terminal.

[1328] Input: The URL of the generated try-on image.

[1329] Output: Try-on image displayed on the user's device.

[1330] Specific operation: The server sends the generated URL to the user's device, and the user opens the link to view the fitting image. The user interface is displayed in a browser or a dedicated app.

[1331] This allows the user to see a concrete image of the item before purchasing.

[1332] 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.

[1333] The present invention provides a system that receives the identifier and image URL of a clothing image selected from multiple clothing images selected by a user on an e-commerce site, and generates a try-on image through acquisition, analysis, synthesis, and display, and further combines it with an emotion engine to provide more personalized try-on images and recommended clothing images. Specific implementation of this system will be described below.

[1334] User selection of clothing

[1335] First, a user accesses an e-commerce site and selects the items they want from various categories (tops, pants, shoes, etc.). For example, if the user selects "red top" and "blue jeans," this selection information is sent from the user's terminal to the server.

[1336] Server receives selection information and acquires images

[1337] The server retrieves the clothing image from the specified URL based on the received identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads the image data from these URLs.

[1338] Image analysis and feature extraction

[1339] The server analyzes the acquired image data and extracts the key features of each garment (e.g., collar, sleeve, trouser leg, etc.). The feature information obtained through this analysis plays an important role in the subsequent synthesis process.

[1340] Image synthesis using generative AI models

[1341] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates an image of how the user would actually try on the clothing they selected. For example, it creates an image of a natural combination of a red top and blue jeans.

[1342] Use of emotion engine

[1343] Furthermore, the system uses an emotion engine to recognize the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at the try-on images to understand their emotional state. Based on this state, it adjusts the generation of try-on images and clothing recommendations.

[1344] Save composite image and generate URL

[1345] The generated try-on image is saved by the server and a URL is generated so that users can access it. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png.

[1346] Sending and displaying composite images

[1347] The server sends the generated URL to the user's device, and when the user opens this URL on their device, an image of the new outfit is displayed. The user can check this image and decide whether to actually purchase it.

[1348] Display of recommended clothing

[1349] Furthermore, based on the emotional information recognized by the emotion engine, the system provides recommended clothing images according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[1350] Specific examples

[1351] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[1352] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[1353] 2. When the user confirms their selection, the device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[1354] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[1355] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[1356] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[1357] 6. The emotion engine analyzes the user's reaction and adds similar items to the recommendation list based on the generated try-on images.

[1358] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[1359] 8. The server sends this URL to the user's device.

[1360] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[1361] This process allows users to get a concrete image of how the clothes will look before they buy, and also provides personalized recommendations based on their emotions, which increases the success rate of purchases and reduces the hassle of returns.

[1362] The processing flow will be explained below.

[1363] Step 1:

[1364] A user accesses an e-commerce site and selects desired items from clothing categories such as tops and pants. For example, the user selects "red top" and "blue jeans."

[1365] Step 2:

[1366] Once the user confirms their selection, the device sends the selected garment's identifier (ID) and image URL to the server. For example, the ID of the red top is top123, the ID of the blue jeans is jeans456, and the image URLs are https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[1367] Step 3:

[1368] Based on the received identifier and image URL, the server retrieves the clothing images from the specified URL. The server downloads the corresponding image data from https: / / example.com / top123.png and https: / / example.com / jeans456.png.

[1369] Step 4:

[1370] The server analyzes the captured image data and extracts key features of each garment, such as the position of the collar and sleeves on a top, or the waist and hem on jeans.

[1371] Step 5:

[1372] The server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. This process generates a try-on image of the user's chosen clothing.

[1373] Step 6:

[1374] The emotion engine recognizes the user's reactions and emotions. For example, it analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when looking at images of items to try on, and understands their emotional state.

[1375] Step 7:

[1376] Based on the user's emotions recognized by the emotion engine, the generated try-on images are adjusted to create a more suitable outfit for the user. For example, if the user expresses positive emotions, the system generates images that emphasize that style.

[1377] Step 8:

[1378] The server saves the generated try-on image and generates a URL that users can access. For example, the URL for the generated try-on image might be https: / / example.com / generated_outfit.png.

[1379] Step 9:

[1380] The server sends the generated URL to the user's device, which then receives the URL and prepares it so that the user can check the outfit image.

[1381] Step 10:

[1382] The device uses the received URL to display the new outfit image on the user interface, allowing the user to see how the item will look when worn and decide whether to purchase it.

[1383] Step 11:

[1384] Based on the emotional information recognized by the emotion engine, recommended clothing images are provided according to the user's preferences. For example, if the user has a positive reaction to an image of the clothes being tried on, clothing of a similar style will be displayed in the recommendation list.

[1385] This specific processing step allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions, thereby increasing the success rate of purchases and reducing the hassle of returns.

[1386] Example 2

[1387] 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."

[1388] On modern e-commerce sites, users have limited options for trying on clothing before purchasing. This can lead to an increase in returns and exchanges, as the product may not meet expectations after purchase. Furthermore, the lack of personalized recommendations based on users' emotions and preferences makes it difficult to increase user satisfaction.

[1389] 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.

[1390] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site, means for acquiring the plurality of clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for saving the generated outfit image and generating a URL for displaying it to the user, means for transmitting the URL to the user terminal and enabling display on the user terminal, and means for recognizing the user's reactions and emotions using an emotion engine and adjusting the generation of try-on images or clothing recommendations based on the user's emotional state. This allows the user to see specific try-on images before purchasing and receive personalized recommendations based on their emotions.

[1391] A "user" is someone who uses this system to select clothing on an e-commerce site and check how it looks when worn.

[1392] An "e-commerce site" is a website for selling and purchasing goods over the Internet.

[1393] A "clothing image" is image data that shows a visual representation of the clothing selected by the user.

[1394] An "identifier" is an ID or code that uniquely identifies a particular clothing image.

[1395] An "image URL" is an address that indicates where a specific clothing image is stored on the Internet.

[1396] The "server" is a computer system that receives information sent by the user, acquires and analyzes images, and synthesizes images for trying on.

[1397] "Image analysis" is a process for extracting feature information from acquired clothing images.

[1398] "Features" are information that indicates specific parts, shapes, patterns, etc. in a clothing image.

[1399] A "generative AI model" is an artificial intelligence algorithm that synthesizes new try-on images based on acquired feature information.

[1400] A "try-on image" is a visual image generated by combining multiple pieces of clothing selected by the user.

[1401] A "URL" is an internet address used to access a particular web page or resource.

[1402] An "emotion engine" is an artificial intelligence system that recognizes users' reactions and emotions and analyzes that data.

[1403] "Reactions" refer to the user's actions, facial expressions, and interactions such as clicking and scrolling when looking at the try-on images.

[1404] "Recommendation" is the process of selecting relevant clothing items based on a user's preferences.

[1405] This system receives the identifier and image URL of a clothing image selected by a user from multiple clothing images selected on an e-commerce site, and generates try-on images through acquisition, analysis, synthesis, and display. Furthermore, by combining an emotion engine, it provides more personalized try-on images and recommended clothing images.

[1406] Hardware and software used

[1407] The server is a computer system with high-performance computing capabilities for receiving data, acquiring images, analyzing them, storing data, and running the generated AI model. For specific implementation, it is recommended to use deep learning libraries such as TensorFlow or PyTorch, and OpenCV for image analysis.

[1408] The user's terminal is typically a smartphone, tablet, or personal computer that accesses the e-commerce site via a web browser.

[1409] An emotion engine requires an algorithm to analyze user reactions and emotions and predict purchasing behavior. Specifically, an analytical model using machine learning is considered.

[1410] Program processing

[1411] A user accesses an e-commerce site and selects the clothing item they want. For example, if the user selects a "red top" and "blue jeans," this selection information is sent from the user's terminal to the server along with an identifier and an image URL.

[1412] The server retrieves clothing images from the specified URLs based on the identifier (ID) and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads images from these URLs.

[1413] The acquired images are analyzed by the server to extract key features using image processing libraries such as OpenCV. After the feature information is extracted, the server runs a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize a try-on image based on the feature information.

[1414] The generated try-on image is saved by the server and a URL that can be accessed by the user is generated. For example, the URL of the generated try-on image is created as https: / / example.com / generated_outfit.png, and the server generates this URL and saves it in the database.

[1415] The server then sends the generated URL to the user's device. When the user opens the URL, an image of the new outfit is displayed. The user can then review the image and decide whether to actually purchase it.

[1416] Furthermore, the server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, the server can understand the user's emotional state from their facial expressions when looking at the try-on images, as well as their actions such as clicking and scrolling, and adjust the generation of try-on images or clothing recommendations based on this.

[1417] Specific examples

[1418] For example, if a user selects a new denim jacket and black jeans, the process might go like this:

[1419] 1. A user selects a "denim jacket" and "black jeans" on an e-commerce site.

[1420] 2. The device sends the IDs jacket123 and jeans456 and the image URLs https: / / example.com / jacket123.png and https: / / example.com / jeans456.png to the server.

[1421] 3. The server accesses the sent URL and downloads images of a denim jacket and black jeans from https: / / example.com / jacket123.png and https: / / example.com / jeans456.png.

[1422] 4. The server performs image analysis to identify the collar and sleeves of the denim jacket and the waist and hem of the black jeans.

[1423] 5. The server uses the generative AI model to combine the denim jacket with the black jeans to generate a new try-on image.

[1424] 6. The emotion engine analyzes the user's reactions and adds similar items to the recommendation list based on the generated try-on images.

[1425] 7. The server saves the generated URL containing the try-on image and recommendation list as https: / / example.com / generated_outfit123456.png.

[1426] 8. The server sends this URL to the user's device.

[1427] 9. The device displays the received URL https: / / example.com / generated_outfit123456.png, and the user sees the new outfit image and recommendation list.

[1428] Using this system, users can easily get a concrete try-on image before purchasing and receive personalized recommendations based on their emotions.

[1429] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1430] Step 1:

[1431] A user accesses an e-commerce site and selects the clothing they want. For example, they select a "red top" and "blue jeans," and then the information (identifier and image URL) is sent from the terminal to the server. The input is the user's selection information, and the output is the clothing identifier and image URL sent to the server.

[1432] Step 2:

[1433] The server retrieves clothing images from the specified URLs based on the identifier and image URL received from the user. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, the server downloads image files from these URLs. The input is the identifier and image URL sent in the previous step, and the output is the downloaded clothing image.

[1434] Step 3:

[1435] The server analyzes the acquired image data and extracts key clothing features. This analysis process uses image processing libraries such as OpenCV. For example, it identifies features such as the collar and sleeves of a red top, or the hem and waist of blue jeans. The input is the downloaded clothing image, and the output is the analyzed feature information.

[1436] Step 4:

[1437] The server uses a generative AI model (e.g., a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. For example, it generates a try-on image of a red top paired with blue jeans. This process uses deep learning libraries such as TensorFlow and PyTorch. The input is the analyzed feature information, and the output is the generated try-on image.

[1438] Step 5:

[1439] The server saves the generated try-on image and generates a URL to make the image accessible. For example, the URL of the generated try-on image is https: / / example.com / generated_outfit.png. The input is the generated try-on image, and the output is the URL of the generated try-on image.

[1440] Step 6:

[1441] The server sends the generated URL to the user's device. When the user accesses this URL on their device, the generated try-on image is displayed. For example, if the user accesses https: / / example.com / generated_outfit.png in their web browser, they can see the synthesized try-on image. The input is the generated URL, and the output is the try-on image displayed on the user's device.

[1442] Step 7:

[1443] The server uses an emotion engine to recognize the user's reactions and emotions and analyzes the data. For example, it collects and analyzes behavioral data such as facial expressions, clicks, and scrolls when the user looks at images of items to try on. The input is the user's reaction data, and the output is analyzed emotional information.

[1444] Step 8:

[1445] The server generates try-on images and adjusts clothing recommendations based on the analyzed emotional information. For example, if the user responds positively, clothing of a similar style is added to the recommendation list. The input is the analyzed emotional information, and the output is the adjusted try-on images and recommendation list.

[1446] This system allows users to see specific try-on images and receive personalized recommendations based on their emotions.

[1447] (Application example 2)

[1448] 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."

[1449] Conventional e-commerce sites lack an effective way to provide users with try-on images of the clothing they select. This leads to users purchasing clothing without visual uncertainty, resulting in a high return rate. Furthermore, the lack of personalized clothing recommendations based on users' emotions and preferences limits the user experience. This creates a need for a system that can facilitate purchasing decisions.

[1450] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1451] In this invention, the server includes means for receiving an identifier and an image URL of a selected clothing image from multiple clothing images selected by a user on an e-commerce site, means for acquiring the multiple clothing images based on the identifier and image URL and analyzing the characteristics of the clothing in each image, means for executing a generative AI model that uses the acquired clothing images to synthesize a generated outfit image, means for recognizing the user's emotional state using an emotion engine and adjusting the generation of try-on images and clothing recommendations based on that state, means for saving the generated outfit images and generating a URL for displaying them to the user, and means for transmitting the URL to a user terminal so that the URL can be displayed on the user terminal. This allows the user to visually check the try-on images and receive personalized recommendations based on the emotion analysis results.

[1452] "User" refers to a customer or consumer who uses the e-commerce site.

[1453] "E-commerce site" means a website where purchases of goods and services are made over the Internet.

[1454] "Garment image" refers to a digital image that shows the visual characteristics of a garment.

[1455] "Identifier" refers to an ID or code that uniquely identifies each clothing image.

[1456] "Image URL" refers to the web address where the clothing image is stored on the Internet.

[1457] "Acquisition" refers to the act of downloading clothing images and related information from an e-commerce site or server.

[1458] "Analysis" refers to the procedure of detecting and extracting features and patterns within clothing images.

[1459] "Generative AI model" refers to an artificial intelligence (AI) algorithm for generating new try-on images from image data.

[1460] "Synthesis" refers to the act of combining multiple clothing images to create a new try-on image.

[1461] "Try-on image" refers to a visual display of a user virtually trying on the clothing combination they have selected.

[1462] An "emotion engine" refers to technology for detecting and analyzing a user's emotions and reactions.

[1463] "Saving" refers to the act of recording the generated try-on image as data on a server or the like.

[1464] "URL generation" refers to the process of creating a web address that allows users to access the generated try-on images.

[1465] "Sending" refers to the act of transferring the URL of the generated try-on image to the user's device.

[1466] "Display" refers to the act of visually presenting try-on images and recommended clothing images generated on the user terminal.

[1467] "Personalization" refers to the act of providing individually tailored services and recommendations based on a user's preferences and emotions.

[1468] "Recommended Clothing Images" refers to additional clothing images that are recommended based on the user's reactions or emotional state.

[1469] The present invention is a system that generates try-on images from multiple clothing images selected by a user on an e-commerce site and uses an emotion engine to provide the user with personalized try-on images and clothing recommendations. A specific embodiment of the present invention will be described below.

[1470] First, a user visits an e-commerce site and selects the desired clothing item from various categories (tops, pants, shoes, etc.). The identifier and image URL of the selected clothing item are sent from the user's device to the server. This information is sent along with the identifier (ID) and image URL. For example, if the image URL for a red top is https: / / example.com / red_top.png and the image URL for blue jeans is https: / / example.com / blue_jeans.png, both of these URLs are sent to the server.

[1471] Next, the server retrieves clothing images from the specified URL based on the received identifier and image URL. The server downloads these image data, for example, image data of a red top and blue jeans. After this, the server analyzes the acquired image data and extracts the main features of each garment (collar, sleeves, hem, etc.). The analyzed feature information plays an important role in the subsequent synthesis process.

[1472] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize each clothing image based on the analyzed feature information. This generates an image of how the user would actually try on the clothing they selected. For example, an image of a natural combination of a red top and blue jeans can be generated.

[1473] Furthermore, an emotion engine is used to analyze the user's reactions and emotions. The system analyzes the user's facial expressions and actions (clicks, scrolls, etc.) when viewing the try-on images to understand their emotional state. Based on these results, the system adjusts the generation of try-on images and clothing recommendations.

[1474] The generated try-on image is saved by the server and a URL is generated so that the user can access it. For example, the URL for the generated try-on image is created as https: / / example.com / generated_outfit.png. The server sends this URL to the user's device, and the user can check the new outfit image on their device.

[1475] Based on the emotional information recognized by the emotion engine, recommended clothing images are also provided according to the user's preferences. For example, if the user has a positive reaction to a fitting image, clothing of a similar style will be displayed in the recommendation list.

[1476] As a concrete example, consider a user who selects a new denim jacket and black jeans. When the user selects "denim jacket" and "black jeans" on an e-commerce website, their respective IDs, jacket123 and jeans456, and their image URLs, https: / / example.com / jacket123.png and https: / / example.com / jeans456.png, are sent to the server. The server downloads the image data of the denim jacket and black jeans from these URLs, analyzes them, and synthesizes them using a generative AI model to generate new try-on images. The emotion engine analyzes the user's reaction and adds similar items to a recommendation list based on the generated try-on images. The server saves the URLs containing the generated try-on images and recommendation list and sends them to the user's device. The user can view the new outfit images and recommendation list.

[1477] An example of a prompt sentence would be a specific instruction input to the generative AI model, such as "Generate an image of someone trying on a denim jacket and black jeans."

[1478] This allows users to see specific clothing images before purchasing and receive personalized recommendations based on their emotions.

[1479] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1480] Step 1:

[1481] A user accesses an e-commerce site and selects a desired clothing item. The user then sends data including the identifier and image URL of the selected clothing item from the terminal to the server. The input includes the identifiers and image URLs of multiple clothing images, and by sending these to the server, information on the clothing selected by the user is collected.

[1482] Step 2:

[1483] The server retrieves clothing images from the specified URLs based on the received identifier and image URL. The server accesses each URL and downloads clothing image data. The identifier and image URL are input, and the required clothing images are prepared by retrieving image data based on these.

[1484] Step 3:

[1485] The server analyzes the acquired clothing images and extracts the key features of each garment (collar, sleeves, hem, etc.). The server uses an image analysis algorithm to identify the features of each clothing image. The downloaded image data is used as input and analysis is performed to obtain feature information.

[1486] Step 4:

[1487] The server uses a generative AI model (such as a generative adversarial network (GAN)) to synthesize clothing images based on the analyzed feature information. It then generates an image of what it would look like if the user tried on the selected clothing. In this process, a specific prompt (e.g., "Generate an image of me trying on a denim jacket and black jeans") is input to the model. The input is the analyzed feature information and the prompt, and the generated try-on image is output.

[1488] Step 5:

[1489] The emotion engine analyzes the user's reactions and emotions when viewing images of clothing to try on. The server collects the user's facial expressions and actions (clicks, scrolls, etc.) through cameras and interaction data to recognize the user's emotional state. The input is the user's reaction data, and the output is their emotional state.

[1490] Step 6:

[1491] The server saves the generated try-on images and generates a URL to display them to the user. The server saves the generated outfit image data and creates a corresponding access URL. The input is the generated try-on image, and the output is the save destination URL.

[1492] Step 7:

[1493] The server sends the URL to the user's device, enabling it to be displayed on the user's device. The user uses the received URL to check the new outfit image. The input is the destination URL, and the output is the display on the user's device.

[1494] Step 8:

[1495] Based on the emotion information recognized by the emotion engine, the server, which provides recommended clothing images according to the user's preferences, creates a recommendation list and generates a URL for the recommended products so that the user can access them. The user can refer to the recommendation list to check related items. The input is emotion information and a recommendation algorithm, and the output is a recommendation list and its URL.

[1496] The above are the specific processing steps for processing from the user terminal to the server, generating try-on images, analyzing emotions, and the recommendation system.

[1497] 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.

[1498] 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.

[1499] 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.

[1500] 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.

[1501] 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.

[1502] 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.

[1503] 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).

[1504] 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.

[1505] 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."

[1506] 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.

[1507] 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).

[1508] 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.

[1509] 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.

[1510] 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.

[1511] 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.

[1512] 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.

[1513] 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.

[1514] 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.

[1515] 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.

[1516] 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.

[1517] 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.

[1518] The following is further disclosed regarding the above embodiment.

[1519] (Claim 1)

[1520] A means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site;

[1521] A means for acquiring the plurality of clothing images based on the identifier and image URL, and analyzing the characteristics of the clothing in each image;

[1522] A means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing image;

[1523] a means for saving the generated outfit image and generating a URL for displaying the outfit image to a user;

[1524] means for transmitting the URL to the user terminal and enabling it to be displayed on the user terminal;

[1525] A system including:

[1526] (Claim 2)

[1527] 2. The system according to claim 1, further comprising a means for managing the storage of the generated outfit images for each user.

[1528] (Claim 3)

[1529] The system of claim 1 , further comprising a generative AI model for realizing a natural arrangement of the clothing images based on the analyzed features.

[1530] "Example 1"

[1531] (Claim 1)

[1532] A means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site;

[1533] A means for acquiring the plurality of clothing images based on the identifier and image URL, and analyzing the characteristics of the clothing in each image;

[1534] A means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing image;

[1535] a means for saving the generated outfit image and generating a URL for displaying the outfit image to a user;

[1536] means for transmitting the URL to the user terminal and enabling it to be displayed on the user terminal;

[1537] A means for generating prompt sentences for the AI ​​model to generate fitting images based on a user's input of a desired clothing combination;

[1538] A system including:

[1539] (Claim 2)

[1540] 2. The system according to claim 1, further comprising a means for managing the storage of the generated outfit images for each user.

[1541] (Claim 3)

[1542] The system of claim 1 , further comprising a generative AI model for realizing a natural arrangement of the clothing images based on the analyzed features.

[1543] "Application Example 1"

[1544] (Claim 1)

[1545] A means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site;

[1546] A means for acquiring the plurality of clothing images based on the identifier and image URL, and analyzing the characteristics of the clothing in each image;

[1547] A means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing image;

[1548] A means for allowing a user to upload a full-body image of himself / herself and for analyzing the acquired clothing image and the user's full-body image;

[1549] A means for executing a generative AI model that generates a fitting image in which the selected clothing is naturally fitted to the user's whole body image based on the analyzed features;

[1550] a means for saving the generated try-on image and generating a URL for displaying it to a user;

[1551] means for transmitting the URL to the user terminal and enabling it to be displayed on the user terminal;

[1552] A system including:

[1553] (Claim 2)

[1554] 2. The system according to claim 1, further comprising means for managing the storage of the generated try-on images for each user.

[1555] (Claim 3)

[1556] 10. The system of claim 1, further comprising a generative AI model for achieving a natural placement of selected clothing on a full-body image of a user based on the analyzed features.

[1557] "Example 2: Combining Emotion Engines"

[1558] (Claim 1)

[1559] A means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site;

[1560] A means for acquiring the plurality of clothing images based on the identifier and image URL, and analyzing the characteristics of the clothing in each image;

[1561] A means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing image;

[1562] a means for saving the generated outfit image and generating a URL for displaying the outfit image to a user;

[1563] means for transmitting the URL to the user terminal and enabling it to be displayed on the user terminal;

[1564] a means for recognizing a user's reactions and emotions using an emotion engine and adjusting the generation of try-on images or clothing recommendations based on the user's emotional state;

[1565] A system including:

[1566] (Claim 2)

[1567] 2. The system according to claim 1, further comprising a means for managing the storage of the generated outfit images for each user.

[1568] (Claim 3)

[1569] The system of claim 1 , further comprising a generative AI model for realizing a natural arrangement of the clothing images based on the analyzed features.

[1570] "Application example 2 when combining emotion engines"

[1571] (Claim 1)

[1572] A means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site;

[1573] A means for acquiring the plurality of clothing images based on the identifier and image URL, and analyzing the characteristics of the clothing in each image;

[1574] A means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing image;

[1575] a means for recognizing a user's emotional state using an emotion engine and adjusting the generation of try-on images and clothing recommendations based on the state;

[1576] a means for saving the generated outfit image and generating a URL for displaying the outfit image to a user;

[1577] means for transmitting the URL to the user terminal and enabling it to be displayed on the user terminal;

[1578] A system including:

[1579] (Claim 2)

[1580] 2. The system according to claim 1, further comprising means for managing the storage of the generated outfit images for each user, and for providing personalized recommended clothing images based on the result of emotion analysis.

[1581] (Claim 3)

[1582] The system of claim 1 further comprising a generative AI model for realizing a natural arrangement of the clothing images based on the analyzed features, and an emotion engine for analyzing the user's emotional state. [Explanation of symbols]

[1583] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for receiving an identifier and an image URL of a selected clothing image from a plurality of clothing images selected by a user on the e-commerce site; A means for acquiring the plurality of clothing images based on the identifier and image URL, and analyzing the characteristics of the clothing in each image; A means for executing a generative AI model that synthesizes a generated outfit image using the acquired clothing image; a means for saving the generated outfit image and generating a URL for displaying the outfit image to a user; means for transmitting the URL to the user terminal and enabling it to be displayed on the user terminal; A system including:

2. The system according to claim 1 , further comprising a means for managing the storage of the generated outfit images for each user.

3. The system of claim 1 , further comprising a generative AI model for realizing a natural arrangement of the clothing images based on the analyzed features.

Images