System

Abstract

A system is provided.SOLUTION: A system, comprising: means for a user to take and upload pictures of the user from multiple angles; means for inputting a desired decoration; a generative model for generating a virtual fitting image of the user wearing the decoration based on the received pictures of the user and the desired decoration design; means for providing the generated virtual fitting image to the user; means for the user to issue a modification instruction for the virtual fitting image; means for re-executing the generative model based on the modification instruction to update the virtual fitting image; and means for determining and storing a finally selected decoration design.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Conventional try-on services are limited by time and the number of items that can be tried on, making it difficult for users to find the perfect accessory. Furthermore, users are unable to check and modify the design of accessories in detail, which can lead to low user satisfaction with their final choice. Furthermore, the need to visit a specific location makes the try-on process cumbersome. [Means for solving the problem]

[0005] The present invention provides a system that allows users to upload photos of themselves taken from multiple angles and input their desired decorative design. Based on the received user photo and desired decorative design, a generative model generates a virtual try-on image of the user wearing the decorative item. The system also has a function that provides the generated virtual try-on image to the user, and allows the user to issue correction instructions for the virtual try-on image. The system includes a means for re-executing the generative model based on the correction instructions to update the virtual try-on image, and for finalizing and saving the selected decorative design. This allows users to try on an unlimited number of decorative items from the comfort of their own home and find the perfect look.

[0006] "User" refers to an individual who utilizes the System to upload photos and try on desired decorations.

[0007] "Photos" refer to image data that users take of themselves from multiple angles and upload to the system.

[0008] "Decoration" refers to the design of the clothing or accessories that the user wishes to try on, such as a dress or kimono.

[0009] "Upload" refers to the operation of sending a photograph or decorative design taken by a user to the system.

[0010] "Generative model" refers to an artificial intelligence model that generates a virtual image of a user wearing an ornament based on the user's photograph and the ornament's design.

[0011] A "virtual try-on image" refers to an image generated by a generative model that virtually depicts a user wearing a specific accessory.

[0012] "Modification instructions" refer to instructions that a user gives to request changes or fine adjustments to the design of the virtual try-on image.

[0013] "Final selection" refers to the operation in which the user checks the virtual try-on images and confirms the decorative design they like.

[0014] "Storage" refers to the operation of the server safely storing data such as received photographs, decorative designs, virtual try-on images, and correction instructions.

[0015] "System" refers to a set of components that provide a series of functions, from uploading a user's photo, virtually trying on accessories, receiving correction instructions, to making a final selection. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0017] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0019] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0020] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0021] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0022] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0024] [First embodiment]

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

[0026] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0027] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0029] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0030] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0033] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0034] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0037] The present invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative model.

[0038] System configuration and operation

[0039] 1. User image upload

[0040] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[0041] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[0042] 2. Input your desired decorative design

[0043] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[0044] 3. Generation of virtual try-on images

[0045] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[0046] 4. Check and edit the virtual try-on image

[0047] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[0048] The server receives the correction instructions and re-inputs them into the generative model. The generative model generates a corrected virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0049] 5. Final dress selection

[0050] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[0051] Specific examples

[0052] If user A wants to choose a wedding dress, he can use the system as follows:

[0053] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[0054] 2. The server receives these photos and stores them in storage.

[0055] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[0056] 4. The server receives this data and inputs it into the generative model.

[0057] 5. The generative model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[0058] 6. The server provides the generated virtual try-on image to User A.

[0059] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[0060] 8. The server re-inputs the correction instructions into the generative model, generates an updated image, and provides it to User A.

[0061] 9. User A finally chooses the dress she likes and confirms it on the system.

[0062] 10. The server saves the selection and completes the process.

[0063] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[0064] The processing flow will be explained below.

[0065] Specific steps in the system programming process

[0066] Image upload

[0067] Step 1:

[0068] Users take photos of themselves from multiple angles and use a terminal to access the system.

[0069] Step 2:

[0070] The terminal allows the user to select a photo they have taken and send it to the server by pressing the upload button.

[0071] Step 3:

[0072] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[0073] Decorative design input

[0074] Step 4:

[0075] The user inputs the desired decorative design in the form of image data, text instructions, or both.

[0076] Step 5:

[0077] The terminal transmits the input decorative design data to the server.

[0078] Step 6:

[0079] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[0080] Virtual try-on image generation

[0081] Step 7:

[0082] The server inputs the received user image and desired decorative design into the generative model.

[0083] Step 8:

[0084] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[0085] Step 9:

[0086] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[0087] Step 10:

[0088] The server receives the generated virtual try-on images and stores them in the user's account.

[0089] Check and edit virtual try-on images

[0090] Step 11:

[0091] The user checks the generated virtual try-on image on the terminal.

[0092] Step 12:

[0093] If a specific part of the image needs to be corrected, the user inputs a correction instruction.

[0094] Step 13:

[0095] The terminal sends a correction instruction to the server.

[0096] Step 14:

[0097] The server re-inputs the received correction instructions into the generative model.

[0098] Step 15:

[0099] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[0100] Step 16:

[0101] The server receives the updated virtual try-on images and stores them in the user's account.

[0102] Step 17:

[0103] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[0104] Final Selection

[0105] Step 18:

[0106] The user finally selects the decorative design they like on the terminal.

[0107] Step 19:

[0108] The terminal transmits the final selected decorative design information to the server.

[0109] Step 20:

[0110] The server determines the selected decorative design and stores the associated data in storage.

[0111] Step 21:

[0112] The server sends the user a confirmation of the final decision, informing them that their dress selection is complete.

[0113] Example 1

[0114] 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."

[0115] Conventional try-on systems require users to physically visit a store, which requires time and effort. Even online try-on systems have the drawback of making it difficult for users to make fine adjustments when trying on accessories, making it difficult to fully replicate the actual wearing experience. Furthermore, there is a lack of a way for users to check the accessories in detail from multiple angles.

[0116] 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.

[0117] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting a desired ornament design, and a generative model for generating a virtual try-on image of the user wearing the ornament based on the received user image and the desired ornament design. This allows users to try on an unlimited number of ornaments from the comfort of their own home and find their perfect outfit. The generated virtual try-on image can also be viewed from a 360-degree angle, making it easy for users to check the details of the desired ornament. Furthermore, by re-executing the generative model based on correction instructions and updating the virtual try-on image, users can adjust the ornament design to their satisfaction.

[0118] "User images" are photographic data taken by a user of himself or herself, taken from multiple angles.

[0119] A "desired accessory design" is a design of accessory or clothing that the user wishes to try on, and may take the form of image data or text instructions.

[0120] A "generative model" is an artificial intelligence model that generates virtual try-on images using a user's image and desired decorative design as input, and has image generation and editing functions.

[0121] A "virtual try-on image" is an image generated by a generative model that virtually recreates the appearance of a user wearing an accessory.

[0122] "Modification instructions" are requests for changes or adjustments made by the user to the virtual try-on image, and include specific changes to parts or styles.

[0123] "360-degree angle" means that the virtual try-on image can be viewed from all directions, allowing the user to check the decorations from multiple angles.

[0124] "Server" means a centralized computer system that receives, stores, and processes images and data from users.

[0125] This invention relates to a system that allows users to virtually try on accessories by taking pictures of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative AI model.

[0126] System configuration and operation

[0127] 1. User image upload

[0128] Users take pictures of themselves from the front, side, and back, and then access the system using a terminal. Users select the images they have taken and press the upload button to send them to the server.

[0129] The server stores the received user image data in cloud storage, confirms that the image has been uploaded correctly, and notifies the user.

[0130] Hardware used: smartphone, PC, cloud server

[0131] Software used: web browser, photo-taking application, cloud storage system (e.g., Amazon S3)

[0132] 2. Input your desired decorative design

[0133] The user inputs the desired decorative design as image data or text instructions. The terminal transmits this data to the server,

[0134] The server stores the received design data in cloud storage.

[0135] Hardware used: smartphone, PC, cloud server

[0136] Software used: Web forms, text input interface

[0137] 3. Generation of virtual try-on images

[0138] The server inputs the received user image and desired decorative design into a generative AI model.

[0139] The generative AI model uses this data to generate virtual try-on images of the user wearing the accessories, which can be viewed from a 360-degree angle.

[0140] The server provides this image to the user.

[0141] Hardware used: Cloud server

[0142] Software used: Generative AI models (e.g., DALL-E, Stable Diffusion), databases

[0143] 4. Check and edit the virtual try-on image

[0144] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[0145] The server receives the correction instructions and feeds them back into the generative AI model, which then generates a modified virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0146] Hardware used: smartphone, PC, cloud server

[0147] Software used: Generative AI models, web forms

[0148] 5. Final selection of decorative design

[0149] The user finally selects the decorative design they like and notifies the server via their terminal.

[0150] The server stores the information and confirms the final decision.

[0151] Hardware used: smartphone, PC, cloud server

[0152] Software used: Web forms, databases

[0153] Specific examples

[0154] If user A wants to choose a wedding dress, he can use the system as follows:

[0155] 1. User A takes pictures of himself from the front, side, and back and uploads them to the system.

[0156] Example prompt: "Please upload a photo of your front, side, and back."

[0157] 2. The server receives these images and stores them in cloud storage.

[0158] Example prompt: "Your image was successfully uploaded."

[0159] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[0160] Example prompt: "Please upload a photo of the dress you would like."

[0161] 4. The server receives this data and inputs it into a generative AI model.

[0162] 5. The generative AI model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[0163] 6. The server provides the generated virtual try-on image to User A.

[0164] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[0165] Example prompt: "Please make the hem a little shorter."

[0166] 8. The server re-inputs the correction instructions into the generative AI model, generates an updated image, and provides it to User A.

[0167] 9. User A finally chooses the dress she likes and confirms it on the system.

[0168] 10. The server saves the selection and completes the process.

[0169] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[0170] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0171] Step 1: Take and upload a user image

[0172] Specific behavior:

[0173] Users take photos of themselves from the front, side, and back using a smartphone or camera.

[0174] Save the captured image file to a folder on your device.

[0175] Input: Captured image files (front, side, back)

[0176] Output: Image file saved on the device

[0177] Step 2: Access the system and upload your images

[0178] Specific behavior:

[0179] The user opens a web browser on the terminal, accesses the system's web application, and logs in.

[0180] Access the image upload form, select the image file you took, and press the upload button.

[0181] Input: Image files stored on the device, user login information

[0182] Output: Image file sent to the server

[0183] Step 3: The server receives and stores the image data.

[0184] Specific behavior:

[0185] The server saves the image data sent by the user in cloud storage and records the file path of the saved data in a database.

[0186] Verify that the image was uploaded correctly, generate a success message, and notify the user.

[0187] Input: Image data sent by the user

[0188] Output: Image data saved in cloud storage, success message

[0189] Step 4: User inputs desired decorative design

[0190] Specific behavior:

[0191] The user inputs image data and text instructions with the desired decorative design into the terminal, for example, by uploading an image of the dress taken with a smartphone.

[0192] Access the input form, select the design data, and press the submit button.

[0193] Input: Image data of decorative design taken with a smartphone or text instructions

[0194] Output: Decorative design data sent to the server

[0195] Step 5: The server receives and stores the design data

[0196] Specific behavior:

[0197] The server saves the decorative design data sent by the user in cloud storage, and records the file path and text content of the saved data in a database.

[0198] Input: Decoration design data submitted by the user

[0199] Output: Design data stored in cloud storage, file paths and text contents recorded in the database

[0200] Step 6: Start generating virtual try-on images

[0201] Specific behavior:

[0202] The server performs preprocessing to input the image data captured by the user and the desired decorative design data into the generative AI model.

[0203] The data is converted into the appropriate format and fed into a generative AI model.

[0204] Input: User image data stored in cloud storage, decorative design data stored in cloud storage

[0205] Output: The data that is fed into the generative AI model

[0206] Step 7: The generative AI model generates virtual try-on images

[0207] Specific behavior:

[0208] The generative AI model generates virtual try-on images based on the input user image data and decorative design data.

[0209] The generated try-on images are processed so that they can be displayed from multiple angles.

[0210] Input: User image data and decorative design data input into the generative AI model

[0211] Output: Generated virtual try-on image

[0212] Step 8: The server stores and serves the generated images.

[0213] Specific behavior:

[0214] The server saves the virtual try-on images sent from the generative AI model to cloud storage and records the file paths of the saved images in a database.

[0215] The user is notified of the URL of the generated virtual try-on image.

[0216] Input: Virtual try-on images sent from the generative AI model

[0217] Output: Virtual try-on images saved in cloud storage, notification to user

[0218] Step 9: The user checks the fitting image and inputs correction instructions.

[0219] Specific behavior:

[0220] The user can check the virtual fitting images provided on the device and input specific correction instructions as needed, such as "Please make the hem a little shorter."

[0221] Input: Provided virtual try-on image, modification instructions

[0222] Output: Correction instructions sent to the server

[0223] Step 10: The server re-inputs the correction instructions into the generative AI model.

[0224] Specific behavior:

[0225] The server performs preprocessing to re-input the received correction instructions into the generative AI model.

[0226] The correction instructions are input into the generative AI model, requesting the generation of an updated virtual try-on image.

[0227] Input: Correction instructions sent to the server

[0228] Output: Corrective instruction data that is input to the generative AI model

[0229] Step 11: The generative AI model generates the corrected image

[0230] Specific behavior:

[0231] The generative AI model generates new virtual try-on images based on the correction instructions and sends them to the server.

[0232] Input: Correction instructions input to the generative AI model

[0233] Output: Corrected virtual try-on image

[0234] Step 12: The server stores and serves the corrected image.

[0235] Specific behavior:

[0236] The server saves the newly generated virtual try-on image in cloud storage, records the file path of the saved image in the database, and notifies the user of the URL of the modified virtual try-on image.

[0237] Input: Modified virtual try-on image sent from the generative AI model

[0238] Output: Edited try-on image saved in cloud storage, notification to user

[0239] Step 13: User selects and notifies final design

[0240] Specific behavior:

[0241] The user finally checks the virtual try-on images to find one that satisfies him / her, selects the final decorative design, and notifies the server of the selection.

[0242] Input: Final selected virtual try-on image, selection details

[0243] Output: Final selection notification sent to the server

[0244] Step 14: The server saves the selection and ends the process

[0245] Specific behavior:

[0246] The server stores the information of the final selection notified by the user in the cloud storage and the database, and notifies the user that the final decision has been made.

[0247] Input: Final selection notification sent to the server

[0248] Output: Final selection information stored in cloud storage and database, completion notification to user

[0249] (Application example 1)

[0250] 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."

[0251] Conventional virtual try-on systems have the drawback of requiring a great deal of time and effort for users to check decorations, and the decorations often deviate from the actual image. The present invention aims to provide a system that allows users to virtually try on decorations in real time, and to quickly and effectively check and modify the decorations.

[0252] 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.

[0253] In this invention, the server includes: means for a user to take and upload photos of themselves from multiple angles; means for inputting desired ornaments; a generative model for generating a virtual try-on image of the user wearing the ornaments based on the received user photo and desired ornament design; means for the user to provide the generated virtual try-on image to the user; means for the user to issue correction instructions for the virtual try-on image; means for re-executing the generative model based on the correction instructions to update the virtual try-on image; means for confirming and saving the final selected ornament design; means for using a smartphone to check the generated virtual try-on image in real time; and means for the user to input the details of the desired ornament design as a prompt statement and for the system to control the generative model based on the prompt statement. This allows the user to check and correct the virtual try-on image of the ornaments in real time using their smartphone and quickly and efficiently select a final ornament design.

[0254] A "smartphone" is a type of mobile information terminal that has mobile phone functions and can connect to the Internet and use various applications.

[0255] A "virtual try-on image" is an image that simulates the appearance of wearing an accessory, created by a generative model based on the user's image and the desired accessory design.

[0256] A "generative model" is an algorithm or machine learning model that uses computer vision technology to generate virtual try-on images based on a received user photo and decorative design.

[0257] A "prompt sentence" is an input sentence that succinctly expresses the content of the decorative design desired by the user, and is important information for the system to control the generative model based on the instructions.

[0258] "Modification instructions" are instructions that the user inputs to make changes or adjustments to the virtual try-on image, and serve as a trigger for the generative model to be re-executed.

[0259] "Real-time" refers to the fact that the user can instantly check and correct the results of the ornament try-on, and means that the system operates with high response speed.

[0260] This invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. The invention is implemented using a server, a terminal, and a generative AI model.

[0261] System configuration and operation

[0262] 1. User image upload

[0263] Users take photos of themselves from various angles with their smartphones, access the system, select the photos they have taken, and press the upload button to send them to the server.

[0264] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[0265] 2. Input your desired decorative design

[0266] The user inputs the desired decorative design as a prompt, such as a simple text like "I want to try on a red dress." The device sends this prompt to the server, which then stores the received data in its storage.

[0267] 3. Generation of virtual try-on images

[0268] The server inputs the received user image and desired accessory design into a generative AI model. Based on this data, the generative AI model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[0269] 4. Check and edit the virtual try-on image

[0270] The user checks the generated virtual try-on image on their smartphone. If they don't like a particular part of the decorative design, they can input a prompt to make corrections. For example, they could give a specific instruction such as "Make the hem of the dress a little longer." The device then sends this correction instruction to the server.

[0271] The server receives the correction instructions and feeds them back into the generative AI model. The generative model generates a revised virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0272] 5. Final dress selection

[0273] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[0274] Hardware and software used

[0275] The system utilizes the following hardware and software:

[0276] Hardware: Smartphones, servers, GPUs (deep learning-compatible GPUs such as NVIDIA)

[0277] Software: Python, FastAPI (web framework), Torch (deep learning library), PIL (Pillow, image processing library)

[0278] As a concrete example, if a user wants to try on a "red dress," he or she inputs the following prompt into the system:

[0279] Prompt: I want to try on a red dress

[0280] This allows the system of the present invention to allow users to try on an unlimited number of decorations from the convenience of their own home and find the perfect one.

[0281] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0282] Step 1:

[0283] Users take photos of themselves from multiple angles and upload them via their devices. They use their smartphones to take photos of the front, side, back, etc., and press the upload button to send them to the system. The input is the user's image data, and the output is image data stored on the server.

[0284] Step 2:

[0285] The server saves the received user photo data in storage, verifies that the photo was uploaded correctly, and notifies the user. The server saves the image file in storage, verifies that it was saved successfully, and notifies the user of this confirmation information.

[0286] Step 3:

[0287] The user inputs the desired decorative design as a prompt sentence and sends it to the server via the terminal. The input includes text data such as "I would like to try on a red dress." The output is the text data saved on the server.

[0288] Step 4:

[0289] The server inputs the received user image and prompt text into the generative AI model. The user image (input image data) and prompt text (text data) are input into the generative model. Based on this data, the generative AI model generates a virtual try-on image. The output is the generated try-on image data.

[0290] Step 5:

[0291] The generative model is executed to generate a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. This image data is returned to the server and saved in the user's account. The output is a virtual try-on image saved in the user's account.

[0292] Step 6:

[0293] The user checks the generated virtual try-on image through the terminal. The generated image is displayed on the terminal, and the user can see the fully decorated try-on image of themselves. This includes both the input and output of the step.

[0294] Step 7:

[0295] If the user does not like a particular part of the decorative design, they can send a prompt to the server from their terminal with instructions to make corrections. For example, they can input an instruction such as "Make the hem of the dress a little longer." The input is the prompt to make corrections, and the output is correction instruction data stored on the server.

[0296] Step 8:

[0297] The server inputs the correction instructions into the generative AI model again, and the generative model regenerates the corrected virtual try-on image. The inputs are the user image, the existing virtual try-on image, and the correction prompt. The output is the corrected try-on image data.

[0298] Step 9:

[0299] The server provides the generated modified virtual try-on image to the user. The data is sent to the user's device, allowing the user to view the new try-on image. The output is the modified try-on image displayed by the user.

[0300] Step 10:

[0301] The user finally selects the decorative design they like and notifies the server via their terminal. The user presses the decision button to send the design to the server. The input at this time is the user's final decision data, and the output is the final selection information stored on the server.

[0302] 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.

[0303] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[0304] System configuration and operation

[0305] Image upload

[0306] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[0307] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[0308] Decorative design input

[0309] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[0310] Virtual try-on image generation

[0311] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[0312] Emotion recognition by emotion engine

[0313] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[0314] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[0315] Check and edit virtual try-on images

[0316] If the user needs to make any corrections to the generated virtual try-on image, the user inputs correction instructions, which are then sent from the terminal to the server.

[0317] The server receives the correction instructions and re-inputs them into the generative model. The generative model regenerates the virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0318] Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a decoration, the emotion engine will suggest a different color.

[0319] Final Selection

[0320] The user finally selects the decorative design they like on the terminal.

[0321] The terminal transmits the final selected decorative design information to the server.

[0322] The server determines the selected decorative design and stores the associated data in storage.

[0323] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[0324] Specific examples

[0325] For example, if user A wants to choose a kimono for his / her coming-of-age ceremony, he / she can use the system as follows:

[0326] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[0327] 2. The server receives these photos and stores them in storage.

[0328] 3. User A takes a picture of a kimono in the catalog with their smartphone and inputs it into the system.

[0329] 4. The server receives this data and inputs it into the generative model.

[0330] 5. The generative model generates a virtual try-on image of User A wearing the kimono and sends it to the server.

[0331] 6. The server provides the generated virtual try-on image to User A.

[0332] 7. When User A checks the try-on images on the device, he / she sends his / her reaction to the emotion engine via the camera and microphone.

[0333] 8. The emotion engine analyzes User A's facial expressions and voice and reports the current emotion to the server.

[0334] 9. If user A is dissatisfied with a particular part of the try-on image, he or she inputs correction instructions.

[0335] 10. The server re-inputs the correction instructions into the generative model, generates an updated image, and serves it to User A. This process is repeated based on feedback from the emotion engine.

[0336] 11. User A finally chooses the decorative design he likes and confirms it on the system.

[0337] 12. The server saves the selection and completes the process.

[0338] In this way, the system of the present invention allows users to try on an infinite number of outfits from the comfort of their own home and find their perfect outfit through emotion recognition.

[0339] The processing flow will be explained below.

[0340] Specific steps in the system programming process

[0341] Image upload

[0342] Step 1:

[0343] Users take photos of themselves from various angles and access the system using a terminal.

[0344] Step 2:

[0345] The device selects multiple photos taken and sends them to the server by pressing the upload button.

[0346] Step 3:

[0347] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[0348] Decorative design input

[0349] Step 4:

[0350] The user inputs the desired decorative design as image data or text instructions.

[0351] Step 5:

[0352] The terminal transmits the input decorative design data to the server.

[0353] Step 6:

[0354] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[0355] Virtual try-on image generation

[0356] Step 7:

[0357] The server inputs the received user image and desired decorative design into the generative model.

[0358] Step 8:

[0359] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[0360] Step 9:

[0361] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[0362] Step 10:

[0363] The server receives the generated virtual try-on images and stores them in the user's account.

[0364] Emotion recognition by emotion engine

[0365] Step 11:

[0366] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[0367] Step 12:

[0368] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions, and evaluates the virtual try-on images based on this information.

[0369] Check and edit virtual try-on images

[0370] Step 13:

[0371] If the user needs to make any corrections to the generated virtual try-on image, they can input specific corrections, such as "change the hem length."

[0372] Step 14:

[0373] The terminal sends a correction instruction to the server.

[0374] Step 15:

[0375] The server re-inputs the received correction instructions into the generative model.

[0376] Step 16:

[0377] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[0378] Step 17:

[0379] The server receives the updated virtual try-on images and stores them in the user's account.

[0380] Step 18:

[0381] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[0382] Step 19:

[0383] The emotion engine monitors the user's emotions, identifies areas of dissatisfaction, and provides that information to the server.

[0384] Step 20:

[0385] Based on the information from the emotion engine, the server suggests appropriate decorative designs and modifications to the user.

[0386] Final Selection

[0387] Step 21:

[0388] The user finally selects the decorative design they like on the terminal.

[0389] Step 22:

[0390] The terminal transmits the final selected decorative design information to the server.

[0391] Step 23:

[0392] The server determines the selected decorative design and stores the associated data in storage.

[0393] Step 24:

[0394] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[0395] Example 2

[0396] 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."

[0397] Conventional virtual try-on systems have a limited process for generating virtual try-on images based on user-provided images and decorative designs, making it difficult to effectively incorporate user emotions and feedback. As a result, users are often dissatisfied with their final selection and are forced to go through multiple trial and error rounds. Furthermore, the quality of try-on images and real-time adjustments are insufficient, leaving room for improvement in the user experience.

[0398] 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.

[0399] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting requested accessories, and an image generation model for generating virtual try-on images of the user wearing the accessories based on the received user images and the requested accessory design. This allows users to generate high-quality virtual try-on images in real time and reflect feedback using an emotion recognition engine.

[0400] A "user" is someone who uses the system to upload a photo of themselves and try on clothes virtually.

[0401] "Terminal" means an electronic device used by a user to access the system and input photos and decorative designs. Examples include smartphones, tablets, and PCs.

[0402] The "server" is a central processing unit that processes data received from users and operates generative AI models and emotion recognition engines.

[0403] A "generative model" is an algorithm or software that generates virtual try-on images based on a received user image and requested accessory design, often using deep learning techniques.

[0404] A "virtual try-on image" is a virtual image generated based on a user's image, showing an outfit with a decorative design specified by the user.

[0405] An "emotion recognition engine" is an algorithm or software that analyzes a user's facial expressions, voice, etc. to identify the user's emotions.

[0406] "Modification instructions" refer to requests for changes made by the user to the virtual try-on image, and specifically include changes to the color or shape of the decorative design.

[0407] "Decorative design" is data that represents the appearance of the clothes, accessories, etc. that the user wishes to try on in the virtual try-on image.

[0408] "Reaction" refers to the facial and vocal feedback given by the user when they view the virtual try-on image.

[0409] "Suggestions" are suggestions for alternative decorative designs to improve user satisfaction, generated by the emotion recognition engine based on the user's reactions.

[0410] These definitions will provide a clear understanding of each claim element.

[0411] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative design. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. Specifically, the system of the present invention is implemented using a server, a terminal, an image generation model, and an emotion recognition engine.

[0412] System configuration

[0413] 1. A device where users take and upload their own images

[0414] Users use devices such as smartphones, tablets, and computers to take and upload images of themselves from various angles.

[0415] The device is equipped with a camera, and the user uses a dedicated application to select and upload images to the system.

[0416] 2. Server

[0417] The server processes the image data received from the user and stores it in storage.

[0418] The server also receives and manages decorative design data sent by users.

[0419] Furthermore, the server operates an image generation model based on the received data to generate a virtual try-on image.

[0420] The server then runs the image generation model again based on the generated virtual try-on images and user feedback.

[0421] 3. Image Generation Model

[0422] The image generation model takes the user's image and decorative design data as input and generates a virtual try-on image, primarily using deep learning technology.

[0423] Virtual try-on images are generated in a format that can be viewed from a 360-degree angle.

[0424] 4. Emotion Recognition Engine

[0425] The emotion recognition engine captures the user's facial expressions and voice through the device's camera and microphone and analyzes the user's emotions in real time.

[0426] Based on the results of this analysis, the emotion recognition engine suggests appropriate decorative designs.

[0427] Specific examples

[0428] Below is a concrete example of how User A uses the system to choose a kimono for her coming-of-age ceremony:

[0429] 1. User A uses a smartphone to take pictures of himself from three angles (front, side, and back) and uploads them to the system via a dedicated application.

[0430] 2. The server receives the image data and stores it in storage.

[0431] 3. User A takes a picture of a kimono from a coming-of-age ceremony catalogue with his smartphone and inputs it into the system as decorative design data.

[0432] 4. The server receives this decorative design data and inputs it into the image generation model.

[0433] 5. The image generation model generates a virtual try-on image of User A wearing the kimono for his coming-of-age ceremony and sends it to the server.

[0434] 6. The server provides the generated virtual try-on image to User A.

[0435] 7. User A checks the virtual try-on image on their device and sends their reaction to the emotion engine via the camera and microphone.

[0436] 8. The emotion engine analyzes User A's facial expressions and voice to determine which designs User A is satisfied or dissatisfied with.

[0437] 9. If user A is dissatisfied with a particular part, he or she enters correction instructions into the terminal and sends them to the server.

[0438] 10. The server receives the correction instructions and again instructs the image generation model to generate updated virtual try-on images.

[0439] 11. The server provides the updated image to User A, and the process repeats until User A is satisfied.

[0440] In this way, the system of the present invention allows users to try on an infinite number of decorations from the comfort of their own home and find their perfect design through emotion recognition.

[0441] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0442] Step 1:

[0443] Users take multiple photos of themselves from various angles, such as from the front, side, and back, select the images through a dedicated application, and press the upload button.

[0444] Input: Images taken by the user from multiple angles

[0445] Output: User's image data

[0446] Step 2:

[0447] The terminal sends the user's image data to the server, where the image data is transferred over the network.

[0448] Input: User's image data

[0449] Output: Image data sent to the server

[0450] Step 3:

[0451] The server stores the received image data of the user in storage, and also generates a message confirming receipt and notifies the user.

[0452] Input: Image data sent from the device

[0453] Output: Image data saved in storage, receipt confirmation message

[0454] Step 4:

[0455] The user inputs the desired decorative design into the terminal as image data or text instructions. For example, the user may take a photo of a kimono for their coming-of-age ceremony and upload it.

[0456] Input: User-provided image data or text instructions for the decorative design

[0457] Output: decorative design data

[0458] Step 5:

[0459] The terminal transmits the input decorative design data to the server, where data transfer also takes place via the network.

[0460] Input: decorative design data

[0461] Output: Decorative design data sent to the server

[0462] Step 6:

[0463] The server stores the received decorative design data in storage, and also generates a confirmation message and notifies the user.

[0464] Input: Decoration design data sent from the device

[0465] Output: Decorative design data saved in storage, confirmation message

[0466] Step 7:

[0467] The server inputs the received user image data and decorative design data into an image generation model, specifically, a deep learning algorithm to generate virtual try-on images.

[0468] Input: User image data, decorative design data

[0469] Output: Generated virtual try-on image

[0470] Step 8:

[0471] The server links the generated virtual try-on image to the user's account and provides it to the user, who can view it on their device.

[0472] Input: Generated virtual try-on images

[0473] Output: A link to the virtual try-on image provided to the user

[0474] Step 9:

[0475] The user checks the virtual fitting images on the device and sends their reactions to the emotion engine via the camera and microphone. The device captures facial expressions and voice.

[0476] Input: Virtual try-on image for user to view

[0477] Output: Captured user reaction data

[0478] Step 10:

[0479] The server analyzes the captured user reaction data using an emotion recognition engine, specifically facial detection and voice tone analysis, to identify user satisfaction or dissatisfaction.

[0480] Input: User response data

[0481] Output: User sentiment analysis results

[0482] Step 11:

[0483] If the user needs to make any modifications to the virtual try-on image, the user can input and send the modification instructions to the terminal, for example, by inputting specific instructions such as "change the color to blue."

[0484] Input: Correction instructions

[0485] Output: Correction instructions typed into the terminal

[0486] Step 12:

[0487] The terminal sends the correction instruction to the server, and data is transferred via the network.

[0488] Input: Correction instructions

[0489] output: Correction instructions sent to the server

[0490] Step 13:

[0491] The server receives the correction instructions and re-inputs them into the generative AI model, which then generates updated virtual try-on images based on the new instructions.

[0492] Input: Correction instructions

[0493] Output: Updated virtual try-on image

[0494] Step 14:

[0495] The server saves the updated virtual try-on image to the user's account and provides it to the user again via the link, and this process is repeated until the user is satisfied.

[0496] Input: Updated virtual try-on image

[0497] output: Updated virtual try-on image link provided to the user

[0498] Step 15:

[0499] The user finally makes a decision to select the decorative design they like, and presses the confirmation button to confirm the selection through the terminal.

[0500] Input: Final decorative design selection

[0501] Output: Selection information

[0502] Step 16:

[0503] The terminal transmits the final selection information to the server, and data transfer occurs via the network.

[0504] Input: Final selection information

[0505] Output: Final selection information sent to the server

[0506] Step 17:

[0507] The server confirms the selected decoration design, stores the related data in storage, and sends a final decision confirmation to the user, informing them that the decoration selection is complete.

[0508] Input: Final selection information

[0509] Output: Selected design data saved in storage, confirmation notification to user

[0510] (Application example 2)

[0511] 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."

[0512] In conventional virtual try-on systems, users had to rely solely on visual confirmation when checking try-on images. This made it difficult to reflect the user's emotions and intuitive reactions, making it difficult for them to select decorative designs that truly satisfied the user. Furthermore, if the user was dissatisfied with the try-on images, suggestions for revisions could not be made efficiently, resulting in a problem of a decline in the quality of the try-on experience. There is a need to solve these problems.

[0513] 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.

[0514] In this invention, the server includes means for a user to take and upload photos of themselves from multiple angles, means for inputting desired decorations, a generative model for generating a virtual try-on image of the user wearing the decorations based on the received user photo and desired decoration design, means for providing the generated virtual try-on image to the user, means for the user to issue correction instructions for the virtual try-on image, means for re-executing the generative model based on the correction instructions to update the virtual try-on image, means for finalizing and saving the selected decoration design, means for capturing the user's facial expressions and voice, recognizing and analyzing emotions, and means for presenting correction suggestions to the user based on the emotion recognition results.

[0515] This allows the system to reflect the user's emotions in real time, efficiently suggest modifications based on intuitive feedback, and select the decorative design that will satisfy the user most.

[0516] "User" refers to a person who uses this system to upload their own photos and try on clothes virtually.

[0517] "Photos" refer to image data taken by a user from multiple angles.

[0518] "Decorations" refer to design elements such as clothing and accessories that the user wishes to try on.

[0519] "Input" refers to the user inputting the desired decorative design into the system.

[0520] "Generative model" refers to a machine learning model or algorithm that generates virtual try-on images based on the user's received photo and desired decorative design.

[0521] A "virtual try-on image" refers to an image that virtually depicts the state in which a user is wearing the accessory.

[0522] "Modification instructions" refer to instructions for changes or modifications that the user makes to the generated virtual try-on image.

[0523] "Emotion recognition" refers to the process of analyzing a user's facial expressions and voice in real time to identify their emotions.

[0524] "Proposed revision" refers to proposed changes to the decorative design that are suggested to the user based on the emotion recognition results.

[0525] "Server" refers to the central system that receives and stores image and design data from users, generates virtual try-on images using generative models, and performs emotion recognition.

[0526] "Storage" refers to the data storage location where the server stores users' photos, decorative design data, and virtual try-on images.

[0527] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[0528] System configuration and operation

[0529] Image upload

[0530] Users access the system by taking photos of themselves from various angles using smart glasses or a smartphone. The user selects the photos and sends them to the server by pressing the upload button. The server stores the received user photo data in storage (e.g., AWS S3) and verifies that the photos have been uploaded correctly.

[0531] Decorative design input

[0532] The user inputs the desired decorative design as image data or text instructions. The device sends this data to the server, which then stores it in storage (e.g., AWS S3).

[0533] Virtual try-on image generation

[0534] The server inputs the received user image and desired accessory design into a generative model (e.g., StyleGAN). Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[0535] Emotion recognition by emotion engine

[0536] The user reviews the virtual try-on images and sends their reactions to the server via their device. The device is equipped with a camera and microphone, which capture the user's facial expressions and voice in real time. The server then uses an emotion engine (e.g., Microsoft Azure Emotion API) to analyze the user's facial expressions and voice and identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[0537] Check and edit virtual try-on images

[0538] If the generated virtual try-on image requires any modifications, the user inputs the modifications. These modifications are sent from the device to the server. The server receives the modifications and inputs them back into the generative model. The generative model then regenerates the virtual try-on image based on the new modifications and sends it to the server. This process is repeated until the user is satisfied. Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a certain decoration, the emotion engine will suggest a different color.

[0539] Final Selection

[0540] The user finally selects their favorite decoration design on the device. The device sends the final selected decoration design information to the server. The server confirms the selected decoration design and saves the related data in storage. The server then sends the user a confirmation of the final decision, notifying them that decoration selection is complete.

[0541] Specific examples

[0542] For example, if user B wants to choose a dress at a brand shop, he can use the system as follows.

[0543] 1. User B uses smart glasses in a store to take photos of himself (front, side, and back).

[0544] 2. The server receives and stores the photo.

[0545] 3. User B takes a picture of the dress they want with their smartphone and enters it into the app.

[0546] 4. The server receives this data and inputs it into a generative model (StyleGAN).

[0547] 5. The generative AI model (StyleGAN) generates a virtual try-on image based on "User B's photo + dress image" and sends it to the server.

[0548] 6. The server provides the generated try-on image to User B.

[0549] 7. User B checks the fitting image. The camera in the smart glasses captures their facial expressions and sends them to the emotion engine.

[0550] 8. The emotion engine (Microsoft Azure Emotion API) analyzes facial expressions and initiates suggestions for corrections to areas of dissatisfaction (e.g., "I don't like the color of the red dress").

[0551] 9. User B inputs color correction instructions. The server again uses the generative model to generate an updated image and provides it to User B.

[0552] 10. User B finally chooses the dress they like and confirms it in the app.

[0553] 11. The server saves your selection and completes the process.

[0554] Example prompts for generative AI models

[0555] "Combine the front, side, and back images of the user to create a virtual image of them trying on a specified decorative design (a red dress)."

[0556] In this way, the system of the present invention provides a support system that allows users to check and select their ideal outfit without trying it on in the store.

[0557] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0558] Step 1:

[0559] Image upload

[0560] Users can use smart glasses or smartphones to take photos of themselves from the front, side, or back, then access the application, select the photo they have taken, and press the upload button to send the photo to the server.

[0561] Input: Image data taken from multiple angles by the user

[0562] The server saves the received image data in storage (e.g. AWS S3) and confirms that it has been uploaded successfully.

[0563] Output: Reference information for image data stored in storage

[0564] Step 2:

[0565] Decorative design input

[0566] The user inputs the desired decorative design into the application as image data or text instructions, which are then sent to the server.

[0567] Input: Image data or text instructions for the decorative design entered by the user

[0568] The server stores the received design data in storage (e.g. AWS S3).

[0569] Output: Reference information for decorative designs stored in storage

[0570] Step 3:

[0571] Virtual try-on image generation

[0572] The server inputs the user's photo data and desired decorative design data into a generative model (e.g., StyleGAN).

[0573] Input: User photo data and decorative design data

[0574] Based on this data, the generative model generates a virtual try-on image of the user wearing the accessories, which can be viewed from a 360-degree angle.

[0575] The server stores the generated virtual try-on image in the user's account and provides it to the user.

[0576] Output: Virtual try-on image data

[0577] Step 4:

[0578] Emotion recognition by emotion engine

[0579] The user checks the provided virtual try-on images, and the device's built-in camera and microphone capture the user's facial expressions and voice in real time as they check.

[0580] Input: User's facial expression data and voice data

[0581] The server analyzes the received facial expression and voice data through an emotion engine (e.g., Microsoft Azure Emotion API) to identify the user's emotions.

[0582] Output: Emotion recognition result data

[0583] Step 5:

[0584] Check and edit virtual try-on images

[0585] If the user needs to make any corrections to the virtual try-on image, the user inputs the correction instructions into the application and sends them to the server.

[0586] Input: User-entered correction instructions

[0587] The server receives and analyzes the correction instructions, and then inputs them back into the generative model to generate a corrected virtual try-on image. The generative model then regenerates the virtual try-on image based on the new instructions.

[0588] The server provides the user with new virtual try-on images, and this process is repeated until the user is satisfied.

[0589] Output: Data of the virtual try-on image after correction

[0590] Step 6:

[0591] Final Selection

[0592] The user finally selects the decorative design they like on the terminal.

[0593] Input: Information on the decorative design that the user finally selected

[0594] The terminal transmits the selected decorative design information to the server.

[0595] The server confirms the selected decorative design, stores the associated data in storage, and sends the user a confirmation of the final decision.

[0596] Output: Final information of the saved decorative design

[0597] This specific processing step allows users to virtually try on clothes and efficiently select the most suitable decorative design through emotion recognition.

[0598] 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.

[0599] 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.

[0600] 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.

[0601] [Second embodiment]

[0602] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0603] 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.

[0604] 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).

[0605] 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.

[0606] 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.

[0607] 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).

[0608] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0609] 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.

[0610] 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.

[0611] 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.

[0612] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0613] 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."

[0614] The present invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative model.

[0615] System configuration and operation

[0616] 1. User image upload

[0617] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[0618] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[0619] 2. Input your desired decorative design

[0620] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[0621] 3. Generation of virtual try-on images

[0622] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[0623] 4. Check and edit the virtual try-on image

[0624] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[0625] The server receives the correction instructions and re-inputs them into the generative model. The generative model generates a corrected virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0626] 5. Final dress selection

[0627] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[0628] Specific examples

[0629] If user A wants to choose a wedding dress, he can use the system as follows:

[0630] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[0631] 2. The server receives these photos and stores them in storage.

[0632] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[0633] 4. The server receives this data and inputs it into the generative model.

[0634] 5. The generative model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[0635] 6. The server provides the generated virtual try-on image to User A.

[0636] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[0637] 8. The server re-inputs the correction instructions into the generative model, generates an updated image, and provides it to User A.

[0638] 9. User A finally chooses the dress she likes and confirms it on the system.

[0639] 10. The server saves the selection and completes the process.

[0640] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[0641] The processing flow will be explained below.

[0642] Specific steps in the system programming process

[0643] Image upload

[0644] Step 1:

[0645] Users take photos of themselves from multiple angles and use a terminal to access the system.

[0646] Step 2:

[0647] The terminal allows the user to select a photo they have taken and send it to the server by pressing the upload button.

[0648] Step 3:

[0649] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[0650] Decorative design input

[0651] Step 4:

[0652] The user inputs the desired decorative design in the form of image data, text instructions, or both.

[0653] Step 5:

[0654] The terminal transmits the input decorative design data to the server.

[0655] Step 6:

[0656] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[0657] Virtual try-on image generation

[0658] Step 7:

[0659] The server inputs the received user image and desired decorative design into the generative model.

[0660] Step 8:

[0661] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[0662] Step 9:

[0663] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[0664] Step 10:

[0665] The server receives the generated virtual try-on images and stores them in the user's account.

[0666] Check and edit virtual try-on images

[0667] Step 11:

[0668] The user checks the generated virtual try-on image on the terminal.

[0669] Step 12:

[0670] If a specific part of the image needs to be corrected, the user inputs a correction instruction.

[0671] Step 13:

[0672] The terminal sends a correction instruction to the server.

[0673] Step 14:

[0674] The server re-inputs the received correction instructions into the generative model.

[0675] Step 15:

[0676] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[0677] Step 16:

[0678] The server receives the updated virtual try-on images and stores them in the user's account.

[0679] Step 17:

[0680] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[0681] Final Selection

[0682] Step 18:

[0683] The user finally selects the decorative design they like on the terminal.

[0684] Step 19:

[0685] The terminal transmits the final selected decorative design information to the server.

[0686] Step 20:

[0687] The server determines the selected decorative design and stores the associated data in storage.

[0688] Step 21:

[0689] The server sends the user a confirmation of the final decision, informing them that their dress selection is complete.

[0690] Example 1

[0691] 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."

[0692] Conventional try-on systems require users to physically visit a store, which requires time and effort. Even online try-on systems have the drawback of making it difficult for users to make fine adjustments when trying on accessories, making it difficult to fully replicate the actual wearing experience. Furthermore, there is a lack of a way for users to check the accessories in detail from multiple angles.

[0693] 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.

[0694] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting a desired ornament design, and a generative model for generating a virtual try-on image of the user wearing the ornament based on the received user image and the desired ornament design. This allows users to try on an unlimited number of ornaments from the comfort of their own home and find their perfect outfit. The generated virtual try-on image can also be viewed from a 360-degree angle, making it easy for users to check the details of the desired ornament. Furthermore, by re-executing the generative model based on correction instructions and updating the virtual try-on image, users can adjust the ornament design to their satisfaction.

[0695] "User images" are photographic data taken by a user of himself or herself, taken from multiple angles.

[0696] A "desired accessory design" is a design of accessory or clothing that the user wishes to try on, and may take the form of image data or text instructions.

[0697] A "generative model" is an artificial intelligence model that generates virtual try-on images using a user's image and desired decorative design as input, and has image generation and editing functions.

[0698] A "virtual try-on image" is an image generated by a generative model that virtually recreates the appearance of a user wearing an accessory.

[0699] "Modification instructions" are requests for changes or adjustments made by the user to the virtual try-on image, and include specific changes to parts or styles.

[0700] "360-degree angle" means that the virtual try-on image can be viewed from all directions, allowing the user to check the decorations from multiple angles.

[0701] "Server" means a centralized computer system that receives, stores, and processes images and data from users.

[0702] This invention relates to a system that allows users to virtually try on accessories by taking pictures of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative AI model.

[0703] System configuration and operation

[0704] 1. User image upload

[0705] Users take pictures of themselves from the front, side, and back, and then access the system using a terminal. Users select the images they have taken and press the upload button to send them to the server.

[0706] The server stores the received user image data in cloud storage, confirms that the image has been uploaded correctly, and notifies the user.

[0707] Hardware used: smartphone, PC, cloud server

[0708] Software used: web browser, photo-taking application, cloud storage system (e.g., Amazon S3)

[0709] 2. Input your desired decorative design

[0710] The user inputs the desired decorative design as image data or text instructions. The terminal transmits this data to the server,

[0711] The server stores the received design data in cloud storage.

[0712] Hardware used: smartphone, PC, cloud server

[0713] Software used: Web forms, text input interface

[0714] 3. Generation of virtual try-on images

[0715] The server inputs the received user image and desired decorative design into a generative AI model.

[0716] The generative AI model uses this data to generate virtual try-on images of the user wearing the accessories, which can be viewed from a 360-degree angle.

[0717] The server provides this image to the user.

[0718] Hardware used: Cloud server

[0719] Software used: Generative AI models (e.g., DALL-E, Stable Diffusion), databases

[0720] 4. Check and edit the virtual try-on image

[0721] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[0722] The server receives the correction instructions and feeds them back into the generative AI model, which then generates a modified virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0723] Hardware used: smartphone, PC, cloud server

[0724] Software used: Generative AI models, web forms

[0725] 5. Final selection of decorative design

[0726] The user finally selects the decorative design they like and notifies the server via their terminal.

[0727] The server stores the information and confirms the final decision.

[0728] Hardware used: smartphone, PC, cloud server

[0729] Software used: Web forms, databases

[0730] Specific examples

[0731] If user A wants to choose a wedding dress, he can use the system as follows:

[0732] 1. User A takes pictures of himself from the front, side, and back and uploads them to the system.

[0733] Example prompt: "Please upload a photo of your front, side, and back."

[0734] 2. The server receives these images and stores them in cloud storage.

[0735] Example prompt: "Your image was successfully uploaded."

[0736] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[0737] Example prompt: "Please upload a photo of the dress you would like."

[0738] 4. The server receives this data and inputs it into a generative AI model.

[0739] 5. The generative AI model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[0740] 6. The server provides the generated virtual try-on image to User A.

[0741] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[0742] Example prompt: "Please make the hem a little shorter."

[0743] 8. The server re-inputs the correction instructions into the generative AI model, generates an updated image, and provides it to User A.

[0744] 9. User A finally chooses the dress she likes and confirms it on the system.

[0745] 10. The server saves the selection and completes the process.

[0746] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[0747] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0748] Step 1: Take and upload a user image

[0749] Specific behavior:

[0750] Users take photos of themselves from the front, side, and back using a smartphone or camera.

[0751] Save the captured image file to a folder on your device.

[0752] Input: Captured image files (front, side, back)

[0753] Output: Image file saved on the device

[0754] Step 2: Access the system and upload your images

[0755] Specific behavior:

[0756] The user opens a web browser on the terminal, accesses the system's web application, and logs in.

[0757] Access the image upload form, select the image file you took, and press the upload button.

[0758] Input: Image files stored on the device, user login information

[0759] Output: Image file sent to the server

[0760] Step 3: The server receives and stores the image data.

[0761] Specific behavior:

[0762] The server saves the image data sent by the user in cloud storage and records the file path of the saved data in a database.

[0763] Verify that the image was uploaded correctly, generate a success message, and notify the user.

[0764] Input: Image data sent by the user

[0765] Output: Image data saved in cloud storage, success message

[0766] Step 4: User inputs desired decorative design

[0767] Specific behavior:

[0768] The user inputs image data and text instructions with the desired decorative design into the terminal, for example, by uploading an image of the dress taken with a smartphone.

[0769] Access the input form, select the design data, and press the submit button.

[0770] Input: Image data of decorative design taken with a smartphone or text instructions

[0771] Output: Decorative design data sent to the server

[0772] Step 5: The server receives and stores the design data

[0773] Specific behavior:

[0774] The server saves the decorative design data sent by the user in cloud storage, and records the file path and text content of the saved data in a database.

[0775] Input: Decoration design data submitted by the user

[0776] Output: Design data stored in cloud storage, file paths and text contents recorded in the database

[0777] Step 6: Start generating virtual try-on images

[0778] Specific behavior:

[0779] The server performs preprocessing to input the image data captured by the user and the desired decorative design data into the generative AI model.

[0780] The data is converted into the appropriate format and fed into a generative AI model.

[0781] Input: User image data stored in cloud storage, decorative design data stored in cloud storage

[0782] Output: The data that is fed into the generative AI model

[0783] Step 7: The generative AI model generates virtual try-on images

[0784] Specific behavior:

[0785] The generative AI model generates virtual try-on images based on the input user image data and decorative design data.

[0786] The generated try-on images are processed so that they can be displayed from multiple angles.

[0787] Input: User image data and decorative design data input into the generative AI model

[0788] Output: Generated virtual try-on image

[0789] Step 8: The server stores and serves the generated images.

[0790] Specific behavior:

[0791] The server saves the virtual try-on images sent from the generative AI model to cloud storage and records the file paths of the saved images in a database.

[0792] The user is notified of the URL of the generated virtual try-on image.

[0793] Input: Virtual try-on images sent from the generative AI model

[0794] Output: Virtual try-on images saved in cloud storage, notification to user

[0795] Step 9: The user checks the fitting image and inputs correction instructions.

[0796] Specific behavior:

[0797] The user can check the virtual fitting images provided on the device and input specific correction instructions as needed, such as "Please make the hem a little shorter."

[0798] Input: Provided virtual try-on image, modification instructions

[0799] Output: Correction instructions sent to the server

[0800] Step 10: The server re-inputs the correction instructions into the generative AI model.

[0801] Specific behavior:

[0802] The server performs preprocessing to re-input the received correction instructions into the generative AI model.

[0803] The correction instructions are input into the generative AI model, requesting the generation of an updated virtual try-on image.

[0804] Input: Correction instructions sent to the server

[0805] Output: Corrective instruction data that is input to the generative AI model

[0806] Step 11: The generative AI model generates the corrected image

[0807] Specific behavior:

[0808] The generative AI model generates new virtual try-on images based on the correction instructions and sends them to the server.

[0809] Input: Correction instructions input to the generative AI model

[0810] Output: Corrected virtual try-on image

[0811] Step 12: The server stores and serves the corrected image.

[0812] Specific behavior:

[0813] The server saves the newly generated virtual try-on image in cloud storage, records the file path of the saved image in the database, and notifies the user of the URL of the modified virtual try-on image.

[0814] Input: Modified virtual try-on image sent from the generative AI model

[0815] Output: Edited try-on image saved in cloud storage, notification to user

[0816] Step 13: User selects and notifies final design

[0817] Specific behavior:

[0818] The user finally checks the virtual try-on images to find one that satisfies him / her, selects the final decorative design, and notifies the server of the selection.

[0819] Input: Final selected virtual try-on image, selection details

[0820] Output: Final selection notification sent to the server

[0821] Step 14: The server saves the selection and ends the process

[0822] Specific behavior:

[0823] The server stores the information of the final selection notified by the user in the cloud storage and the database, and notifies the user that the final decision has been made.

[0824] Input: Final selection notification sent to the server

[0825] Output: Final selection information stored in cloud storage and database, completion notification to user

[0826] (Application example 1)

[0827] 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."

[0828] Conventional virtual try-on systems have the drawback of requiring a great deal of time and effort for users to check decorations, and the decorations often deviate from the actual image. The present invention aims to provide a system that allows users to virtually try on decorations in real time, and to quickly and effectively check and modify the decorations.

[0829] 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.

[0830] In this invention, the server includes: means for a user to take and upload photos of themselves from multiple angles; means for inputting desired ornaments; a generative model for generating a virtual try-on image of the user wearing the ornaments based on the received user photo and desired ornament design; means for the user to provide the generated virtual try-on image to the user; means for the user to issue correction instructions for the virtual try-on image; means for re-executing the generative model based on the correction instructions to update the virtual try-on image; means for confirming and saving the final selected ornament design; means for using a smartphone to check the generated virtual try-on image in real time; and means for the user to input the details of the desired ornament design as a prompt statement and for the system to control the generative model based on the prompt statement. This allows the user to check and correct the virtual try-on image of the ornaments in real time using their smartphone and quickly and efficiently select a final ornament design.

[0831] A "smartphone" is a type of mobile information terminal that has mobile phone functions and can connect to the Internet and use various applications.

[0832] A "virtual try-on image" is an image that simulates the appearance of wearing an accessory, created by a generative model based on the user's image and the desired accessory design.

[0833] A "generative model" is an algorithm or machine learning model that uses computer vision technology to generate virtual try-on images based on a received user photo and decorative design.

[0834] A "prompt sentence" is an input sentence that succinctly expresses the content of the decorative design desired by the user, and is important information for the system to control the generative model based on the instructions.

[0835] "Modification instructions" are instructions that the user inputs to make changes or adjustments to the virtual try-on image, and serve as a trigger for the generative model to be re-executed.

[0836] "Real-time" refers to the fact that the user can instantly check and correct the results of the ornament try-on, and means that the system operates with high response speed.

[0837] This invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. The invention is implemented using a server, a terminal, and a generative AI model.

[0838] System configuration and operation

[0839] 1. User image upload

[0840] Users take photos of themselves from various angles with their smartphones, access the system, select the photos they have taken, and press the upload button to send them to the server.

[0841] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[0842] 2. Input your desired decorative design

[0843] The user inputs the desired decorative design as a prompt, such as a simple text like "I want to try on a red dress." The device sends this prompt to the server, which then stores the received data in its storage.

[0844] 3. Generation of virtual try-on images

[0845] The server inputs the received user image and desired accessory design into a generative AI model. Based on this data, the generative AI model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[0846] 4. Check and edit the virtual try-on image

[0847] The user checks the generated virtual try-on image on their smartphone. If they don't like a particular part of the decorative design, they can input a prompt to make corrections. For example, they could give a specific instruction such as "Make the hem of the dress a little longer." The device then sends this correction instruction to the server.

[0848] The server receives the correction instructions and feeds them back into the generative AI model. The generative model generates a revised virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0849] 5. Final dress selection

[0850] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[0851] Hardware and software used

[0852] The system utilizes the following hardware and software:

[0853] Hardware: Smartphones, servers, GPUs (deep learning-compatible GPUs such as NVIDIA)

[0854] Software: Python, FastAPI (web framework), Torch (deep learning library), PIL (Pillow, image processing library)

[0855] As a concrete example, if a user wants to try on a "red dress," he or she inputs the following prompt into the system:

[0856] Prompt: I want to try on a red dress

[0857] This allows the system of the present invention to allow users to try on an unlimited number of decorations from the convenience of their own home and find the perfect one.

[0858] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0859] Step 1:

[0860] Users take photos of themselves from multiple angles and upload them via their devices. They use their smartphones to take photos of the front, side, back, etc., and press the upload button to send them to the system. The input is the user's image data, and the output is image data stored on the server.

[0861] Step 2:

[0862] The server saves the received user photo data in storage, verifies that the photo was uploaded correctly, and notifies the user. The server saves the image file in storage, verifies that it was saved successfully, and notifies the user of this confirmation information.

[0863] Step 3:

[0864] The user inputs the desired decorative design as a prompt sentence and sends it to the server via the terminal. The input includes text data such as "I would like to try on a red dress." The output is the text data saved on the server.

[0865] Step 4:

[0866] The server inputs the received user image and prompt text into the generative AI model. The user image (input image data) and prompt text (text data) are input into the generative model. Based on this data, the generative AI model generates a virtual try-on image. The output is the generated try-on image data.

[0867] Step 5:

[0868] The generative model is executed to generate a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. This image data is returned to the server and saved in the user's account. The output is a virtual try-on image saved in the user's account.

[0869] Step 6:

[0870] The user checks the generated virtual try-on image through the terminal. The generated image is displayed on the terminal, and the user can see the fully decorated try-on image of themselves. This includes both the input and output of the step.

[0871] Step 7:

[0872] If the user does not like a particular part of the decorative design, they can send a prompt to the server from their terminal with instructions to make corrections. For example, they can input an instruction such as "Make the hem of the dress a little longer." The input is the prompt to make corrections, and the output is correction instruction data stored on the server.

[0873] Step 8:

[0874] The server inputs the correction instructions into the generative AI model again, and the generative model regenerates the corrected virtual try-on image. The inputs are the user image, the existing virtual try-on image, and the correction prompt. The output is the corrected try-on image data.

[0875] Step 9:

[0876] The server provides the generated modified virtual try-on image to the user. The data is sent to the user's device, allowing the user to view the new try-on image. The output is the modified try-on image displayed by the user.

[0877] Step 10:

[0878] The user finally selects the decorative design they like and notifies the server via their terminal. The user presses the decision button to send the design to the server. The input at this time is the user's final decision data, and the output is the final selection information stored on the server.

[0879] 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.

[0880] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[0881] System configuration and operation

[0882] Image upload

[0883] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[0884] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[0885] Decorative design input

[0886] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[0887] Virtual try-on image generation

[0888] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[0889] Emotion recognition by emotion engine

[0890] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[0891] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[0892] Check and edit virtual try-on images

[0893] If the user needs to make any corrections to the generated virtual try-on image, the user inputs correction instructions, which are then sent from the terminal to the server.

[0894] The server receives the correction instructions and re-inputs them into the generative model. The generative model regenerates the virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[0895] Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a decoration, the emotion engine will suggest a different color.

[0896] Final Selection

[0897] The user finally selects the decorative design they like on the terminal.

[0898] The terminal transmits the final selected decorative design information to the server.

[0899] The server determines the selected decorative design and stores the associated data in storage.

[0900] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[0901] Specific examples

[0902] For example, if user A wants to choose a kimono for his / her coming-of-age ceremony, he / she can use the system as follows:

[0903] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[0904] 2. The server receives these photos and stores them in storage.

[0905] 3. User A takes a picture of a kimono in the catalog with their smartphone and inputs it into the system.

[0906] 4. The server receives this data and inputs it into the generative model.

[0907] 5. The generative model generates a virtual try-on image of User A wearing the kimono and sends it to the server.

[0908] 6. The server provides the generated virtual try-on image to User A.

[0909] 7. When User A checks the try-on images on the device, he / she sends his / her reaction to the emotion engine via the camera and microphone.

[0910] 8. The emotion engine analyzes User A's facial expressions and voice and reports the current emotion to the server.

[0911] 9. If user A is dissatisfied with a particular part of the try-on image, he or she inputs correction instructions.

[0912] 10. The server re-inputs the correction instructions into the generative model, generates an updated image, and serves it to User A. This process is repeated based on feedback from the emotion engine.

[0913] 11. User A finally chooses the decorative design he likes and confirms it on the system.

[0914] 12. The server saves the selection and completes the process.

[0915] In this way, the system of the present invention allows users to try on an infinite number of outfits from the comfort of their own home and find their perfect outfit through emotion recognition.

[0916] The processing flow will be explained below.

[0917] Specific steps in the system programming process

[0918] Image upload

[0919] Step 1:

[0920] Users take photos of themselves from various angles and access the system using a terminal.

[0921] Step 2:

[0922] The device selects multiple photos taken and sends them to the server by pressing the upload button.

[0923] Step 3:

[0924] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[0925] Decorative design input

[0926] Step 4:

[0927] The user inputs the desired decorative design as image data or text instructions.

[0928] Step 5:

[0929] The terminal transmits the input decorative design data to the server.

[0930] Step 6:

[0931] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[0932] Virtual try-on image generation

[0933] Step 7:

[0934] The server inputs the received user image and desired decorative design into the generative model.

[0935] Step 8:

[0936] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[0937] Step 9:

[0938] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[0939] Step 10:

[0940] The server receives the generated virtual try-on images and stores them in the user's account.

[0941] Emotion recognition by emotion engine

[0942] Step 11:

[0943] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[0944] Step 12:

[0945] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions, and evaluates the virtual try-on images based on this information.

[0946] Check and edit virtual try-on images

[0947] Step 13:

[0948] If the user needs to make any corrections to the generated virtual try-on image, they can input specific corrections, such as "change the hem length."

[0949] Step 14:

[0950] The terminal sends a correction instruction to the server.

[0951] Step 15:

[0952] The server re-inputs the received correction instructions into the generative model.

[0953] Step 16:

[0954] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[0955] Step 17:

[0956] The server receives the updated virtual try-on images and stores them in the user's account.

[0957] Step 18:

[0958] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[0959] Step 19:

[0960] The emotion engine monitors the user's emotions, identifies areas of dissatisfaction, and provides that information to the server.

[0961] Step 20:

[0962] Based on the information from the emotion engine, the server suggests appropriate decorative designs and modifications to the user.

[0963] Final Selection

[0964] Step 21:

[0965] The user finally selects the decorative design they like on the terminal.

[0966] Step 22:

[0967] The terminal transmits the final selected decorative design information to the server.

[0968] Step 23:

[0969] The server determines the selected decorative design and stores the associated data in storage.

[0970] Step 24:

[0971] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[0972] Example 2

[0973] 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."

[0974] Conventional virtual try-on systems have a limited process for generating virtual try-on images based on user-provided images and decorative designs, making it difficult to effectively incorporate user emotions and feedback. As a result, users are often dissatisfied with their final selection and are forced to go through multiple trial and error rounds. Furthermore, the quality of try-on images and real-time adjustments are insufficient, leaving room for improvement in the user experience.

[0975] 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.

[0976] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting requested accessories, and an image generation model for generating virtual try-on images of the user wearing the accessories based on the received user images and the requested accessory design. This allows users to generate high-quality virtual try-on images in real time and reflect feedback using an emotion recognition engine.

[0977] A "user" is someone who uses the system to upload a photo of themselves and try on clothes virtually.

[0978] "Terminal" means an electronic device used by a user to access the system and input photos and decorative designs. Examples include smartphones, tablets, and PCs.

[0979] The "server" is a central processing unit that processes data received from users and operates generative AI models and emotion recognition engines.

[0980] A "generative model" is an algorithm or software that generates virtual try-on images based on a received user image and requested accessory design, often using deep learning techniques.

[0981] A "virtual try-on image" is a virtual image generated based on a user's image, showing an outfit with a decorative design specified by the user.

[0982] An "emotion recognition engine" is an algorithm or software that analyzes a user's facial expressions, voice, etc. to identify the user's emotions.

[0983] "Modification instructions" refer to requests for changes made by the user to the virtual try-on image, and specifically include changes to the color or shape of the decorative design.

[0984] "Decorative design" is data that represents the appearance of the clothes, accessories, etc. that the user wishes to try on in the virtual try-on image.

[0985] "Reaction" refers to the facial and vocal feedback given by the user when they view the virtual try-on image.

[0986] "Suggestions" are suggestions for alternative decorative designs to improve user satisfaction, generated by the emotion recognition engine based on the user's reactions.

[0987] These definitions will provide a clear understanding of each claim element.

[0988] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative design. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. Specifically, the system of the present invention is implemented using a server, a terminal, an image generation model, and an emotion recognition engine.

[0989] System configuration

[0990] 1. A device where users take and upload their own images

[0991] Users use devices such as smartphones, tablets, and computers to take and upload images of themselves from various angles.

[0992] The device is equipped with a camera, and the user uses a dedicated application to select and upload images to the system.

[0993] 2. Server

[0994] The server processes the image data received from the user and stores it in storage.

[0995] The server also receives and manages decorative design data sent by users.

[0996] Furthermore, the server operates an image generation model based on the received data to generate a virtual try-on image.

[0997] The server then runs the image generation model again based on the generated virtual try-on images and user feedback.

[0998] 3. Image Generation Model

[0999] The image generation model takes the user's image and decorative design data as input and generates a virtual try-on image, primarily using deep learning technology.

[1000] Virtual try-on images are generated in a format that can be viewed from a 360-degree angle.

[1001] 4. Emotion Recognition Engine

[1002] The emotion recognition engine captures the user's facial expressions and voice through the device's camera and microphone and analyzes the user's emotions in real time.

[1003] Based on the results of this analysis, the emotion recognition engine suggests appropriate decorative designs.

[1004] Specific examples

[1005] Below is a concrete example of how User A uses the system to choose a kimono for her coming-of-age ceremony:

[1006] 1. User A uses a smartphone to take pictures of himself from three angles (front, side, and back) and uploads them to the system via a dedicated application.

[1007] 2. The server receives the image data and stores it in storage.

[1008] 3. User A takes a picture of a kimono from a coming-of-age ceremony catalogue with his smartphone and inputs it into the system as decorative design data.

[1009] 4. The server receives this decorative design data and inputs it into the image generation model.

[1010] 5. The image generation model generates a virtual try-on image of User A wearing the kimono for his coming-of-age ceremony and sends it to the server.

[1011] 6. The server provides the generated virtual try-on image to User A.

[1012] 7. User A checks the virtual try-on image on their device and sends their reaction to the emotion engine via the camera and microphone.

[1013] 8. The emotion engine analyzes User A's facial expressions and voice to determine which designs User A is satisfied or dissatisfied with.

[1014] 9. If user A is dissatisfied with a particular part, he or she enters correction instructions into the terminal and sends them to the server.

[1015] 10. The server receives the correction instructions and again instructs the image generation model to generate updated virtual try-on images.

[1016] 11. The server provides the updated image to User A, and the process repeats until User A is satisfied.

[1017] In this way, the system of the present invention allows users to try on an infinite number of decorations from the comfort of their own home and find their perfect design through emotion recognition.

[1018] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1019] Step 1:

[1020] Users take multiple photos of themselves from various angles, such as from the front, side, and back, select the images through a dedicated application, and press the upload button.

[1021] Input: Images taken by the user from multiple angles

[1022] Output: User's image data

[1023] Step 2:

[1024] The terminal sends the user's image data to the server, where the image data is transferred over the network.

[1025] Input: User's image data

[1026] Output: Image data sent to the server

[1027] Step 3:

[1028] The server stores the received image data of the user in storage, and also generates a message confirming receipt and notifies the user.

[1029] Input: Image data sent from the device

[1030] Output: Image data saved in storage, receipt confirmation message

[1031] Step 4:

[1032] The user inputs the desired decorative design into the terminal as image data or text instructions. For example, the user may take a photo of a kimono for their coming-of-age ceremony and upload it.

[1033] Input: User-provided image data or text instructions for the decorative design

[1034] Output: decorative design data

[1035] Step 5:

[1036] The terminal transmits the input decorative design data to the server, where data transfer also takes place via the network.

[1037] Input: decorative design data

[1038] Output: Decorative design data sent to the server

[1039] Step 6:

[1040] The server stores the received decorative design data in storage, and also generates a confirmation message and notifies the user.

[1041] Input: Decoration design data sent from the device

[1042] Output: Decorative design data saved in storage, confirmation message

[1043] Step 7:

[1044] The server inputs the received user image data and decorative design data into an image generation model, specifically, a deep learning algorithm to generate virtual try-on images.

[1045] Input: User image data, decorative design data

[1046] Output: Generated virtual try-on image

[1047] Step 8:

[1048] The server links the generated virtual try-on image to the user's account and provides it to the user, who can view it on their device.

[1049] Input: Generated virtual try-on images

[1050] Output: A link to the virtual try-on image provided to the user

[1051] Step 9:

[1052] The user checks the virtual fitting images on the device and sends their reactions to the emotion engine via the camera and microphone. The device captures facial expressions and voice.

[1053] Input: Virtual try-on image for user to view

[1054] Output: Captured user reaction data

[1055] Step 10:

[1056] The server analyzes the captured user reaction data using an emotion recognition engine, specifically facial detection and voice tone analysis, to identify user satisfaction or dissatisfaction.

[1057] Input: User response data

[1058] Output: User sentiment analysis results

[1059] Step 11:

[1060] If the user needs to make any modifications to the virtual try-on image, the user can input and send the modification instructions to the terminal, for example, by inputting specific instructions such as "change the color to blue."

[1061] Input: Correction instructions

[1062] Output: Correction instructions typed into the terminal

[1063] Step 12:

[1064] The terminal sends the correction instruction to the server, and data is transferred via the network.

[1065] Input: Correction instructions

[1066] output: Correction instructions sent to the server

[1067] Step 13:

[1068] The server receives the correction instructions and re-inputs them into the generative AI model, which then generates updated virtual try-on images based on the new instructions.

[1069] Input: Correction instructions

[1070] Output: Updated virtual try-on image

[1071] Step 14:

[1072] The server saves the updated virtual try-on image to the user's account and provides it to the user again via the link, and this process is repeated until the user is satisfied.

[1073] Input: Updated virtual try-on image

[1074] output: Updated virtual try-on image link provided to the user

[1075] Step 15:

[1076] The user finally makes a decision to select the decorative design they like, and presses the confirmation button to confirm the selection through the terminal.

[1077] Input: Final decorative design selection

[1078] Output: Selection information

[1079] Step 16:

[1080] The terminal transmits the final selection information to the server, and data transfer occurs via the network.

[1081] Input: Final selection information

[1082] Output: Final selection information sent to the server

[1083] Step 17:

[1084] The server confirms the selected decoration design, stores the related data in storage, and sends a final decision confirmation to the user, informing them that the decoration selection is complete.

[1085] Input: Final selection information

[1086] Output: Selected design data saved in storage, confirmation notification to user

[1087] (Application example 2)

[1088] 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."

[1089] In conventional virtual try-on systems, users had to rely solely on visual confirmation when checking try-on images. This made it difficult to reflect the user's emotions and intuitive reactions, making it difficult for them to select decorative designs that truly satisfied the user. Furthermore, if the user was dissatisfied with the try-on images, suggestions for revisions could not be made efficiently, resulting in a problem of a decline in the quality of the try-on experience. There is a need to solve these problems.

[1090] 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.

[1091] In this invention, the server includes means for a user to take and upload photos of themselves from multiple angles, means for inputting desired decorations, a generative model for generating a virtual try-on image of the user wearing the decorations based on the received user photo and desired decoration design, means for providing the generated virtual try-on image to the user, means for the user to issue correction instructions for the virtual try-on image, means for re-executing the generative model based on the correction instructions to update the virtual try-on image, means for finalizing and saving the selected decoration design, means for capturing the user's facial expressions and voice, recognizing and analyzing emotions, and means for presenting correction suggestions to the user based on the emotion recognition results.

[1092] This allows the system to reflect the user's emotions in real time, efficiently suggest modifications based on intuitive feedback, and select the decorative design that will satisfy the user most.

[1093] "User" refers to a person who uses this system to upload their own photos and try on clothes virtually.

[1094] "Photos" refer to image data taken by a user from multiple angles.

[1095] "Decorations" refer to design elements such as clothing and accessories that the user wishes to try on.

[1096] "Input" refers to the user inputting the desired decorative design into the system.

[1097] "Generative model" refers to a machine learning model or algorithm that generates virtual try-on images based on the user's received photo and desired decorative design.

[1098] A "virtual try-on image" refers to an image that virtually depicts the state in which a user is wearing the accessory.

[1099] "Modification instructions" refer to instructions for changes or modifications that the user makes to the generated virtual try-on image.

[1100] "Emotion recognition" refers to the process of analyzing a user's facial expressions and voice in real time to identify their emotions.

[1101] "Proposed revision" refers to proposed changes to the decorative design that are suggested to the user based on the emotion recognition results.

[1102] "Server" refers to the central system that receives and stores image and design data from users, generates virtual try-on images using generative models, and performs emotion recognition.

[1103] "Storage" refers to the data storage location where the server stores users' photos, decorative design data, and virtual try-on images.

[1104] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[1105] System configuration and operation

[1106] Image upload

[1107] Users access the system by taking photos of themselves from various angles using smart glasses or a smartphone. The user selects the photos and sends them to the server by pressing the upload button. The server stores the received user photo data in storage (e.g., AWS S3) and verifies that the photos have been uploaded correctly.

[1108] Decorative design input

[1109] The user inputs the desired decorative design as image data or text instructions. The device sends this data to the server, which then stores it in storage (e.g., AWS S3).

[1110] Virtual try-on image generation

[1111] The server inputs the received user image and desired accessory design into a generative model (e.g., StyleGAN). Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[1112] Emotion recognition by emotion engine

[1113] The user reviews the virtual try-on images and sends their reactions to the server via their device. The device is equipped with a camera and microphone, which capture the user's facial expressions and voice in real time. The server then uses an emotion engine (e.g., Microsoft Azure Emotion API) to analyze the user's facial expressions and voice and identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[1114] Check and edit virtual try-on images

[1115] If the generated virtual try-on image requires any modifications, the user inputs the modifications. These modifications are sent from the device to the server. The server receives the modifications and inputs them back into the generative model. The generative model then regenerates the virtual try-on image based on the new modifications and sends it to the server. This process is repeated until the user is satisfied. Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a certain decoration, the emotion engine will suggest a different color.

[1116] Final Selection

[1117] The user finally selects their favorite decoration design on the device. The device sends the final selected decoration design information to the server. The server confirms the selected decoration design and saves the related data in storage. The server then sends the user a confirmation of the final decision, notifying them that decoration selection is complete.

[1118] Specific examples

[1119] For example, if user B wants to choose a dress at a brand shop, he can use the system as follows.

[1120] 1. User B uses smart glasses in a store to take photos of himself (front, side, and back).

[1121] 2. The server receives and stores the photo.

[1122] 3. User B takes a picture of the dress they want with their smartphone and enters it into the app.

[1123] 4. The server receives this data and inputs it into a generative model (StyleGAN).

[1124] 5. The generative AI model (StyleGAN) generates a virtual try-on image based on "User B's photo + dress image" and sends it to the server.

[1125] 6. The server provides the generated try-on image to User B.

[1126] 7. User B checks the fitting image. The camera in the smart glasses captures their facial expressions and sends them to the emotion engine.

[1127] 8. The emotion engine (Microsoft Azure Emotion API) analyzes facial expressions and initiates suggestions for corrections to areas of dissatisfaction (e.g., "I don't like the color of the red dress").

[1128] 9. User B inputs color correction instructions. The server again uses the generative model to generate an updated image and provides it to User B.

[1129] 10. User B finally chooses the dress they like and confirms it in the app.

[1130] 11. The server saves your selection and completes the process.

[1131] Example prompts for generative AI models

[1132] "Combine the front, side, and back images of the user to create a virtual image of them trying on a specified decorative design (a red dress)."

[1133] In this way, the system of the present invention provides a support system that allows users to check and select their ideal outfit without trying it on in the store.

[1134] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1135] Step 1:

[1136] Image upload

[1137] Users can use smart glasses or smartphones to take photos of themselves from the front, side, or back, then access the application, select the photo they have taken, and press the upload button to send the photo to the server.

[1138] Input: Image data taken from multiple angles by the user

[1139] The server saves the received image data in storage (e.g. AWS S3) and confirms that it has been uploaded successfully.

[1140] Output: Reference information for image data stored in storage

[1141] Step 2:

[1142] Decorative design input

[1143] The user inputs the desired decorative design into the application as image data or text instructions, which are then sent to the server.

[1144] Input: Image data or text instructions for the decorative design entered by the user

[1145] The server stores the received design data in storage (e.g. AWS S3).

[1146] Output: Reference information for decorative designs stored in storage

[1147] Step 3:

[1148] Virtual try-on image generation

[1149] The server inputs the user's photo data and desired decorative design data into a generative model (e.g., StyleGAN).

[1150] Input: User photo data and decorative design data

[1151] Based on this data, the generative model generates a virtual try-on image of the user wearing the accessories, which can be viewed from a 360-degree angle.

[1152] The server stores the generated virtual try-on image in the user's account and provides it to the user.

[1153] Output: Virtual try-on image data

[1154] Step 4:

[1155] Emotion recognition by emotion engine

[1156] The user checks the provided virtual try-on images, and the device's built-in camera and microphone capture the user's facial expressions and voice in real time as they check.

[1157] Input: User's facial expression data and voice data

[1158] The server analyzes the received facial expression and voice data through an emotion engine (e.g., Microsoft Azure Emotion API) to identify the user's emotions.

[1159] Output: Emotion recognition result data

[1160] Step 5:

[1161] Check and edit virtual try-on images

[1162] If the user needs to make any corrections to the virtual try-on image, the user inputs the correction instructions into the application and sends them to the server.

[1163] Input: User-entered correction instructions

[1164] The server receives and analyzes the correction instructions, and then inputs them back into the generative model to generate a corrected virtual try-on image. The generative model then regenerates the virtual try-on image based on the new instructions.

[1165] The server provides the user with new virtual try-on images, and this process is repeated until the user is satisfied.

[1166] Output: Data of the virtual try-on image after correction

[1167] Step 6:

[1168] Final Selection

[1169] The user finally selects the decorative design they like on the terminal.

[1170] Input: Information on the decorative design that the user finally selected

[1171] The terminal transmits the selected decorative design information to the server.

[1172] The server confirms the selected decorative design, stores the associated data in storage, and sends the user a confirmation of the final decision.

[1173] Output: Final information of the saved decorative design

[1174] This specific processing step allows users to virtually try on clothes and efficiently select the most suitable decorative design through emotion recognition.

[1175] 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.

[1176] 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.

[1177] 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.

[1178] [Third embodiment]

[1179] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1180] 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.

[1181] 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).

[1182] 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.

[1183] 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.

[1184] 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).

[1185] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1186] 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.

[1187] 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.

[1188] 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.

[1189] 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.

[1190] 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."

[1191] The present invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative model.

[1192] System configuration and operation

[1193] 1. User image upload

[1194] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[1195] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[1196] 2. Input your desired decorative design

[1197] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[1198] 3. Generation of virtual try-on images

[1199] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[1200] 4. Check and edit the virtual try-on image

[1201] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[1202] The server receives the correction instructions and re-inputs them into the generative model. The generative model generates a corrected virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[1203] 5. Final dress selection

[1204] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[1205] Specific examples

[1206] If user A wants to choose a wedding dress, he can use the system as follows:

[1207] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[1208] 2. The server receives these photos and stores them in storage.

[1209] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[1210] 4. The server receives this data and inputs it into the generative model.

[1211] 5. The generative model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[1212] 6. The server provides the generated virtual try-on image to User A.

[1213] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[1214] 8. The server re-inputs the correction instructions into the generative model, generates an updated image, and provides it to User A.

[1215] 9. User A finally chooses the dress she likes and confirms it on the system.

[1216] 10. The server saves the selection and completes the process.

[1217] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[1218] The processing flow will be explained below.

[1219] Specific steps in the system programming process

[1220] Image upload

[1221] Step 1:

[1222] Users take photos of themselves from multiple angles and use a terminal to access the system.

[1223] Step 2:

[1224] The terminal allows the user to select a photo they have taken and send it to the server by pressing the upload button.

[1225] Step 3:

[1226] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[1227] Decorative design input

[1228] Step 4:

[1229] The user inputs the desired decorative design in the form of image data, text instructions, or both.

[1230] Step 5:

[1231] The terminal transmits the input decorative design data to the server.

[1232] Step 6:

[1233] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[1234] Virtual try-on image generation

[1235] Step 7:

[1236] The server inputs the received user image and desired decorative design into the generative model.

[1237] Step 8:

[1238] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[1239] Step 9:

[1240] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[1241] Step 10:

[1242] The server receives the generated virtual try-on images and stores them in the user's account.

[1243] Check and edit virtual try-on images

[1244] Step 11:

[1245] The user checks the generated virtual try-on image on the terminal.

[1246] Step 12:

[1247] If a specific part of the image needs to be corrected, the user inputs a correction instruction.

[1248] Step 13:

[1249] The terminal sends a correction instruction to the server.

[1250] Step 14:

[1251] The server re-inputs the received correction instructions into the generative model.

[1252] Step 15:

[1253] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[1254] Step 16:

[1255] The server receives the updated virtual try-on images and stores them in the user's account.

[1256] Step 17:

[1257] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[1258] Final Selection

[1259] Step 18:

[1260] The user finally selects the decorative design they like on the terminal.

[1261] Step 19:

[1262] The terminal transmits the final selected decorative design information to the server.

[1263] Step 20:

[1264] The server determines the selected decorative design and stores the associated data in storage.

[1265] Step 21:

[1266] The server sends the user a confirmation of the final decision, informing them that their dress selection is complete.

[1267] Example 1

[1268] 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."

[1269] Conventional try-on systems require users to physically visit a store, which requires time and effort. Even online try-on systems have the drawback of making it difficult for users to make fine adjustments when trying on accessories, making it difficult to fully replicate the actual wearing experience. Furthermore, there is a lack of a way for users to check the accessories in detail from multiple angles.

[1270] 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.

[1271] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting a desired ornament design, and a generative model for generating a virtual try-on image of the user wearing the ornament based on the received user image and the desired ornament design. This allows users to try on an unlimited number of ornaments from the comfort of their own home and find their perfect outfit. The generated virtual try-on image can also be viewed from a 360-degree angle, making it easy for users to check the details of the desired ornament. Furthermore, by re-executing the generative model based on correction instructions and updating the virtual try-on image, users can adjust the ornament design to their satisfaction.

[1272] "User images" are photographic data taken by a user of himself or herself, taken from multiple angles.

[1273] A "desired accessory design" is a design of accessory or clothing that the user wishes to try on, and may take the form of image data or text instructions.

[1274] A "generative model" is an artificial intelligence model that generates virtual try-on images using a user's image and desired decorative design as input, and has image generation and editing functions.

[1275] A "virtual try-on image" is an image generated by a generative model that virtually recreates the appearance of a user wearing an accessory.

[1276] "Modification instructions" are requests for changes or adjustments made by the user to the virtual try-on image, and include specific changes to parts or styles.

[1277] "360-degree angle" means that the virtual try-on image can be viewed from all directions, allowing the user to check the decorations from multiple angles.

[1278] "Server" means a centralized computer system that receives, stores, and processes images and data from users.

[1279] This invention relates to a system that allows users to virtually try on accessories by taking pictures of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative AI model.

[1280] System configuration and operation

[1281] 1. User image upload

[1282] Users take pictures of themselves from the front, side, and back, and then access the system using a terminal. Users select the images they have taken and press the upload button to send them to the server.

[1283] The server stores the received user image data in cloud storage, confirms that the image has been uploaded correctly, and notifies the user.

[1284] Hardware used: smartphone, PC, cloud server

[1285] Software used: web browser, photo-taking application, cloud storage system (e.g., Amazon S3)

[1286] 2. Input your desired decorative design

[1287] The user inputs the desired decorative design as image data or text instructions. The terminal transmits this data to the server,

[1288] The server stores the received design data in cloud storage.

[1289] Hardware used: smartphone, PC, cloud server

[1290] Software used: Web forms, text input interface

[1291] 3. Generation of virtual try-on images

[1292] The server inputs the received user image and desired decorative design into a generative AI model.

[1293] The generative AI model uses this data to generate virtual try-on images of the user wearing the accessories, which can be viewed from a 360-degree angle.

[1294] The server provides this image to the user.

[1295] Hardware used: Cloud server

[1296] Software used: Generative AI models (e.g., DALL-E, Stable Diffusion), databases

[1297] 4. Check and edit the virtual try-on image

[1298] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[1299] The server receives the correction instructions and feeds them back into the generative AI model, which then generates a modified virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[1300] Hardware used: smartphone, PC, cloud server

[1301] Software used: Generative AI models, web forms

[1302] 5. Final selection of decorative design

[1303] The user finally selects the decorative design they like and notifies the server via their terminal.

[1304] The server stores the information and confirms the final decision.

[1305] Hardware used: smartphone, PC, cloud server

[1306] Software used: Web forms, databases

[1307] Specific examples

[1308] If user A wants to choose a wedding dress, he can use the system as follows:

[1309] 1. User A takes pictures of himself from the front, side, and back and uploads them to the system.

[1310] Example prompt: "Please upload a photo of your front, side, and back."

[1311] 2. The server receives these images and stores them in cloud storage.

[1312] Example prompt: "Your image was successfully uploaded."

[1313] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[1314] Example prompt: "Please upload a photo of the dress you would like."

[1315] 4. The server receives this data and inputs it into a generative AI model.

[1316] 5. The generative AI model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[1317] 6. The server provides the generated virtual try-on image to User A.

[1318] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[1319] Example prompt: "Please make the hem a little shorter."

[1320] 8. The server re-inputs the correction instructions into the generative AI model, generates an updated image, and provides it to User A.

[1321] 9. User A finally chooses the dress she likes and confirms it on the system.

[1322] 10. The server saves the selection and completes the process.

[1323] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[1324] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1325] Step 1: Take and upload a user image

[1326] Specific behavior:

[1327] Users take photos of themselves from the front, side, and back using a smartphone or camera.

[1328] Save the captured image file to a folder on your device.

[1329] Input: Captured image files (front, side, back)

[1330] Output: Image file saved on the device

[1331] Step 2: Access the system and upload your images

[1332] Specific behavior:

[1333] The user opens a web browser on the terminal, accesses the system's web application, and logs in.

[1334] Access the image upload form, select the image file you took, and press the upload button.

[1335] Input: Image files stored on the device, user login information

[1336] Output: Image file sent to the server

[1337] Step 3: The server receives and stores the image data.

[1338] Specific behavior:

[1339] The server saves the image data sent by the user in cloud storage and records the file path of the saved data in a database.

[1340] Verify that the image was uploaded correctly, generate a success message, and notify the user.

[1341] Input: Image data sent by the user

[1342] Output: Image data saved in cloud storage, success message

[1343] Step 4: User inputs desired decorative design

[1344] Specific behavior:

[1345] The user inputs image data and text instructions with the desired decorative design into the terminal, for example, by uploading an image of the dress taken with a smartphone.

[1346] Access the input form, select the design data, and press the submit button.

[1347] Input: Image data of decorative design taken with a smartphone or text instructions

[1348] Output: Decorative design data sent to the server

[1349] Step 5: The server receives and stores the design data

[1350] Specific behavior:

[1351] The server saves the decorative design data sent by the user in cloud storage, and records the file path and text content of the saved data in a database.

[1352] Input: Decoration design data submitted by the user

[1353] Output: Design data stored in cloud storage, file paths and text contents recorded in the database

[1354] Step 6: Start generating virtual try-on images

[1355] Specific behavior:

[1356] The server performs preprocessing to input the image data captured by the user and the desired decorative design data into the generative AI model.

[1357] The data is converted into the appropriate format and fed into a generative AI model.

[1358] Input: User image data stored in cloud storage, decorative design data stored in cloud storage

[1359] Output: The data that is fed into the generative AI model

[1360] Step 7: The generative AI model generates virtual try-on images

[1361] Specific behavior:

[1362] The generative AI model generates virtual try-on images based on the input user image data and decorative design data.

[1363] The generated try-on images are processed so that they can be displayed from multiple angles.

[1364] Input: User image data and decorative design data input into the generative AI model

[1365] Output: Generated virtual try-on image

[1366] Step 8: The server stores and serves the generated images.

[1367] Specific behavior:

[1368] The server saves the virtual try-on images sent from the generative AI model to cloud storage and records the file paths of the saved images in a database.

[1369] The user is notified of the URL of the generated virtual try-on image.

[1370] Input: Virtual try-on images sent from the generative AI model

[1371] Output: Virtual try-on images saved in cloud storage, notification to user

[1372] Step 9: The user checks the fitting image and inputs correction instructions.

[1373] Specific behavior:

[1374] The user can check the virtual fitting images provided on the device and input specific correction instructions as needed, such as "Please make the hem a little shorter."

[1375] Input: Provided virtual try-on image, modification instructions

[1376] Output: Correction instructions sent to the server

[1377] Step 10: The server re-inputs the correction instructions into the generative AI model.

[1378] Specific behavior:

[1379] The server performs preprocessing to re-input the received correction instructions into the generative AI model.

[1380] The correction instructions are input into the generative AI model, requesting the generation of an updated virtual try-on image.

[1381] Input: Correction instructions sent to the server

[1382] Output: Corrective instruction data that is input to the generative AI model

[1383] Step 11: The generative AI model generates the corrected image

[1384] Specific behavior:

[1385] The generative AI model generates new virtual try-on images based on the correction instructions and sends them to the server.

[1386] Input: Correction instructions input to the generative AI model

[1387] Output: Corrected virtual try-on image

[1388] Step 12: The server stores and serves the corrected image.

[1389] Specific behavior:

[1390] The server saves the newly generated virtual try-on image in cloud storage, records the file path of the saved image in the database, and notifies the user of the URL of the modified virtual try-on image.

[1391] Input: Modified virtual try-on image sent from the generative AI model

[1392] Output: Edited try-on image saved in cloud storage, notification to user

[1393] Step 13: User selects and notifies final design

[1394] Specific behavior:

[1395] The user finally checks the virtual try-on images to find one that satisfies him / her, selects the final decorative design, and notifies the server of the selection.

[1396] Input: Final selected virtual try-on image, selection details

[1397] Output: Final selection notification sent to the server

[1398] Step 14: The server saves the selection and ends the process

[1399] Specific behavior:

[1400] The server stores the information of the final selection notified by the user in the cloud storage and the database, and notifies the user that the final decision has been made.

[1401] Input: Final selection notification sent to the server

[1402] Output: Final selection information stored in cloud storage and database, completion notification to user

[1403] (Application example 1)

[1404] 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."

[1405] Conventional virtual try-on systems have the drawback of requiring a great deal of time and effort for users to check decorations, and the decorations often deviate from the actual image. The present invention aims to provide a system that allows users to virtually try on decorations in real time, and to quickly and effectively check and modify the decorations.

[1406] 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.

[1407] In this invention, the server includes: means for a user to take and upload photos of themselves from multiple angles; means for inputting desired ornaments; a generative model for generating a virtual try-on image of the user wearing the ornaments based on the received user photo and desired ornament design; means for the user to provide the generated virtual try-on image to the user; means for the user to issue correction instructions for the virtual try-on image; means for re-executing the generative model based on the correction instructions to update the virtual try-on image; means for confirming and saving the final selected ornament design; means for using a smartphone to check the generated virtual try-on image in real time; and means for the user to input the details of the desired ornament design as a prompt statement and for the system to control the generative model based on the prompt statement. This allows the user to check and correct the virtual try-on image of the ornaments in real time using their smartphone and quickly and efficiently select a final ornament design.

[1408] A "smartphone" is a type of mobile information terminal that has mobile phone functions and can connect to the Internet and use various applications.

[1409] A "virtual try-on image" is an image that simulates the appearance of wearing an accessory, created by a generative model based on the user's image and the desired accessory design.

[1410] A "generative model" is an algorithm or machine learning model that uses computer vision technology to generate virtual try-on images based on a received user photo and decorative design.

[1411] A "prompt sentence" is an input sentence that succinctly expresses the content of the decorative design desired by the user, and is important information for the system to control the generative model based on the instructions.

[1412] "Modification instructions" are instructions that the user inputs to make changes or adjustments to the virtual try-on image, and serve as a trigger for the generative model to be re-executed.

[1413] "Real-time" refers to the fact that the user can instantly check and correct the results of the ornament try-on, and means that the system operates with high response speed.

[1414] This invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. The invention is implemented using a server, a terminal, and a generative AI model.

[1415] System configuration and operation

[1416] 1. User image upload

[1417] Users take photos of themselves from various angles with their smartphones, access the system, select the photos they have taken, and press the upload button to send them to the server.

[1418] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[1419] 2. Input your desired decorative design

[1420] The user inputs the desired decorative design as a prompt, such as a simple text like "I want to try on a red dress." The device sends this prompt to the server, which then stores the received data in its storage.

[1421] 3. Generation of virtual try-on images

[1422] The server inputs the received user image and desired accessory design into a generative AI model. Based on this data, the generative AI model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[1423] 4. Check and edit the virtual try-on image

[1424] The user checks the generated virtual try-on image on their smartphone. If they don't like a particular part of the decorative design, they can input a prompt to make corrections. For example, they could give a specific instruction such as "Make the hem of the dress a little longer." The device then sends this correction instruction to the server.

[1425] The server receives the correction instructions and feeds them back into the generative AI model. The generative model generates a revised virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[1426] 5. Final dress selection

[1427] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[1428] Hardware and software used

[1429] The system utilizes the following hardware and software:

[1430] Hardware: Smartphones, servers, GPUs (deep learning-compatible GPUs such as NVIDIA)

[1431] Software: Python, FastAPI (web framework), Torch (deep learning library), PIL (Pillow, image processing library)

[1432] As a concrete example, if a user wants to try on a "red dress," he or she inputs the following prompt into the system:

[1433] Prompt: I want to try on a red dress

[1434] This allows the system of the present invention to allow users to try on an unlimited number of decorations from the convenience of their own home and find the perfect one.

[1435] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1436] Step 1:

[1437] Users take photos of themselves from multiple angles and upload them via their devices. They use their smartphones to take photos of the front, side, back, etc., and press the upload button to send them to the system. The input is the user's image data, and the output is image data stored on the server.

[1438] Step 2:

[1439] The server saves the received user photo data in storage, verifies that the photo was uploaded correctly, and notifies the user. The server saves the image file in storage, verifies that it was saved successfully, and notifies the user of this confirmation information.

[1440] Step 3:

[1441] The user inputs the desired decorative design as a prompt sentence and sends it to the server via the terminal. The input includes text data such as "I would like to try on a red dress." The output is the text data saved on the server.

[1442] Step 4:

[1443] The server inputs the received user image and prompt text into the generative AI model. The user image (input image data) and prompt text (text data) are input into the generative model. Based on this data, the generative AI model generates a virtual try-on image. The output is the generated try-on image data.

[1444] Step 5:

[1445] The generative model is executed to generate a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. This image data is returned to the server and saved in the user's account. The output is a virtual try-on image saved in the user's account.

[1446] Step 6:

[1447] The user checks the generated virtual try-on image through the terminal. The generated image is displayed on the terminal, and the user can see the fully decorated try-on image of themselves. This includes both the input and output of the step.

[1448] Step 7:

[1449] If the user does not like a particular part of the decorative design, they can send a prompt to the server from their terminal with instructions to make corrections. For example, they can input an instruction such as "Make the hem of the dress a little longer." The input is the prompt to make corrections, and the output is correction instruction data stored on the server.

[1450] Step 8:

[1451] The server inputs the correction instructions into the generative AI model again, and the generative model regenerates the corrected virtual try-on image. The inputs are the user image, the existing virtual try-on image, and the correction prompt. The output is the corrected try-on image data.

[1452] Step 9:

[1453] The server provides the generated modified virtual try-on image to the user. The data is sent to the user's device, allowing the user to view the new try-on image. The output is the modified try-on image displayed by the user.

[1454] Step 10:

[1455] The user finally selects the decorative design they like and notifies the server via their terminal. The user presses the decision button to send the design to the server. The input at this time is the user's final decision data, and the output is the final selection information stored on the server.

[1456] 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.

[1457] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[1458] System configuration and operation

[1459] Image upload

[1460] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[1461] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[1462] Decorative design input

[1463] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[1464] Virtual try-on image generation

[1465] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[1466] Emotion recognition by emotion engine

[1467] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[1468] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[1469] Check and edit virtual try-on images

[1470] If the user needs to make any corrections to the generated virtual try-on image, the user inputs correction instructions, which are then sent from the terminal to the server.

[1471] The server receives the correction instructions and re-inputs them into the generative model. The generative model regenerates the virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[1472] Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a decoration, the emotion engine will suggest a different color.

[1473] Final Selection

[1474] The user finally selects the decorative design they like on the terminal.

[1475] The terminal transmits the final selected decorative design information to the server.

[1476] The server determines the selected decorative design and stores the associated data in storage.

[1477] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[1478] Specific examples

[1479] For example, if user A wants to choose a kimono for his / her coming-of-age ceremony, he / she can use the system as follows:

[1480] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[1481] 2. The server receives these photos and stores them in storage.

[1482] 3. User A takes a picture of a kimono in the catalog with their smartphone and inputs it into the system.

[1483] 4. The server receives this data and inputs it into the generative model.

[1484] 5. The generative model generates a virtual try-on image of User A wearing the kimono and sends it to the server.

[1485] 6. The server provides the generated virtual try-on image to User A.

[1486] 7. When User A checks the try-on images on the device, he / she sends his / her reaction to the emotion engine via the camera and microphone.

[1487] 8. The emotion engine analyzes User A's facial expressions and voice and reports the current emotion to the server.

[1488] 9. If user A is dissatisfied with a particular part of the try-on image, he or she inputs correction instructions.

[1489] 10. The server re-inputs the correction instructions into the generative model, generates an updated image, and serves it to User A. This process is repeated based on feedback from the emotion engine.

[1490] 11. User A finally chooses the decorative design he likes and confirms it on the system.

[1491] 12. The server saves the selection and completes the process.

[1492] In this way, the system of the present invention allows users to try on an infinite number of outfits from the comfort of their own home and find their perfect outfit through emotion recognition.

[1493] The processing flow will be explained below.

[1494] Specific steps in the system programming process

[1495] Image upload

[1496] Step 1:

[1497] Users take photos of themselves from various angles and access the system using a terminal.

[1498] Step 2:

[1499] The device selects multiple photos taken and sends them to the server by pressing the upload button.

[1500] Step 3:

[1501] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[1502] Decorative design input

[1503] Step 4:

[1504] The user inputs the desired decorative design as image data or text instructions.

[1505] Step 5:

[1506] The terminal transmits the input decorative design data to the server.

[1507] Step 6:

[1508] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[1509] Virtual try-on image generation

[1510] Step 7:

[1511] The server inputs the received user image and desired decorative design into the generative model.

[1512] Step 8:

[1513] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[1514] Step 9:

[1515] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[1516] Step 10:

[1517] The server receives the generated virtual try-on images and stores them in the user's account.

[1518] Emotion recognition by emotion engine

[1519] Step 11:

[1520] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[1521] Step 12:

[1522] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions, and evaluates the virtual try-on images based on this information.

[1523] Check and edit virtual try-on images

[1524] Step 13:

[1525] If the user needs to make any corrections to the generated virtual try-on image, they can input specific corrections, such as "change the hem length."

[1526] Step 14:

[1527] The terminal sends a correction instruction to the server.

[1528] Step 15:

[1529] The server re-inputs the received correction instructions into the generative model.

[1530] Step 16:

[1531] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[1532] Step 17:

[1533] The server receives the updated virtual try-on images and stores them in the user's account.

[1534] Step 18:

[1535] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[1536] Step 19:

[1537] The emotion engine monitors the user's emotions, identifies areas of dissatisfaction, and provides that information to the server.

[1538] Step 20:

[1539] Based on the information from the emotion engine, the server suggests appropriate decorative designs and modifications to the user.

[1540] Final Selection

[1541] Step 21:

[1542] The user finally selects the decorative design they like on the terminal.

[1543] Step 22:

[1544] The terminal transmits the final selected decorative design information to the server.

[1545] Step 23:

[1546] The server determines the selected decorative design and stores the associated data in storage.

[1547] Step 24:

[1548] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[1549] Example 2

[1550] 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."

[1551] Conventional virtual try-on systems have a limited process for generating virtual try-on images based on user-provided images and decorative designs, making it difficult to effectively incorporate user emotions and feedback. As a result, users are often dissatisfied with their final selection and are forced to go through multiple trial and error rounds. Furthermore, the quality of try-on images and real-time adjustments are insufficient, leaving room for improvement in the user experience.

[1552] 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.

[1553] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting requested accessories, and an image generation model for generating virtual try-on images of the user wearing the accessories based on the received user images and the requested accessory design. This allows users to generate high-quality virtual try-on images in real time and reflect feedback using an emotion recognition engine.

[1554] A "user" is someone who uses the system to upload a photo of themselves and try on clothes virtually.

[1555] "Terminal" means an electronic device used by a user to access the system and input photos and decorative designs. Examples include smartphones, tablets, and PCs.

[1556] The "server" is a central processing unit that processes data received from users and operates generative AI models and emotion recognition engines.

[1557] A "generative model" is an algorithm or software that generates virtual try-on images based on a received user image and requested accessory design, often using deep learning techniques.

[1558] A "virtual try-on image" is a virtual image generated based on a user's image, showing an outfit with a decorative design specified by the user.

[1559] An "emotion recognition engine" is an algorithm or software that analyzes a user's facial expressions, voice, etc. to identify the user's emotions.

[1560] "Modification instructions" refer to requests for changes made by the user to the virtual try-on image, and specifically include changes to the color or shape of the decorative design.

[1561] "Decorative design" is data that represents the appearance of the clothes, accessories, etc. that the user wishes to try on in the virtual try-on image.

[1562] "Reaction" refers to the facial and vocal feedback given by the user when they view the virtual try-on image.

[1563] "Suggestions" are suggestions for alternative decorative designs to improve user satisfaction, generated by the emotion recognition engine based on the user's reactions.

[1564] These definitions will provide a clear understanding of each claim element.

[1565] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative design. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. Specifically, the system of the present invention is implemented using a server, a terminal, an image generation model, and an emotion recognition engine.

[1566] System configuration

[1567] 1. A device where users take and upload their own images

[1568] Users use devices such as smartphones, tablets, and computers to take and upload images of themselves from various angles.

[1569] The device is equipped with a camera, and the user uses a dedicated application to select and upload images to the system.

[1570] 2. Server

[1571] The server processes the image data received from the user and stores it in storage.

[1572] The server also receives and manages decorative design data sent by users.

[1573] Furthermore, the server operates an image generation model based on the received data to generate a virtual try-on image.

[1574] The server then runs the image generation model again based on the generated virtual try-on images and user feedback.

[1575] 3. Image Generation Model

[1576] The image generation model takes the user's image and decorative design data as input and generates a virtual try-on image, primarily using deep learning technology.

[1577] Virtual try-on images are generated in a format that can be viewed from a 360-degree angle.

[1578] 4. Emotion Recognition Engine

[1579] The emotion recognition engine captures the user's facial expressions and voice through the device's camera and microphone and analyzes the user's emotions in real time.

[1580] Based on the results of this analysis, the emotion recognition engine suggests appropriate decorative designs.

[1581] Specific examples

[1582] Below is a concrete example of how User A uses the system to choose a kimono for her coming-of-age ceremony:

[1583] 1. User A uses a smartphone to take pictures of himself from three angles (front, side, and back) and uploads them to the system via a dedicated application.

[1584] 2. The server receives the image data and stores it in storage.

[1585] 3. User A takes a picture of a kimono from a coming-of-age ceremony catalogue with his smartphone and inputs it into the system as decorative design data.

[1586] 4. The server receives this decorative design data and inputs it into the image generation model.

[1587] 5. The image generation model generates a virtual try-on image of User A wearing the kimono for his coming-of-age ceremony and sends it to the server.

[1588] 6. The server provides the generated virtual try-on image to User A.

[1589] 7. User A checks the virtual try-on image on their device and sends their reaction to the emotion engine via the camera and microphone.

[1590] 8. The emotion engine analyzes User A's facial expressions and voice to determine which designs User A is satisfied or dissatisfied with.

[1591] 9. If user A is dissatisfied with a particular part, he or she enters correction instructions into the terminal and sends them to the server.

[1592] 10. The server receives the correction instructions and again instructs the image generation model to generate updated virtual try-on images.

[1593] 11. The server provides the updated image to User A, and the process repeats until User A is satisfied.

[1594] In this way, the system of the present invention allows users to try on an infinite number of decorations from the comfort of their own home and find their perfect design through emotion recognition.

[1595] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1596] Step 1:

[1597] Users take multiple photos of themselves from various angles, such as from the front, side, and back, select the images through a dedicated application, and press the upload button.

[1598] Input: Images taken by the user from multiple angles

[1599] Output: User's image data

[1600] Step 2:

[1601] The terminal sends the user's image data to the server, where the image data is transferred over the network.

[1602] Input: User's image data

[1603] Output: Image data sent to the server

[1604] Step 3:

[1605] The server stores the received image data of the user in storage, and also generates a message confirming receipt and notifies the user.

[1606] Input: Image data sent from the device

[1607] Output: Image data saved in storage, receipt confirmation message

[1608] Step 4:

[1609] The user inputs the desired decorative design into the terminal as image data or text instructions. For example, the user may take a photo of a kimono for their coming-of-age ceremony and upload it.

[1610] Input: User-provided image data or text instructions for the decorative design

[1611] Output: decorative design data

[1612] Step 5:

[1613] The terminal transmits the input decorative design data to the server, where data transfer also takes place via the network.

[1614] Input: decorative design data

[1615] Output: Decorative design data sent to the server

[1616] Step 6:

[1617] The server stores the received decorative design data in storage, and also generates a confirmation message and notifies the user.

[1618] Input: Decoration design data sent from the device

[1619] Output: Decorative design data saved in storage, confirmation message

[1620] Step 7:

[1621] The server inputs the received user image data and decorative design data into an image generation model, specifically, a deep learning algorithm to generate virtual try-on images.

[1622] Input: User image data, decorative design data

[1623] Output: Generated virtual try-on image

[1624] Step 8:

[1625] The server links the generated virtual try-on image to the user's account and provides it to the user, who can view it on their device.

[1626] Input: Generated virtual try-on images

[1627] Output: A link to the virtual try-on image provided to the user

[1628] Step 9:

[1629] The user checks the virtual fitting images on the device and sends their reactions to the emotion engine via the camera and microphone. The device captures facial expressions and voice.

[1630] Input: Virtual try-on image for user to view

[1631] Output: Captured user reaction data

[1632] Step 10:

[1633] The server analyzes the captured user reaction data using an emotion recognition engine, specifically facial detection and voice tone analysis, to identify user satisfaction or dissatisfaction.

[1634] Input: User response data

[1635] Output: User sentiment analysis results

[1636] Step 11:

[1637] If the user needs to make any modifications to the virtual try-on image, the user can input and send the modification instructions to the terminal, for example, by inputting specific instructions such as "change the color to blue."

[1638] Input: Correction instructions

[1639] Output: Correction instructions typed into the terminal

[1640] Step 12:

[1641] The terminal sends the correction instruction to the server, and data is transferred via the network.

[1642] Input: Correction instructions

[1643] output: Correction instructions sent to the server

[1644] Step 13:

[1645] The server receives the correction instructions and re-inputs them into the generative AI model, which then generates updated virtual try-on images based on the new instructions.

[1646] Input: Correction instructions

[1647] Output: Updated virtual try-on image

[1648] Step 14:

[1649] The server saves the updated virtual try-on image to the user's account and provides it to the user again via the link, and this process is repeated until the user is satisfied.

[1650] Input: Updated virtual try-on image

[1651] output: Updated virtual try-on image link provided to the user

[1652] Step 15:

[1653] The user finally makes a decision to select the decorative design they like, and presses the confirmation button to confirm the selection through the terminal.

[1654] Input: Final decorative design selection

[1655] Output: Selection information

[1656] Step 16:

[1657] The terminal transmits the final selection information to the server, and data transfer occurs via the network.

[1658] Input: Final selection information

[1659] Output: Final selection information sent to the server

[1660] Step 17:

[1661] The server confirms the selected decoration design, stores the related data in storage, and sends a final decision confirmation to the user, informing them that the decoration selection is complete.

[1662] Input: Final selection information

[1663] Output: Selected design data saved in storage, confirmation notification to user

[1664] (Application example 2)

[1665] 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."

[1666] In conventional virtual try-on systems, users had to rely solely on visual confirmation when checking try-on images. This made it difficult to reflect the user's emotions and intuitive reactions, making it difficult for them to select decorative designs that truly satisfied the user. Furthermore, if the user was dissatisfied with the try-on images, suggestions for revisions could not be made efficiently, resulting in a problem of a decline in the quality of the try-on experience. There is a need to solve these problems.

[1667] 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.

[1668] In this invention, the server includes means for a user to take and upload photos of themselves from multiple angles, means for inputting desired decorations, a generative model for generating a virtual try-on image of the user wearing the decorations based on the received user photo and desired decoration design, means for providing the generated virtual try-on image to the user, means for the user to issue correction instructions for the virtual try-on image, means for re-executing the generative model based on the correction instructions to update the virtual try-on image, means for finalizing and saving the selected decoration design, means for capturing the user's facial expressions and voice, recognizing and analyzing emotions, and means for presenting correction suggestions to the user based on the emotion recognition results.

[1669] This allows the system to reflect the user's emotions in real time, efficiently suggest modifications based on intuitive feedback, and select the decorative design that will satisfy the user most.

[1670] "User" refers to a person who uses this system to upload their own photos and try on clothes virtually.

[1671] "Photos" refer to image data taken by a user from multiple angles.

[1672] "Decorations" refer to design elements such as clothing and accessories that the user wishes to try on.

[1673] "Input" refers to the user inputting the desired decorative design into the system.

[1674] "Generative model" refers to a machine learning model or algorithm that generates virtual try-on images based on the user's received photo and desired decorative design.

[1675] A "virtual try-on image" refers to an image that virtually depicts the state in which a user is wearing the accessory.

[1676] "Modification instructions" refer to instructions for changes or modifications that the user makes to the generated virtual try-on image.

[1677] "Emotion recognition" refers to the process of analyzing a user's facial expressions and voice in real time to identify their emotions.

[1678] "Proposed revision" refers to proposed changes to the decorative design that are suggested to the user based on the emotion recognition results.

[1679] "Server" refers to the central system that receives and stores image and design data from users, generates virtual try-on images using generative models, and performs emotion recognition.

[1680] "Storage" refers to the data storage location where the server stores users' photos, decorative design data, and virtual try-on images.

[1681] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[1682] System configuration and operation

[1683] Image upload

[1684] Users access the system by taking photos of themselves from various angles using smart glasses or a smartphone. The user selects the photos and sends them to the server by pressing the upload button. The server stores the received user photo data in storage (e.g., AWS S3) and verifies that the photos have been uploaded correctly.

[1685] Decorative design input

[1686] The user inputs the desired decorative design as image data or text instructions. The device sends this data to the server, which then stores it in storage (e.g., AWS S3).

[1687] Virtual try-on image generation

[1688] The server inputs the received user image and desired accessory design into a generative model (e.g., StyleGAN). Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[1689] Emotion recognition by emotion engine

[1690] The user reviews the virtual try-on images and sends their reactions to the server via their device. The device is equipped with a camera and microphone, which capture the user's facial expressions and voice in real time. The server then uses an emotion engine (e.g., Microsoft Azure Emotion API) to analyze the user's facial expressions and voice and identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[1691] Check and edit virtual try-on images

[1692] If the generated virtual try-on image requires any modifications, the user inputs the modifications. These modifications are sent from the device to the server. The server receives the modifications and inputs them back into the generative model. The generative model then regenerates the virtual try-on image based on the new modifications and sends it to the server. This process is repeated until the user is satisfied. Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a certain decoration, the emotion engine will suggest a different color.

[1693] Final Selection

[1694] The user finally selects their favorite decoration design on the device. The device sends the final selected decoration design information to the server. The server confirms the selected decoration design and saves the related data in storage. The server then sends the user a confirmation of the final decision, notifying them that decoration selection is complete.

[1695] Specific examples

[1696] For example, if user B wants to choose a dress at a brand shop, he can use the system as follows.

[1697] 1. User B uses smart glasses in a store to take photos of himself (front, side, and back).

[1698] 2. The server receives and stores the photo.

[1699] 3. User B takes a picture of the dress they want with their smartphone and enters it into the app.

[1700] 4. The server receives this data and inputs it into a generative model (StyleGAN).

[1701] 5. The generative AI model (StyleGAN) generates a virtual try-on image based on "User B's photo + dress image" and sends it to the server.

[1702] 6. The server provides the generated try-on image to User B.

[1703] 7. User B checks the fitting image. The camera in the smart glasses captures their facial expressions and sends them to the emotion engine.

[1704] 8. The emotion engine (Microsoft Azure Emotion API) analyzes facial expressions and initiates suggestions for corrections to areas of dissatisfaction (e.g., "I don't like the color of the red dress").

[1705] 9. User B inputs color correction instructions. The server again uses the generative model to generate an updated image and provides it to User B.

[1706] 10. User B finally chooses the dress they like and confirms it in the app.

[1707] 11. The server saves your selection and completes the process.

[1708] Example prompts for generative AI models

[1709] "Combine the front, side, and back images of the user to create a virtual image of them trying on a specified decorative design (a red dress)."

[1710] In this way, the system of the present invention provides a support system that allows users to check and select their ideal outfit without trying it on in the store.

[1711] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1712] Step 1:

[1713] Image upload

[1714] Users can use smart glasses or smartphones to take photos of themselves from the front, side, or back, then access the application, select the photo they have taken, and press the upload button to send the photo to the server.

[1715] Input: Image data taken from multiple angles by the user

[1716] The server saves the received image data in storage (e.g. AWS S3) and confirms that it has been uploaded successfully.

[1717] Output: Reference information for image data stored in storage

[1718] Step 2:

[1719] Decorative design input

[1720] The user inputs the desired decorative design into the application as image data or text instructions, which are then sent to the server.

[1721] Input: Image data or text instructions for the decorative design entered by the user

[1722] The server stores the received design data in storage (e.g. AWS S3).

[1723] Output: Reference information for decorative designs stored in storage

[1724] Step 3:

[1725] Virtual try-on image generation

[1726] The server inputs the user's photo data and desired decorative design data into a generative model (e.g., StyleGAN).

[1727] Input: User photo data and decorative design data

[1728] Based on this data, the generative model generates a virtual try-on image of the user wearing the accessories, which can be viewed from a 360-degree angle.

[1729] The server stores the generated virtual try-on image in the user's account and provides it to the user.

[1730] Output: Virtual try-on image data

[1731] Step 4:

[1732] Emotion recognition by emotion engine

[1733] The user checks the provided virtual try-on images, and the device's built-in camera and microphone capture the user's facial expressions and voice in real time as they check.

[1734] Input: User's facial expression data and voice data

[1735] The server analyzes the received facial expression and voice data through an emotion engine (e.g., Microsoft Azure Emotion API) to identify the user's emotions.

[1736] Output: Emotion recognition result data

[1737] Step 5:

[1738] Check and edit virtual try-on images

[1739] If the user needs to make any corrections to the virtual try-on image, the user inputs the correction instructions into the application and sends them to the server.

[1740] Input: User-entered correction instructions

[1741] The server receives and analyzes the correction instructions, and then inputs them back into the generative model to generate a corrected virtual try-on image. The generative model then regenerates the virtual try-on image based on the new instructions.

[1742] The server provides the user with new virtual try-on images, and this process is repeated until the user is satisfied.

[1743] Output: Data of the virtual try-on image after correction

[1744] Step 6:

[1745] Final Selection

[1746] The user finally selects the decorative design they like on the terminal.

[1747] Input: Information on the decorative design that the user finally selected

[1748] The terminal transmits the selected decorative design information to the server.

[1749] The server confirms the selected decorative design, stores the associated data in storage, and sends the user a confirmation of the final decision.

[1750] Output: Final information of the saved decorative design

[1751] This specific processing step allows users to virtually try on clothes and efficiently select the most suitable decorative design through emotion recognition.

[1752] 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.

[1753] 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.

[1754] 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.

[1755] [Fourth embodiment]

[1756] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1757] 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.

[1758] 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).

[1759] 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.

[1760] 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.

[1761] 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).

[1762] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1763] 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.

[1764] 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.

[1765] 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.

[1766] 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.

[1767] 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.

[1768] 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."

[1769] The present invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative model.

[1770] System configuration and operation

[1771] 1. User image upload

[1772] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[1773] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[1774] 2. Input your desired decorative design

[1775] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[1776] 3. Generation of virtual try-on images

[1777] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[1778] 4. Check and edit the virtual try-on image

[1779] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[1780] The server receives the correction instructions and re-inputs them into the generative model. The generative model generates a corrected virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[1781] 5. Final dress selection

[1782] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[1783] Specific examples

[1784] If user A wants to choose a wedding dress, he can use the system as follows:

[1785] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[1786] 2. The server receives these photos and stores them in storage.

[1787] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[1788] 4. The server receives this data and inputs it into the generative model.

[1789] 5. The generative model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[1790] 6. The server provides the generated virtual try-on image to User A.

[1791] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[1792] 8. The server re-inputs the correction instructions into the generative model, generates an updated image, and provides it to User A.

[1793] 9. User A finally chooses the dress she likes and confirms it on the system.

[1794] 10. The server saves the selection and completes the process.

[1795] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[1796] The processing flow will be explained below.

[1797] Specific steps in the system programming process

[1798] Image upload

[1799] Step 1:

[1800] Users take photos of themselves from multiple angles and use a terminal to access the system.

[1801] Step 2:

[1802] The terminal allows the user to select a photo they have taken and send it to the server by pressing the upload button.

[1803] Step 3:

[1804] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[1805] Decorative design input

[1806] Step 4:

[1807] The user inputs the desired decorative design in the form of image data, text instructions, or both.

[1808] Step 5:

[1809] The terminal transmits the input decorative design data to the server.

[1810] Step 6:

[1811] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[1812] Virtual try-on image generation

[1813] Step 7:

[1814] The server inputs the received user image and desired decorative design into the generative model.

[1815] Step 8:

[1816] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[1817] Step 9:

[1818] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[1819] Step 10:

[1820] The server receives the generated virtual try-on images and stores them in the user's account.

[1821] Check and edit virtual try-on images

[1822] Step 11:

[1823] The user checks the generated virtual try-on image on the terminal.

[1824] Step 12:

[1825] If a specific part of the image needs to be corrected, the user inputs a correction instruction.

[1826] Step 13:

[1827] The terminal sends a correction instruction to the server.

[1828] Step 14:

[1829] The server re-inputs the received correction instructions into the generative model.

[1830] Step 15:

[1831] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[1832] Step 16:

[1833] The server receives the updated virtual try-on images and stores them in the user's account.

[1834] Step 17:

[1835] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[1836] Final Selection

[1837] Step 18:

[1838] The user finally selects the decorative design they like on the terminal.

[1839] Step 19:

[1840] The terminal transmits the final selected decorative design information to the server.

[1841] Step 20:

[1842] The server determines the selected decorative design and stores the associated data in storage.

[1843] Step 21:

[1844] The server sends the user a confirmation of the final decision, informing them that their dress selection is complete.

[1845] Example 1

[1846] 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."

[1847] Conventional try-on systems require users to physically visit a store, which requires time and effort. Even online try-on systems have the drawback of making it difficult for users to make fine adjustments when trying on accessories, making it difficult to fully replicate the actual wearing experience. Furthermore, there is a lack of a way for users to check the accessories in detail from multiple angles.

[1848] 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.

[1849] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting a desired ornament design, and a generative model for generating a virtual try-on image of the user wearing the ornament based on the received user image and the desired ornament design. This allows users to try on an unlimited number of ornaments from the comfort of their own home and find their perfect outfit. The generated virtual try-on image can also be viewed from a 360-degree angle, making it easy for users to check the details of the desired ornament. Furthermore, by re-executing the generative model based on correction instructions and updating the virtual try-on image, users can adjust the ornament design to their satisfaction.

[1850] "User images" are photographic data taken by a user of himself or herself, taken from multiple angles.

[1851] A "desired accessory design" is a design of accessory or clothing that the user wishes to try on, and may take the form of image data or text instructions.

[1852] A "generative model" is an artificial intelligence model that generates virtual try-on images using a user's image and desired decorative design as input, and has image generation and editing functions.

[1853] A "virtual try-on image" is an image generated by a generative model that virtually recreates the appearance of a user wearing an accessory.

[1854] "Modification instructions" are requests for changes or adjustments made by the user to the virtual try-on image, and include specific changes to parts or styles.

[1855] "360-degree angle" means that the virtual try-on image can be viewed from all directions, allowing the user to check the decorations from multiple angles.

[1856] "Server" means a centralized computer system that receives, stores, and processes images and data from users.

[1857] This invention relates to a system that allows users to virtually try on accessories by taking pictures of themselves from multiple angles and inputting the desired accessory design. This system is implemented using a server, a terminal, and a generative AI model.

[1858] System configuration and operation

[1859] 1. User image upload

[1860] Users take pictures of themselves from the front, side, and back, and then access the system using a terminal. Users select the images they have taken and press the upload button to send them to the server.

[1861] The server stores the received user image data in cloud storage, confirms that the image has been uploaded correctly, and notifies the user.

[1862] Hardware used: smartphone, PC, cloud server

[1863] Software used: web browser, photo-taking application, cloud storage system (e.g., Amazon S3)

[1864] 2. Input your desired decorative design

[1865] The user inputs the desired decorative design as image data or text instructions. The terminal transmits this data to the server,

[1866] The server stores the received design data in cloud storage.

[1867] Hardware used: smartphone, PC, cloud server

[1868] Software used: Web forms, text input interface

[1869] 3. Generation of virtual try-on images

[1870] The server inputs the received user image and desired decorative design into a generative AI model.

[1871] The generative AI model uses this data to generate virtual try-on images of the user wearing the accessories, which can be viewed from a 360-degree angle.

[1872] The server provides this image to the user.

[1873] Hardware used: Cloud server

[1874] Software used: Generative AI models (e.g., DALL-E, Stable Diffusion), databases

[1875] 4. Check and edit the virtual try-on image

[1876] The user can check the generated virtual try-on image on their device. If they don't like a particular part of the decorative design, they can input correction instructions. These correction instructions are sent from the device to the server.

[1877] The server receives the correction instructions and feeds them back into the generative AI model, which then generates a modified virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[1878] Hardware used: smartphone, PC, cloud server

[1879] Software used: Generative AI models, web forms

[1880] 5. Final selection of decorative design

[1881] The user finally selects the decorative design they like and notifies the server via their terminal.

[1882] The server stores the information and confirms the final decision.

[1883] Hardware used: smartphone, PC, cloud server

[1884] Software used: Web forms, databases

[1885] Specific examples

[1886] If user A wants to choose a wedding dress, he can use the system as follows:

[1887] 1. User A takes pictures of himself from the front, side, and back and uploads them to the system.

[1888] Example prompt: "Please upload a photo of your front, side, and back."

[1889] 2. The server receives these images and stores them in cloud storage.

[1890] Example prompt: "Your image was successfully uploaded."

[1891] 3. User A takes a photo of a dress in a magazine with their smartphone and inputs it into the system.

[1892] Example prompt: "Please upload a photo of the dress you would like."

[1893] 4. The server receives this data and inputs it into a generative AI model.

[1894] 5. The generative AI model generates a virtual try-on image of User A wearing the dress and sends it to the server.

[1895] 6. The server provides the generated virtual try-on image to User A.

[1896] 7. User A checks the image on the terminal and inputs instructions to correct the hem length, for example.

[1897] Example prompt: "Please make the hem a little shorter."

[1898] 8. The server re-inputs the correction instructions into the generative AI model, generates an updated image, and provides it to User A.

[1899] 9. User A finally chooses the dress she likes and confirms it on the system.

[1900] 10. The server saves the selection and completes the process.

[1901] In this way, the system of the present invention allows users to try on an unlimited number of outfits from the comfort of their own home and find the perfect one.

[1902] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1903] Step 1: Take and upload a user image

[1904] Specific behavior:

[1905] Users take photos of themselves from the front, side, and back using a smartphone or camera.

[1906] Save the captured image file to a folder on your device.

[1907] Input: Captured image files (front, side, back)

[1908] Output: Image file saved on the device

[1909] Step 2: Access the system and upload your images

[1910] Specific behavior:

[1911] The user opens a web browser on the terminal, accesses the system's web application, and logs in.

[1912] Access the image upload form, select the image file you took, and press the upload button.

[1913] Input: Image files stored on the device, user login information

[1914] Output: Image file sent to the server

[1915] Step 3: The server receives and stores the image data.

[1916] Specific behavior:

[1917] The server saves the image data sent by the user in cloud storage and records the file path of the saved data in a database.

[1918] Verify that the image was uploaded correctly, generate a success message, and notify the user.

[1919] Input: Image data sent by the user

[1920] Output: Image data saved in cloud storage, success message

[1921] Step 4: User inputs desired decorative design

[1922] Specific behavior:

[1923] The user inputs image data and text instructions with the desired decorative design into the terminal, for example, by uploading an image of the dress taken with a smartphone.

[1924] Access the input form, select the design data, and press the submit button.

[1925] Input: Image data of decorative design taken with a smartphone or text instructions

[1926] Output: Decorative design data sent to the server

[1927] Step 5: The server receives and stores the design data

[1928] Specific behavior:

[1929] The server saves the decorative design data sent by the user in cloud storage, and records the file path and text content of the saved data in a database.

[1930] Input: Decoration design data submitted by the user

[1931] Output: Design data stored in cloud storage, file paths and text contents recorded in the database

[1932] Step 6: Start generating virtual try-on images

[1933] Specific behavior:

[1934] The server performs preprocessing to input the image data captured by the user and the desired decorative design data into the generative AI model.

[1935] The data is converted into the appropriate format and fed into a generative AI model.

[1936] Input: User image data stored in cloud storage, decorative design data stored in cloud storage

[1937] Output: The data that is fed into the generative AI model

[1938] Step 7: The generative AI model generates virtual try-on images

[1939] Specific behavior:

[1940] The generative AI model generates virtual try-on images based on the input user image data and decorative design data.

[1941] The generated try-on images are processed so that they can be displayed from multiple angles.

[1942] Input: User image data and decorative design data input into the generative AI model

[1943] Output: Generated virtual try-on image

[1944] Step 8: The server stores and serves the generated images.

[1945] Specific behavior:

[1946] The server saves the virtual try-on images sent from the generative AI model to cloud storage and records the file paths of the saved images in a database.

[1947] The user is notified of the URL of the generated virtual try-on image.

[1948] Input: Virtual try-on images sent from the generative AI model

[1949] Output: Virtual try-on images saved in cloud storage, notification to user

[1950] Step 9: The user checks the fitting image and inputs correction instructions.

[1951] Specific behavior:

[1952] The user can check the virtual fitting images provided on the device and input specific correction instructions as needed, such as "Please make the hem a little shorter."

[1953] Input: Provided virtual try-on image, modification instructions

[1954] Output: Correction instructions sent to the server

[1955] Step 10: The server re-inputs the correction instructions into the generative AI model.

[1956] Specific behavior:

[1957] The server performs preprocessing to re-input the received correction instructions into the generative AI model.

[1958] The correction instructions are input into the generative AI model, requesting the generation of an updated virtual try-on image.

[1959] Input: Correction instructions sent to the server

[1960] Output: Corrective instruction data that is input to the generative AI model

[1961] Step 11: The generative AI model generates the corrected image

[1962] Specific behavior:

[1963] The generative AI model generates new virtual try-on images based on the correction instructions and sends them to the server.

[1964] Input: Correction instructions input to the generative AI model

[1965] Output: Corrected virtual try-on image

[1966] Step 12: The server stores and serves the corrected image.

[1967] Specific behavior:

[1968] The server saves the newly generated virtual try-on image in cloud storage, records the file path of the saved image in the database, and notifies the user of the URL of the modified virtual try-on image.

[1969] Input: Modified virtual try-on image sent from the generative AI model

[1970] Output: Edited try-on image saved in cloud storage, notification to user

[1971] Step 13: User selects and notifies final design

[1972] Specific behavior:

[1973] The user finally checks the virtual try-on images to find one that satisfies him / her, selects the final decorative design, and notifies the server of the selection.

[1974] Input: Final selected virtual try-on image, selection details

[1975] Output: Final selection notification sent to the server

[1976] Step 14: The server saves the selection and ends the process

[1977] Specific behavior:

[1978] The server stores the information of the final selection notified by the user in the cloud storage and the database, and notifies the user that the final decision has been made.

[1979] Input: Final selection notification sent to the server

[1980] Output: Final selection information stored in cloud storage and database, completion notification to user

[1981] (Application example 1)

[1982] 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."

[1983] Conventional virtual try-on systems have the drawback of requiring a great deal of time and effort for users to check decorations, and the decorations often deviate from the actual image. The present invention aims to provide a system that allows users to virtually try on decorations in real time, and to quickly and effectively check and modify the decorations.

[1984] 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.

[1985] In this invention, the server includes: means for a user to take and upload photos of themselves from multiple angles; means for inputting desired ornaments; a generative model for generating a virtual try-on image of the user wearing the ornaments based on the received user photo and desired ornament design; means for the user to provide the generated virtual try-on image to the user; means for the user to issue correction instructions for the virtual try-on image; means for re-executing the generative model based on the correction instructions to update the virtual try-on image; means for confirming and saving the final selected ornament design; means for using a smartphone to check the generated virtual try-on image in real time; and means for the user to input the details of the desired ornament design as a prompt statement and for the system to control the generative model based on the prompt statement. This allows the user to check and correct the virtual try-on image of the ornaments in real time using their smartphone and quickly and efficiently select a final ornament design.

[1986] A "smartphone" is a type of mobile information terminal that has mobile phone functions and can connect to the Internet and use various applications.

[1987] A "virtual try-on image" is an image that simulates the appearance of wearing an accessory, created by a generative model based on the user's image and the desired accessory design.

[1988] A "generative model" is an algorithm or machine learning model that uses computer vision technology to generate virtual try-on images based on a received user photo and decorative design.

[1989] A "prompt sentence" is an input sentence that succinctly expresses the content of the decorative design desired by the user, and is important information for the system to control the generative model based on the instructions.

[1990] "Modification instructions" are instructions that the user inputs to make changes or adjustments to the virtual try-on image, and serve as a trigger for the generative model to be re-executed.

[1991] "Real-time" refers to the fact that the user can instantly check and correct the results of the ornament try-on, and means that the system operates with high response speed.

[1992] This invention relates to a system that allows users to virtually try on accessories by taking photos of themselves from multiple angles and inputting the desired accessory design. The invention is implemented using a server, a terminal, and a generative AI model.

[1993] System configuration and operation

[1994] 1. User image upload

[1995] Users take photos of themselves from various angles with their smartphones, access the system, select the photos they have taken, and press the upload button to send them to the server.

[1996] The server stores the received user photo data in storage, confirms that the photo has been uploaded correctly, and notifies the user.

[1997] 2. Input your desired decorative design

[1998] The user inputs the desired decorative design as a prompt, such as a simple text like "I want to try on a red dress." The device sends this prompt to the server, which then stores the received data in its storage.

[1999] 3. Generation of virtual try-on images

[2000] The server inputs the received user image and desired accessory design into a generative AI model. Based on this data, the generative AI model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[2001] 4. Check and edit the virtual try-on image

[2002] The user checks the generated virtual try-on image on their smartphone. If they don't like a particular part of the decorative design, they can input a prompt to make corrections. For example, they could give a specific instruction such as "Make the hem of the dress a little longer." The device then sends this correction instruction to the server.

[2003] The server receives the correction instructions and feeds them back into the generative AI model. The generative model generates a revised virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[2004] 5. Final dress selection

[2005] The user finally selects the decorative design they like and notifies the server via their device, which then stores the information and confirms the final decision.

[2006] Hardware and software used

[2007] The system utilizes the following hardware and software:

[2008] Hardware: Smartphones, servers, GPUs (deep learning-compatible GPUs such as NVIDIA)

[2009] Software: Python, FastAPI (web framework), Torch (deep learning library), PIL (Pillow, image processing library)

[2010] As a concrete example, if a user wants to try on a "red dress," he or she inputs the following prompt into the system:

[2011] Prompt: I want to try on a red dress

[2012] This allows the system of the present invention to allow users to try on an unlimited number of decorations from the convenience of their own home and find the perfect one.

[2013] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[2014] Step 1:

[2015] Users take photos of themselves from multiple angles and upload them via their devices. They use their smartphones to take photos of the front, side, back, etc., and press the upload button to send them to the system. The input is the user's image data, and the output is image data stored on the server.

[2016] Step 2:

[2017] The server saves the received user photo data in storage, verifies that the photo was uploaded correctly, and notifies the user. The server saves the image file in storage, verifies that it was saved successfully, and notifies the user of this confirmation information.

[2018] Step 3:

[2019] The user inputs the desired decorative design as a prompt sentence and sends it to the server via the terminal. The input includes text data such as "I would like to try on a red dress." The output is the text data saved on the server.

[2020] Step 4:

[2021] The server inputs the received user image and prompt text into the generative AI model. The user image (input image data) and prompt text (text data) are input into the generative model. Based on this data, the generative AI model generates a virtual try-on image. The output is the generated try-on image data.

[2022] Step 5:

[2023] The generative model is executed to generate a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. This image data is returned to the server and saved in the user's account. The output is a virtual try-on image saved in the user's account.

[2024] Step 6:

[2025] The user checks the generated virtual try-on image through the terminal. The generated image is displayed on the terminal, and the user can see the fully decorated try-on image of themselves. This includes both the input and output of the step.

[2026] Step 7:

[2027] If the user does not like a particular part of the decorative design, they can send a prompt to the server from their terminal with instructions to make corrections. For example, they can input an instruction such as "Make the hem of the dress a little longer." The input is the prompt to make corrections, and the output is correction instruction data stored on the server.

[2028] Step 8:

[2029] The server inputs the correction instructions into the generative AI model again, and the generative model regenerates the corrected virtual try-on image. The inputs are the user image, the existing virtual try-on image, and the correction prompt. The output is the corrected try-on image data.

[2030] Step 9:

[2031] The server provides the generated modified virtual try-on image to the user. The data is sent to the user's device, allowing the user to view the new try-on image. The output is the modified try-on image displayed by the user.

[2032] Step 10:

[2033] The user finally selects the decorative design they like and notifies the server via their terminal. The user presses the decision button to send the design to the server. The input at this time is the user's final decision data, and the output is the final selection information stored on the server.

[2034] 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.

[2035] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[2036] System configuration and operation

[2037] Image upload

[2038] Users take their own photos from various angles and access the system using a terminal. They select the photos they have taken and press the upload button to send them to the server.

[2039] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[2040] Decorative design input

[2041] The user inputs the desired decorative design as image data or text instructions, and the device sends this data to the server, which then stores the received data in storage.

[2042] Virtual try-on image generation

[2043] The server inputs the received user image and the desired accessory design into a generative model. Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[2044] Emotion recognition by emotion engine

[2045] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[2046] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[2047] Check and edit virtual try-on images

[2048] If the user needs to make any corrections to the generated virtual try-on image, the user inputs correction instructions, which are then sent from the terminal to the server.

[2049] The server receives the correction instructions and re-inputs them into the generative model. The generative model regenerates the virtual try-on image based on the new instructions and sends it to the server. This process is repeated until the user is satisfied.

[2050] Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a decoration, the emotion engine will suggest a different color.

[2051] Final Selection

[2052] The user finally selects the decorative design they like on the terminal.

[2053] The terminal transmits the final selected decorative design information to the server.

[2054] The server determines the selected decorative design and stores the associated data in storage.

[2055] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[2056] Specific examples

[2057] For example, if user A wants to choose a kimono for his / her coming-of-age ceremony, he / she can use the system as follows:

[2058] 1. User A takes photos of himself from the front, side, and back and uploads them to the system.

[2059] 2. The server receives these photos and stores them in storage.

[2060] 3. User A takes a picture of a kimono in the catalog with their smartphone and inputs it into the system.

[2061] 4. The server receives this data and inputs it into the generative model.

[2062] 5. The generative model generates a virtual try-on image of User A wearing the kimono and sends it to the server.

[2063] 6. The server provides the generated virtual try-on image to User A.

[2064] 7. When User A checks the try-on images on the device, he / she sends his / her reaction to the emotion engine via the camera and microphone.

[2065] 8. The emotion engine analyzes User A's facial expressions and voice and reports the current emotion to the server.

[2066] 9. If user A is dissatisfied with a particular part of the try-on image, he or she inputs correction instructions.

[2067] 10. The server re-inputs the correction instructions into the generative model, generates an updated image, and serves it to User A. This process is repeated based on feedback from the emotion engine.

[2068] 11. User A finally chooses the decorative design he likes and confirms it on the system.

[2069] 12. The server saves the selection and completes the process.

[2070] In this way, the system of the present invention allows users to try on an infinite number of outfits from the comfort of their own home and find their perfect outfit through emotion recognition.

[2071] The processing flow will be explained below.

[2072] Specific steps in the system programming process

[2073] Image upload

[2074] Step 1:

[2075] Users take photos of themselves from various angles and access the system using a terminal.

[2076] Step 2:

[2077] The device selects multiple photos taken and sends them to the server by pressing the upload button.

[2078] Step 3:

[2079] The server stores the received user photo data in storage and verifies that the photo has been uploaded correctly.

[2080] Decorative design input

[2081] Step 4:

[2082] The user inputs the desired decorative design as image data or text instructions.

[2083] Step 5:

[2084] The terminal transmits the input decorative design data to the server.

[2085] Step 6:

[2086] The server stores the received decorative design data in storage and notifies the user that the input is complete.

[2087] Virtual try-on image generation

[2088] Step 7:

[2089] The server inputs the received user image and desired decorative design into the generative model.

[2090] Step 8:

[2091] The generative model generates a virtual try-on image of the user wearing the accessory based on the user's photo and the input accessory design.

[2092] Step 9:

[2093] The generative model creates virtual try-on images that can be viewed from a 360-degree angle and sends them to the server.

[2094] Step 10:

[2095] The server receives the generated virtual try-on images and stores them in the user's account.

[2096] Emotion recognition by emotion engine

[2097] Step 11:

[2098] The user checks the virtual fitting images and sends their reactions to the server via the device, which is equipped with a camera and microphone to capture the user's facial expressions and voice in real time.

[2099] Step 12:

[2100] The server uses an emotion engine to analyze the user's facial expressions and voice to identify their emotions, and evaluates the virtual try-on images based on this information.

[2101] Check and edit virtual try-on images

[2102] Step 13:

[2103] If the user needs to make any corrections to the generated virtual try-on image, they can input specific corrections, such as "change the hem length."

[2104] Step 14:

[2105] The terminal sends a correction instruction to the server.

[2106] Step 15:

[2107] The server re-inputs the received correction instructions into the generative model.

[2108] Step 16:

[2109] The generative model regenerates the virtual try-on images based on the new instructions and sends them to the server.

[2110] Step 17:

[2111] The server receives the updated virtual try-on images and stores them in the user's account.

[2112] Step 18:

[2113] The user reviews the image again and makes further corrections if necessary, and the process is repeated until the user is satisfied.

[2114] Step 19:

[2115] The emotion engine monitors the user's emotions, identifies areas of dissatisfaction, and provides that information to the server.

[2116] Step 20:

[2117] Based on the information from the emotion engine, the server suggests appropriate decorative designs and modifications to the user.

[2118] Final Selection

[2119] Step 21:

[2120] The user finally selects the decorative design they like on the terminal.

[2121] Step 22:

[2122] The terminal transmits the final selected decorative design information to the server.

[2123] Step 23:

[2124] The server determines the selected decorative design and stores the associated data in storage.

[2125] Step 24:

[2126] The server will then send the user a confirmation of their final decision, informing them that their decoration selection is complete.

[2127] Example 2

[2128] 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."

[2129] Conventional virtual try-on systems have a limited process for generating virtual try-on images based on user-provided images and decorative designs, making it difficult to effectively incorporate user emotions and feedback. As a result, users are often dissatisfied with their final selection and are forced to go through multiple trial and error rounds. Furthermore, the quality of try-on images and real-time adjustments are insufficient, leaving room for improvement in the user experience.

[2130] 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.

[2131] In this invention, the server includes a means for users to take and upload images of themselves from multiple angles, a means for inputting requested accessories, and an image generation model for generating virtual try-on images of the user wearing the accessories based on the received user images and the requested accessory design. This allows users to generate high-quality virtual try-on images in real time and reflect feedback using an emotion recognition engine.

[2132] A "user" is someone who uses the system to upload a photo of themselves and try on clothes virtually.

[2133] "Terminal" means an electronic device used by a user to access the system and input photos and decorative designs. Examples include smartphones, tablets, and PCs.

[2134] The "server" is a central processing unit that processes data received from users and operates generative AI models and emotion recognition engines.

[2135] A "generative model" is an algorithm or software that generates virtual try-on images based on a received user image and requested accessory design, often using deep learning techniques.

[2136] A "virtual try-on image" is a virtual image generated based on a user's image, showing an outfit with a decorative design specified by the user.

[2137] An "emotion recognition engine" is an algorithm or software that analyzes a user's facial expressions, voice, etc. to identify the user's emotions.

[2138] "Modification instructions" refer to requests for changes made by the user to the virtual try-on image, and specifically include changes to the color or shape of the decorative design.

[2139] "Decorative design" is data that represents the appearance of the clothes, accessories, etc. that the user wishes to try on in the virtual try-on image.

[2140] "Reaction" refers to the facial and vocal feedback given by the user when they view the virtual try-on image.

[2141] "Suggestions" are suggestions for alternative decorative designs to improve user satisfaction, generated by the emotion recognition engine based on the user's reactions.

[2142] These definitions will provide a clear understanding of each claim element.

[2143] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative design. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. Specifically, the system of the present invention is implemented using a server, a terminal, an image generation model, and an emotion recognition engine.

[2144] System configuration

[2145] 1. A device where users take and upload their own images

[2146] Users use devices such as smartphones, tablets, and computers to take and upload images of themselves from various angles.

[2147] The device is equipped with a camera, and the user uses a dedicated application to select and upload images to the system.

[2148] 2. Server

[2149] The server processes the image data received from the user and stores it in storage.

[2150] The server also receives and manages decorative design data sent by users.

[2151] Furthermore, the server operates an image generation model based on the received data to generate a virtual try-on image.

[2152] The server then runs the image generation model again based on the generated virtual try-on images and user feedback.

[2153] 3. Image Generation Model

[2154] The image generation model takes the user's image and decorative design data as input and generates a virtual try-on image, primarily using deep learning technology.

[2155] Virtual try-on images are generated in a format that can be viewed from a 360-degree angle.

[2156] 4. Emotion Recognition Engine

[2157] The emotion recognition engine captures the user's facial expressions and voice through the device's camera and microphone and analyzes the user's emotions in real time.

[2158] Based on the results of this analysis, the emotion recognition engine suggests appropriate decorative designs.

[2159] Specific examples

[2160] Below is a concrete example of how User A uses the system to choose a kimono for her coming-of-age ceremony:

[2161] 1. User A uses a smartphone to take pictures of himself from three angles (front, side, and back) and uploads them to the system via a dedicated application.

[2162] 2. The server receives the image data and stores it in storage.

[2163] 3. User A takes a picture of a kimono from a coming-of-age ceremony catalogue with his smartphone and inputs it into the system as decorative design data.

[2164] 4. The server receives this decorative design data and inputs it into the image generation model.

[2165] 5. The image generation model generates a virtual try-on image of User A wearing the kimono for his coming-of-age ceremony and sends it to the server.

[2166] 6. The server provides the generated virtual try-on image to User A.

[2167] 7. User A checks the virtual try-on image on their device and sends their reaction to the emotion engine via the camera and microphone.

[2168] 8. The emotion engine analyzes User A's facial expressions and voice to determine which designs User A is satisfied or dissatisfied with.

[2169] 9. If user A is dissatisfied with a particular part, he or she enters correction instructions into the terminal and sends them to the server.

[2170] 10. The server receives the correction instructions and again instructs the image generation model to generate updated virtual try-on images.

[2171] 11. The server provides the updated image to User A, and the process repeats until User A is satisfied.

[2172] In this way, the system of the present invention allows users to try on an infinite number of decorations from the comfort of their own home and find their perfect design through emotion recognition.

[2173] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2174] Step 1:

[2175] Users take multiple photos of themselves from various angles, such as from the front, side, and back, select the images through a dedicated application, and press the upload button.

[2176] Input: Images taken by the user from multiple angles

[2177] Output: User's image data

[2178] Step 2:

[2179] The terminal sends the user's image data to the server, where the image data is transferred over the network.

[2180] Input: User's image data

[2181] Output: Image data sent to the server

[2182] Step 3:

[2183] The server stores the received image data of the user in storage, and also generates a message confirming receipt and notifies the user.

[2184] Input: Image data sent from the device

[2185] Output: Image data saved in storage, receipt confirmation message

[2186] Step 4:

[2187] The user inputs the desired decorative design into the terminal as image data or text instructions. For example, the user may take a photo of a kimono for their coming-of-age ceremony and upload it.

[2188] Input: User-provided image data or text instructions for the decorative design

[2189] Output: decorative design data

[2190] Step 5:

[2191] The terminal transmits the input decorative design data to the server, where data transfer also takes place via the network.

[2192] Input: decorative design data

[2193] Output: Decorative design data sent to the server

[2194] Step 6:

[2195] The server stores the received decorative design data in storage, and also generates a confirmation message and notifies the user.

[2196] Input: Decoration design data sent from the device

[2197] Output: Decorative design data saved in storage, confirmation message

[2198] Step 7:

[2199] The server inputs the received user image data and decorative design data into an image generation model, specifically, a deep learning algorithm to generate virtual try-on images.

[2200] Input: User image data, decorative design data

[2201] Output: Generated virtual try-on image

[2202] Step 8:

[2203] The server links the generated virtual try-on image to the user's account and provides it to the user, who can view it on their device.

[2204] Input: Generated virtual try-on images

[2205] Output: A link to the virtual try-on image provided to the user

[2206] Step 9:

[2207] The user checks the virtual fitting images on the device and sends their reactions to the emotion engine via the camera and microphone. The device captures facial expressions and voice.

[2208] Input: Virtual try-on image for user to view

[2209] Output: Captured user reaction data

[2210] Step 10:

[2211] The server analyzes the captured user reaction data using an emotion recognition engine, specifically facial detection and voice tone analysis, to identify user satisfaction or dissatisfaction.

[2212] Input: User response data

[2213] Output: User sentiment analysis results

[2214] Step 11:

[2215] If the user needs to make any modifications to the virtual try-on image, the user can input and send the modification instructions to the terminal, for example, by inputting specific instructions such as "change the color to blue."

[2216] Input: Correction instructions

[2217] Output: Correction instructions typed into the terminal

[2218] Step 12:

[2219] The terminal sends the correction instruction to the server, and data is transferred via the network.

[2220] Input: Correction instructions

[2221] output: Correction instructions sent to the server

[2222] Step 13:

[2223] The server receives the correction instructions and re-inputs them into the generative AI model, which then generates updated virtual try-on images based on the new instructions.

[2224] Input: Correction instructions

[2225] Output: Updated virtual try-on image

[2226] Step 14:

[2227] The server saves the updated virtual try-on image to the user's account and provides it to the user again via the link, and this process is repeated until the user is satisfied.

[2228] Input: Updated virtual try-on image

[2229] output: Updated virtual try-on image link provided to the user

[2230] Step 15:

[2231] The user finally makes a decision to select the decorative design they like, and presses the confirmation button to confirm the selection through the terminal.

[2232] Input: Final decorative design selection

[2233] Output: Selection information

[2234] Step 16:

[2235] The terminal transmits the final selection information to the server, and data transfer occurs via the network.

[2236] Input: Final selection information

[2237] Output: Final selection information sent to the server

[2238] Step 17:

[2239] The server confirms the selected decoration design, stores the related data in storage, and sends a final decision confirmation to the user, informing them that the decoration selection is complete.

[2240] Input: Final selection information

[2241] Output: Selected design data saved in storage, confirmation notification to user

[2242] (Application example 2)

[2243] 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."

[2244] In conventional virtual try-on systems, users had to rely solely on visual confirmation when checking try-on images. This made it difficult to reflect the user's emotions and intuitive reactions, making it difficult for them to select decorative designs that truly satisfied the user. Furthermore, if the user was dissatisfied with the try-on images, suggestions for revisions could not be made efficiently, resulting in a problem of a decline in the quality of the try-on experience. There is a need to solve these problems.

[2245] 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.

[2246] In this invention, the server includes means for a user to take and upload photos of themselves from multiple angles, means for inputting desired decorations, a generative model for generating a virtual try-on image of the user wearing the decorations based on the received user photo and desired decoration design, means for providing the generated virtual try-on image to the user, means for the user to issue correction instructions for the virtual try-on image, means for re-executing the generative model based on the correction instructions to update the virtual try-on image, means for finalizing and saving the selected decoration design, means for capturing the user's facial expressions and voice, recognizing and analyzing emotions, and means for presenting correction suggestions to the user based on the emotion recognition results.

[2247] This allows the system to reflect the user's emotions in real time, efficiently suggest modifications based on intuitive feedback, and select the decorative design that will satisfy the user most.

[2248] "User" refers to a person who uses this system to upload their own photos and try on clothes virtually.

[2249] "Photos" refer to image data taken by a user from multiple angles.

[2250] "Decorations" refer to design elements such as clothing and accessories that the user wishes to try on.

[2251] "Input" refers to the user inputting the desired decorative design into the system.

[2252] "Generative model" refers to a machine learning model or algorithm that generates virtual try-on images based on the user's received photo and desired decorative design.

[2253] A "virtual try-on image" refers to an image that virtually depicts the state in which a user is wearing the accessory.

[2254] "Modification instructions" refer to instructions for changes or modifications that the user makes to the generated virtual try-on image.

[2255] "Emotion recognition" refers to the process of analyzing a user's facial expressions and voice in real time to identify their emotions.

[2256] "Proposed revision" refers to proposed changes to the decorative design that are suggested to the user based on the emotion recognition results.

[2257] "Server" refers to the central system that receives and stores image and design data from users, generates virtual try-on images using generative models, and performs emotion recognition.

[2258] "Storage" refers to the data storage location where the server stores users' photos, decorative design data, and virtual try-on images.

[2259] The present invention is a system that allows users to virtually try on clothes by taking photos of themselves from multiple angles and inputting their desired decorative designs. Furthermore, this system also incorporates an emotion engine that recognizes the user's emotions. This system is implemented using a server, a terminal, a generative model, and an emotion engine.

[2260] System configuration and operation

[2261] Image upload

[2262] Users access the system by taking photos of themselves from various angles using smart glasses or a smartphone. The user selects the photos and sends them to the server by pressing the upload button. The server stores the received user photo data in storage (e.g., AWS S3) and verifies that the photos have been uploaded correctly.

[2263] Decorative design input

[2264] The user inputs the desired decorative design as image data or text instructions. The device sends this data to the server, which then stores it in storage (e.g., AWS S3).

[2265] Virtual try-on image generation

[2266] The server inputs the received user image and desired accessory design into a generative model (e.g., StyleGAN). Based on this data, the generative model generates a virtual try-on image of the user wearing the accessory. The generated virtual try-on image is created so that it can be viewed from a 360-degree angle. The server saves this image in the user's account and provides it to the user.

[2267] Emotion recognition by emotion engine

[2268] The user reviews the virtual try-on images and sends their reactions to the server via their device. The device is equipped with a camera and microphone, which capture the user's facial expressions and voice in real time. The server then uses an emotion engine (e.g., Microsoft Azure Emotion API) to analyze the user's facial expressions and voice and identify their emotions. This information is used to evaluate the virtual try-on images and determine which aspects the user is satisfied or dissatisfied with.

[2269] Check and edit virtual try-on images

[2270] If the generated virtual try-on image requires any modifications, the user inputs the modifications. These modifications are sent from the device to the server. The server receives the modifications and inputs them back into the generative model. The generative model then regenerates the virtual try-on image based on the new modifications and sends it to the server. This process is repeated until the user is satisfied. Between steps, the emotion engine constantly monitors the user's emotions and suggests appropriate modifications if the user is dissatisfied. For example, if the user expresses dissatisfaction with the color of a certain decoration, the emotion engine will suggest a different color.

[2271] Final Selection

[2272] The user finally selects their favorite decoration design on the device. The device sends the final selected decoration design information to the server. The server confirms the selected decoration design and saves the related data in storage. The server then sends the user a confirmation of the final decision, notifying them that decoration selection is complete.

[2273] Specific examples

[2274] For example, if user B wants to choose a dress at a brand shop, he can use the system as follows.

[2275] 1. User B uses smart glasses in a store to take photos of himself (front, side, and back).

[2276] 2. The server receives and stores the photo.

[2277] 3. User B takes a picture of the dress they want with their smartphone and enters it into the app.

[2278] 4. The server receives this data and inputs it into a generative model (StyleGAN).

[2279] 5. The generative AI model (StyleGAN) generates a virtual try-on image based on "User B's photo + dress image" and sends it to the server.

[2280] 6. The server provides the generated try-on image to User B.

[2281] 7. User B checks the fitting image. The camera in the smart glasses captures their facial expressions and sends them to the emotion engine.

[2282] 8. The emotion engine (Microsoft Azure Emotion API) analyzes facial expressions and initiates suggestions for corrections to areas of dissatisfaction (e.g., "I don't like the color of the red dress").

[2283] 9. User B inputs color correction instructions. The server again uses the generative model to generate an updated image and provides it to User B.

[2284] 10. User B finally chooses the dress they like and confirms it in the app.

[2285] 11. The server saves your selection and completes the process.

[2286] Example prompts for generative AI models

[2287] "Combine the front, side, and back images of the user to create a virtual image of them trying on a specified decorative design (a red dress)."

[2288] In this way, the system of the present invention provides a support system that allows users to check and select their ideal outfit without trying it on in the store.

[2289] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2290] Step 1:

[2291] Image upload

[2292] Users can use smart glasses or smartphones to take photos of themselves from the front, side, or back, then access the application, select the photo they have taken, and press the upload button to send the photo to the server.

[2293] Input: Image data taken from multiple angles by the user

[2294] The server saves the received image data in storage (e.g. AWS S3) and confirms that it has been uploaded successfully.

[2295] Output: Reference information for image data stored in storage

[2296] Step 2:

[2297] Decorative design input

[2298] The user inputs the desired decorative design into the application as image data or text instructions, which are then sent to the server.

[2299] Input: Image data or text instructions for the decorative design entered by the user

[2300] The server stores the received design data in storage (e.g. AWS S3).

[2301] Output: Reference information for decorative designs stored in storage

[2302] Step 3:

[2303] Virtual try-on image generation

[2304] The server inputs the user's photo data and desired decorative design data into a generative model (e.g., StyleGAN).

[2305] Input: User photo data and decorative design data

[2306] Based on this data, the generative model generates a virtual try-on image of the user wearing the accessories, which can be viewed from a 360-degree angle.

[2307] The server stores the generated virtual try-on image in the user's account and provides it to the user.

[2308] Output: Virtual try-on image data

[2309] Step 4:

[2310] Emotion recognition by emotion engine

[2311] The user checks the provided virtual try-on images, and the device's built-in camera and microphone capture the user's facial expressions and voice in real time as they check.

[2312] Input: User's facial expression data and voice data

[2313] The server analyzes the received facial expression and voice data through an emotion engine (e.g., Microsoft Azure Emotion API) to identify the user's emotions.

[2314] Output: Emotion recognition result data

[2315] Step 5:

[2316] Check and edit virtual try-on images

[2317] If the user needs to make any corrections to the virtual try-on image, the user inputs the correction instructions into the application and sends them to the server.

[2318] Input: User-entered correction instructions

[2319] The server receives and analyzes the correction instructions, and then inputs them back into the generative model to generate a corrected virtual try-on image. The generative model then regenerates the virtual try-on image based on the new instructions.

[2320] The server provides the user with new virtual try-on images, and this process is repeated until the user is satisfied.

[2321] Output: Data of the virtual try-on image after correction

[2322] Step 6:

[2323] Final Selection

[2324] The user finally selects the decorative design they like on the terminal.

[2325] Input: Information on the decorative design that the user finally selected

[2326] The terminal transmits the selected decorative design information to the server.

[2327] The server confirms the selected decorative design, stores the relevant data in storage, and sends the user a confirmation of the final decision.

[2328] Output: Final information of the saved decorative design

[2329] This specific processing step allows users to virtually try on clothes and efficiently select the most suitable decorative design through emotion recognition.

[2330] 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.

[2331] 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.

[2332] 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.

[2333] 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.

[2334] 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.

[2335] 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.

[2336] 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).

[2337] 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.

[2338] 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."

[2339] 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.

[2340] 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).

[2341] 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.

[2342] 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.

[2343] 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.

[2344] 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.

[2345] 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....

Claims

1. a means for users to take and upload photos of themselves from multiple angles; A way to input the desired decorations, a generative model for generating a virtual try-on image of the user wearing the ornament based on the received user's photo and desired ornament design; A means for providing the generated virtual try-on image to a user; A means for a user to issue correction instructions to the virtual try-on image; A means for re-executing the generative model based on the correction instruction to update the virtual try-on image; a means for finalizing and saving the selected decorative design; A system including:

2. 2. The system of claim 1, wherein the means by which the user inputs the desired decorative design is by image data or text instructions.

3. The system of claim 1 , wherein the generated virtual try-on images are provided to the user from a 360-degree angle.

Images