Automated method for selecting and requesting retro product designs

Abstract

An automated method for selecting and requesting a retro product design is provided. The equipment according to one embodiment can collect retro product pictures from an online shopping mall related to retro and an offline shopping mall related to retro, select a retro product to be produced by using the collected retro product pictures, and send the produced retro product to a terminal of a designer for production.

Description

Technical Field

[0001] The following examples relate to automated methods for selecting vintage products and design commissions. Background Technology

[0002] Vintage clothing is popular among many consumers due to its unique historical value and original designs. Especially in the fashion world, vintage clothing possesses a special charm and is highly sought after by those who appreciate original style. However, the process of selecting and designing vintage clothing is both time-consuming and laborious, and it's difficult to do it effectively without the help of professionals. Therefore, a system is needed that can easily select vintage clothing and automate design requests.

[0003] There are several key challenges in selecting vintage clothing and commissioning design services. First, choosing vintage clothing is time-consuming, as it involves sifting through a vast array of options to find pieces with unique historical value and quality. Furthermore, accurately evaluating and selecting vintage clothing requires considerable expertise. The average consumer lacks this knowledge, making it difficult to choose suitable pieces. After selecting vintage clothing, designing it is a daunting task—adding a modern touch while maintaining the garment's uniqueness. Maintaining this consistency throughout the design process is not easy, and poor communication with the designer often results in outcomes that fall short of expectations.

[0004] Therefore, it is necessary to develop automated methods for selecting retro products and requesting designs.

[0005] Existing technical documents

[0006] Patent documents

[0007] (Patent Document 1) Korean Patent Registration No. 10-2458694 (Announcement 2022.10.25)

[0008] (Patent Document 2) Korean Patent Registration No. 10-2597643 (Announcement 2023.11.02)

[0009] (Patent Document 3) Korean Patent No. 10-2018-0078845 (Published on July 10, 2018)

[0010] (Patent Document 4) Korean Patent No. 10-2020-0001766 (Published on January 7, 2020) Summary of the Invention

[0011] The problem that the invention aims to solve

[0012] The implementation plan aims to save time and effort through automated systems, making it easier for non-professionals to select vintage clothing and to request consistent designs.

[0013] The implementation plan aims to enhance collaboration with designers through effective communication tools.

[0014] The embodiments are designed to save consumers time and effort by providing a system for quickly and efficiently selecting vintage clothing.

[0015] means for solving problems

[0016] According to one embodiment, in an automated method for selecting and designing antique products, the process is performed by a device to collect images of antique products from online and offline antique-related marketplaces; select antique products for production using the collected images; and transmit the produced antique products to a designer's terminal and request their production; it can be used to include...

[0017] The steps for collecting vintage product images from vintage-related online stores include: checking product pages where users spend more time than a preset standard, and identifying and collecting images from product pages where users confirm that the exposure time exceeds the preset standard, as vintage product images; and collecting vintage product images from vintage-related offline stores, including identifying and collecting products whose exposure time exceeds a preset standard from videos captured by users' cameras and vintage product images. The steps for selecting vintage products to create using collected vintage product images include: checking the similarity of collected vintage product images using an image similarity analysis algorithm; selecting product images with similarity lower than the preset standard as unusual vintage product images; searching for unusual vintage product images using an image search service; selecting unusual vintage product images using an image similarity analysis algorithm; and selecting product images with similarity lower than the preset standard. This may include selecting products corresponding to the aforementioned specific vintage product images as the basis for producing vintage products.

[0018] The steps for selecting vintage products to be made using collected vintage product images include checking the uploaded product images to the user's social network account, using image similarity analysis algorithms to check the similarity between the collected vintage product images and the uploaded product images, and identifying the images to be checked from the collected vintage product images with similarity higher than a preset standard. This may include uploading the images required for verification to the social network account, confirming the preferences of the uploaded posts, and selecting products that correspond to the images posted in the posts and whose preferences are higher than the preset standard as the vintage products to be made.

[0019] The stage of transmitting the produced retro products to the designer's terminal and requesting production involves analyzing the produced retro products, generating data such as material information, production methods, expected production time, and estimated production costs, checking whether the designer can produce based on the data, and if the designer cannot produce the retro products, transmitting the data to the designer's terminal that can produce the retro products. If the retro products can be produced, this may include transmitting the intention to produce to the organizer's terminal, transmitting the estimated production costs, including material costs, labor costs, and other ancillary expenses, to the organizer's terminal, and the organizer confirming the estimated production costs and requesting confirmation from the organizer's terminal.

[0020] The steps for selecting and producing vintage products using collected vintage product images include further extracting product components from the collected vintage product images, including tearing, color, material, and size; checking common components from multiple product images; selecting the most common configuration from the identified common components; generating an AI image based on the selected vintage product components; and selecting the generated image as the vintage product to be produced.

[0021] The device according to one embodiment can be controlled by combining a computer program stored on a medium with hardware to perform any of the methods described above.

[0022] Invention Effects

[0023] The implementation can save time and effort through automated systems, make it easier for non-professionals to select vintage clothing, and achieve consistent design commissions.

[0024] Implementation examples can enhance collaboration with designers through effective communication tools.

[0025] The embodiments provide a system for quickly and efficiently selecting vintage clothing, saving consumers time and effort. Attached Figure Description

[0026] Figure 1 It is a diagram used to illustrate the system configuration according to a single embodiment.

[0027] Figure 2 This is a flowchart illustrating an automated method for selecting retro products and requesting designs based on common implementation examples.

[0028] Figure 3 It is a flowchart illustrating the process of selecting and producing antique products in online and offline shopping malls related to antiques, based on a routine implementation.

[0029] Figure 4This is a flowchart illustrating the process of selecting retro products produced via social networks, based on an embodiment.

[0030] Figure 5 It is a flowchart illustrating the process of requesting designers to proceed with production based on a single embodiment.

[0031] Figure 6 It is a flowchart illustrating the process of extracting product ingredients and selecting products from the production year according to an example.

[0032] Figure 7 This is a preliminary schematic diagram of the structure of the device according to an embodiment.

[0033] Explanation of reference numerals in the attached figures

[0034] 100: Terminal; 400: Device

[0035] 401: Processor; 402: Memory Detailed Implementation

[0036] The embodiments are described in detail below with reference to the accompanying drawings. However, various modifications can be made to the embodiments, and therefore the scope of the patent application is not limited to or restricted by these embodiments. Any alterations, equivalents, or substitutions to the implementation methods should be understood to be included within the scope of the claims.

[0037] The specific structural or functional descriptions of the embodiments are for illustrative purposes only and can be modified and implemented in various forms. Therefore, the embodiments are not limited to the specific form of disclosure, and the scope of this specification includes changes, uniformities, or substitutions incorporated into the descriptive ideas.

[0038] Terms such as "first" or "second" can be used to describe various components, but the interpretation of these terms should only be used to distinguish one component from another. For example, a first component can be named a second component, and similarly, a second component can be named a first component.

[0039] When a component is said to be "connected" to another component, it should be understood that it may be directly connected to another component or connected to that other component, but there may be another component.

[0040] The terminology used in the embodiments is for illustrative purposes only and should not be construed as limiting. Singular expressions include plural expressions unless the context clearly implies otherwise. In this specification, the terms "comprising" or "having" should be understood to mean the presence of the features, numbers, steps, actions, components, parts, or combinations thereof described herein, and should not exclude the presence or addition of one or more other features or numbers, steps, actions, components, parts, or combinations thereof.

[0041] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments pertain. Terms as defined in common dictionaries should be interpreted as having the same meaning as they have in the relevant description and should not be interpreted in an idealistic or overly formal sense unless expressly defined in this application.

[0042] Furthermore, when describing the accompanying drawings, regardless of the drawing codes, the same reference numerals should be assigned to the same components, and identical repetitive descriptions should be omitted. When describing embodiments, detailed descriptions should be omitted if it is determined that a specific description of the relevant notifying technology might unnecessarily obscure the essence of the embodiment.

[0043] The embodiments can be implemented in a variety of products, including personal computers, laptops, tablets, smartphones, televisions, smart home appliances, smart cars, self-service terminals, and wearable devices.

[0044] Figure 1 It is a diagram used to illustrate the system configuration according to a single embodiment.

[0045] See Figure 1 The system according to the embodiment may include a user terminal 100 and a device 400, and the system according to the embodiment may communicate with each other through a communication network.

[0046] First, regardless of the communication method, such as wired or wireless, a communication network can be configured and implemented in various forms to enable communication between servers and between servers and terminals.

[0047] The user's terminal 100 can be the terminal from which the user requests the production of retro products. The user's terminal 100 can be a desktop computer, laptop computer, tablet computer, smartphone, etc. For example, ... Figure 1 As shown, the user's terminal 100 can be a smartphone, and can be implemented in different ways depending on the specific embodiment.

[0048] The user's terminal 100 can be configured to perform all or part of the computing, storage / reference, input / output, and control functions of a regular computer. The user's terminal 100 can be configured to communicate with the device 400 via wired or wireless communication.

[0049] The user's terminal 100 can access web pages operated by individuals or organizations using services provided by device 400, or can install applications developed and distributed by individuals or organizations using services provided by device 400. The user's device 100 can link to device 400 via web pages or applications.

[0050] The user's terminal 100 can access device 400 through a webpage or application provided by device 400.

[0051] The singular form in a claim can be understood to include the plural form. For example, the user in a statement can refer to one or more users.

[0052] Device 400 can provide a platform service that offers educational content for non-face-to-face programming training to businesses.

[0053] Device 400 may be its own server, owned by an individual or organization using services provided by device 400, or it may be a cloud server or a peer-to-peer (P2P) distributed set of nodes. Device 400 may be configured to perform all or part of the computing, storage / reference, input / output, and control functions of a conventional computer.

[0054] The device 400 can be configured to communicate with the user's terminal 100 via wired or wireless communication, and can control the operation of the user's terminal 100 and control what information is displayed on the screen of the user's terminal 100.

[0055] On the other hand, for ease of explanation, Figure 1 Only user terminal 100 is shown, but the number of terminals may vary depending on the embodiment. There is no particular limitation on the number of terminals and printers in terms of the processing power of device 400.

[0056] In one embodiment, the database may be located within device 400, but is not limited to this; the database may be configured independently of device 400. Device 400 may contain multiple artificial neural networks for performing machine learning algorithms.

[0057] Figure 2 This is a flowchart illustrating an automated method for selecting retro products and requesting designs based on common implementation examples.

[0058] See Figure 2 First, in step S201, the device 400 is able to collect images of antique products from online stores related to antiques.

[0059] Retro-themed online stores can be online shopping malls that gain exposure through searches for retro-related keywords. Users can select desired products and perform actions on the product pages. Users can view product pages where they spend more time than a preset threshold. Based on the assumption that products attract longer user attention are more likely to be retro, images on product pages that exceed the user's viewing time threshold can be identified as retro product images. This process effectively collects retro product images that users are highly interested in.

[0060] At this time, below Figure 3 The document describes in detail the process of checking product pages where users spend more than a preset time.

[0061] In addition, during phase S202, device 400 is able to collect images of antique products from offline stores related to antiques.

[0062] Retro product images can be collected by analyzing user behavior in offline environments. By analyzing images captured by users' cameras, products that have been exposed for more than a preset reference time can be identified and collected, thereby identifying and collecting retro product images.

[0063] Currently, the following Figure 3 The process of losing products exposed for more than a preset reference time will be explained in detail, and described using images of vintage products.

[0064] In step S203, the device 400 can use the collected year product images to select the year of manufacture of the product.

[0065] First, image similarity analysis algorithms can be used to check the similarity of collected product images from different years, selecting images with similarity scores below a preset standard as unique product images for that year. Then, an image search service can be used to search for unusual vintage product images. If the number of images found is below a preset standard, these unusual vintage product images can be selected for vintage product production.

[0066] In addition, the system can examine product images uploaded to users' social network accounts. It can check the similarity between collected retro product images and those uploaded using image similarity analysis algorithms. Images with similarity scores higher than a preset standard can be extracted and identified as images requiring verification. These images are then uploaded to the social network account, and preferences are checked through posts. Products corresponding to the images posted in the posts, if confirmed to have preferences higher than the preset standard, can be selected for retro product production.

[0067] In addition, it can extract product ingredients from collected vintage product images, check common ingredients from product images, select the most common ingredients from the identified common ingredients as vintage product ingredients, generate AI images based on the ingredients, and select the generated images as products from the production year.

[0068] A detailed explanation of the process for extracting product ingredients from collected product images from different years will be provided later. Figure 6 Describe it.

[0069] In the S204 stage, equipment 400 can transfer the manufactured retro products to the design terminal for production.

[0070] First, by analyzing the selected retro products, data can be generated from material information, production methods, expected production time and estimated production costs. Based on the generated data, the feasibility of production by the designer can be confirmed.

[0071] Furthermore, if the designer is unable to produce it, they can request it by transmitting data to the terminal of the designer who can produce it. If production is possible, the designer receiving the production request can convey the intention to produce it to the customer's terminal, along with the estimated production cost. The estimated production cost includes, but is not limited to, materials, labor, and other ancillary costs.

[0072] In addition, the client confirms the estimated production cost, and once the client's end confirms the request, the designer proceeds with production.

[0073] This invention automatically selects vintage products and automates the design request process, thus providing convenience and saving users time and money. In particular, by analyzing images collected from various sources and selecting products based on similarity and preferences, vintage products can be produced more accurately and efficiently. Furthermore, the image generation and analysis methods using artificial intelligence technology can significantly contribute to simplifying the production process while maintaining the originality and uniqueness of vintage products.

[0074] Figure 3 It is a flowchart illustrating the process of selecting and producing antique products in online and offline shopping malls related to antiques, based on a routine implementation.

[0075] Reference Figure 3 First, in step S301, device 400 can see product pages where the user spends more time than a preset baseline.

[0076] Device 400 can track the time users spend on each product page of an online store related to antiques. In this context, the time spent on each product page can predict which products users will pay attention to over a longer period, and the predicted products may be valuable as a date, with the assumption of a reference time serving as a basis for determining user interest in the products.

[0077] In step S302, device 400 may collect images contained on the product page, which are determined to be retro product images that exceed a preset threshold of user viewing.

[0078] For example, if the standard time is set to be greater than 2 minutes and less than 5 minutes, product images that have been viewed for more than 2 minutes without being turned are considered to be product images of interest to the user. If the user stays on the page for more than 5 minutes, it can be considered that the user is absent. Factors determining user activity include, but are not limited to, clicks, scrolling, and input.

[0079] In step S303, the device 400 can collect images of vintage products that have been exposed for more than a preset reference time from images captured by the user's camera.

[0080] Device 400 can measure the exposure time of product images by analyzing images directly captured by a user's camera in an offline shopping mall related to antiques. For example, if a reference time is set to 5 seconds according to step S302, and the exposure time of a product extracted from the entire video is greater than the reference time, it can be identified as an image of a product of interest to the user, while products with exposure times less than 5 seconds can be determined as part of the video included in the recording process. The criteria for judging the exposure of interest during shooting include, but are not limited to, the proportion of the entire video screen, the position of the camera screen of the product being filmed, and the speed of video movement.

[0081] In step S304, device 400 uses an image similarity analysis algorithm to determine the similarity of the collected year product images.

[0082] By comparing the retro product images corresponding to the products of interest collected in steps S302 and S303, similarity can be checked using an image similarity analysis algorithm.

[0083] Here, image similarity analysis algorithms primarily use CNNs (Convolutional Neural Networks) as a technique to evaluate the similarity between two given images and express their degree of similarity numerically. CNNs are deep learning models used for image analysis that learn image features by extracting spatial features. After extracting the feature vectors of the images, the similarity can be measured by calculating the distance between the feature vectors of two images. Typical CNN models include, but are not limited to, VGGNet, GoogLeNet, and ResNet.

[0084] In step S305, the device 400 can select an image with a similarity lower than a preset standard as an image of a product from a specific year.

[0085] Vintage items possess elements that differentiate them from existing products. Images with low similarity to existing items can be judged as having differentiating elements and thus selected as unique vintage products. This process plays a crucial role in identifying unique designs and original vintage products from various images collected from online and offline vintage-related stores, identifying potential vintage product candidates, and ultimately generating a distinctive vintage product image.

[0086] In step S306, device 400 can use an image search service to search for images of unusual antique products.

[0087] The selection of unusual year product images, chosen in phase S305, involves comparing collected images of products from different years for similarity. Searching for unusual year product images increases the reliability of the selection. For example, you can use image search services like Google Image Search or Bing Image Search to search for unusual antique products, use the search results to determine if similar products exist, and analyze the number and similarity of the searched images to determine if the image already exists.

[0088] In step S307, if the number of retrieved images is lower than a preset standard, the device 400 can select the product corresponding to the product image of a specific year as the product of the manufacturing year.

[0089] If a large number of images are collected through searching to improve reliability, and the number of collected images is below a preset threshold, then unusual antique products used in the search can be selected as antique products with unique designs and original production. Finally, the selected images can be sent to designers for production.

[0090] In this invention, by analyzing images collected from various sources and selecting products based on similarity and specificity, more accurate and efficient antique products can be produced, while maintaining the originality and uniqueness of antique products and simplifying the production process.

[0091] Figure 4 This is a flowchart illustrating the process of selecting retro products produced via social networks, based on an embodiment.

[0092] Reference Figure 4 First, in step S401, device 400 can check the uploaded product images uploaded to the user's social network account.

[0093] It analyzes images users upload to their social network accounts and allows users to identify uploaded product images included in year-related categories. Within social networks, user image analysis can use text and image analytics techniques to identify keywords or hashtags related to vintage products and select relevant images. Social networks include, but are not limited to, Facebook, Instagram, and Twitter.

[0094] In step S402, device 400 uses an image similarity analysis algorithm to determine the similarity between the collected retro product images and the uploaded product images.

[0095] The similarity between images can be derived from the feature vectors of each image using deep learning models such as convolutional neural networks (CNNs). Furthermore, CNNs can effectively extract spatial features from images, generating unique feature vectors that allow for comparison of the similarity between two images.

[0096] Similarity calculation includes, but is not limited to, cosine similarity Euclidean distance and Manhattan distance methods.

[0097] Specifically, cosine similarity measures similarity based on the angle between two vectors, Euclidean distance measures the straight-line distance between two vectors, and Manhattan distance measures the distance traveled along an axis between two vectors. By selecting a suitable similarity measurement method from these methods, the similarity between collected retro product images and uploaded product images can be calculated, thus allowing the selection of images with high similarity.

[0098] In step S403, the device 400 can determine that images with values ​​higher than a preset standard among the collected year product images are images that need to be identified.

[0099] Through similarity analysis, images with similarity exceeding the standard are selected as those requiring verification. As an example of a similarity measurement method, cosine similarity can represent... The value is 0.5, which ranges from -1 to 1. If the value is 0.5 or greater, it can be considered similar. Therefore, images that need to be confirmed by the 0.5 standard can be selected to effectively identify retro products with unique designs.

[0100] In step S404, device 400 can upload the image that needs to be verified to a social network account.

[0101] The selected image to be verified will be uploaded to the user's social network account. The uploaded image will be exposed to the user's friends or followers, who can see their preferences and reactions to the image. The uploaded image can be used for example, to ask questions like, "Which products do you prefer?" and to gather feedback from online users.

[0102] In step S405, device 400 can check the preferences of uploaded posts.

[0103] By analyzing reactions to posts on social networks, you can assess image preferences. Higher preferences indicate a more positive user response to retro products. You can collect and analyze user reaction data to evaluate the image's popularity. Reactions on social networks include, but are not limited to, likes, comments, and shares.

[0104] In step S406, device 400 can select products corresponding to the images posted in the post, with a preference higher than a preset threshold, for production as retro products.

[0105] Based on responses on social networks, images exceeding preference criteria can be selected for the final retro product. For example, if a post has over 100 likes, that image can be chosen as the retro product to be created, and the final product can be determined by reflecting the collective wisdom of users. In this case, preference criteria can be flexibly set, and responses over a certain period can be aggregated to determine the final product.

[0106] By reflecting actual user feedback, this invention can increase the likelihood of market success, maximize the value of antique products, and play a significant role in providing customized designs.

[0107] Figure 5 It is a flowchart illustrating the process of requesting designers to proceed with production based on a single embodiment.

[0108] Reference Figure 5 First, in step S501, the device 400 can analyze the manufactured retro product and generate data on material information, production methods, estimated production time, and estimated production costs.

[0109] You can identify all the materials that make up a product and collect detailed information about each material. For example, you can analyze the properties of the materials in detail, such as the type of fabric, color, thickness, and decorative materials, and store this information in a database. You can also define the methods for each step required to produce the product. This includes, but is not limited to, processes such as cutting, sewing, dyeing, and attaching embellishments.

[0110] Furthermore, the total production time is calculated by estimating the time required for each production step, and the estimated production cost is calculated by combining material costs, labor costs, and incidental costs. All of this information can be stored and used in the database.

[0111] In step S502, the device 400 can determine the production availability for designers based on the generated data.

[0112] The analyzed data (material information, production methods, estimated production time and cost) can be sent to the designer to determine production feasibility by considering his or her work plan, technical feasibility and material availability. The designer can review the data and communicate with the client, including all information related to production feasibility.

[0113] During the S503 phase, device 400 can send manufacturing information to the initiator terminal.

[0114] If a designer determines that a commissioned product is feasible, he or she will send an intention statement to the client, who can then confirm and decide whether to proceed with production. In this case, factors influencing the designer's determination of product feasibility include, but are not limited to, the designer's work schedule, technical feasibility, and material availability.

[0115] In the S504 phase, data can be transmitted to the designer's terminal under manufacturing conditions.

[0116] If, during the S502 phase, the designers determine that this is impossible, the data can be sent to alternative designers to reconfirm feasibility. This can be accomplished by finding alternative designers, evaluating each designer's capabilities and qualifications to select the best one, sending data to multiple designers sequentially, and collecting responses from each designer to choose the optimal one.

[0117] In step S505, device 400 can transmit the estimated manufacturing costs, including material costs, labor costs and other ancillary costs, to the sponsor's terminal.

[0118] If designers can proceed with production through steps S503 and S504 described above, the estimated production cost of the product can be transmitted to the customer's end. The estimated cost of manufacturing the product includes, but is not limited to, material costs, labor costs, and other ancillary costs.

[0119] Furthermore, if an estimated production cost, including a description of each cost item and the total cost, is calculated and sent to the sponsor, the client can review the details and decide whether to proceed with production.

[0120] In the S506 phase, device 400 allows the sponsor to confirm the estimated production cost and chooses to confirm the request on the sponsor's terminal.

[0121] After confirming the estimated production cost, the client can consult with the designer if necessary. Once the consultation is complete, the client can confirm the production request by pressing a button and send the confirmation to the designer.

[0122] This invention provides convenience for users and designers by automatically selecting retro products and automating the design request process, saving time and money. Furthermore, accurate analysis and rapid decision-making based on various data can greatly improve the efficiency of retro product production.

[0123] Figure 6 It is a flowchart illustrating the process of extracting product ingredients and selecting products from the production year according to an example.

[0124] Reference Figure 6 First, in step S601, the device 400 is able to extract the composition of the product from the collected vintage product images, including tearing, color, material and size.

[0125] By analyzing the collected images of antique products, you can extract detailed features from each image.

[0126] For example, it can identify materials through image texture analysis, extract key colors from color histograms, detect tears or damage through object recognition, and estimate product dimensions by analyzing image size and proportions. At this point, detailed components analyzed can be extracted and stored in a database, and the features of each image can be organized into data.

[0127] Product components extracted from vintage product images include, but are not limited to, tears, colors, materials, and sizes.

[0128] In the S602 stage, device 400 can identify common configurations in product images.

[0129] For example, if a product image consists of a first, second, third, and fourth product, the first, second, and third products might have a torn configuration, the second and third products might have the same color, and the second, third, and fourth products might have the same material. In this case, the common components of the products can be identified, and these common components can become a means of identifying the originality and uniqueness of vintage products.

[0130] In step S603, device 400 can select the most commonly used configuration from the identified general configurations to the configuration of the retro product.

[0131] For example, if a product image consists of a first, second, third, and fourth product, and the first, second, and third products have a tear configuration while the fourth product does not, then the tear configuration can be selected as a retro product configuration.

[0132] Furthermore, if the second and third products are the same color, while the first and fourth products are different colors, then the colors of the second and third products can be chosen as components of the vintage product.

[0133] In the S604 stage, device 400 is able to generate AI images based on the selected product configuration.

[0134] Vintage product components (including teardrops, colors, materials, and sizes) selected from the vintage product images collected in step S603 can be generated as text, and images can be generated using a text-based AI image generator.

[0135] At this point, AI image generation is a technology that generates suitable text images based on input text. It is achieved by combining natural language processing (NLP) and computer vision (CV). By using sequence models such as RNN, LSTM, GRU, or transformer models, text can be converted and encoded into vector form, and images that conform to the text description can be generated by receiving the encoded text vector as input.

[0136] In step S605, the device 400 can select the generated image as a manufactured retro product.

[0137] With the selected retro product configuration, the text-based AI image generator generates retro images that can be selected for production. The selected retro products are then sent to the designer for confirmation of whether they can be produced. If the designer reviews the details of the image and needs additional modifications, the modifications can be made and the work can be completed with the client's approval.

[0138] This invention utilizes text-based AI image generation technology to automate the design of retro products, which helps in the efficient production of customized products.

[0139] Figure 7 This is a preliminary schematic diagram of the structure of the device according to an embodiment.

[0140] Device 400 according to one embodiment includes a processor 401 and a memory 402. Device 400 according to one embodiment may be the aforementioned server or terminal. The processor can... Figures 1 to 6 Includes at least one of the aforementioned devices, or through Figures 1 to 6 At least one of the aforementioned methods is performed. Memory 402 may store information related to the aforementioned methods, or it may store programs that implement the methods. Memory 402 may be volatile or non-volatile memory.

[0141] Processor 401 can execute programs and control device 400. The code of the program executed by processor 401 can be stored in memory 402. Device 400 is connected to external devices (e.g., personal computers or networks) via input / output devices (not shown) and can exchange data.

[0142] The above embodiments can be implemented as hardware components, software components, and / or combinations of hardware and software components. For example, the devices, methods, and components described in the embodiments can be implemented using one or more general-purpose or special-purpose computers, such as processors, controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field-programmable gate arrays (FPGAs), programmable logic units (PLUs), microprocessors, or any other device capable of executing and responding to instructions. The processing unit can execute an operating system (OS) and one or more software applications running on the operating system. The processing unit can also access, store, manipulate, process, and generate data in response to software execution. For ease of understanding, the processing unit can be described as used as a whole; however, those skilled in the art will recognize that a processing unit may contain multiple processing elements and / or various types of processing elements. For example, a processing unit may include multiple processors or one processor and one controller. Furthermore, other processing configurations, such as parallel processors, may be used.

[0143] The method according to the embodiments can be implemented in the form of program instructions, which can be executed by various computer means and recorded on a computer-readable medium. The computer-readable medium may contain program instructions, data files, data structures, etc., and can be used alone or in combination. The program commands recorded in the medium may be specifically designed and configured for this embodiment, or may be well known and available to computer software articulators. Examples of computer-readable recording media include magnetic media (such as hard disks, floppy disks, and magnetic tapes), optical media (such as CD-ROMs and DVDs), magneto-optical media (such as floppy disks), and hardware devices specifically configured to store and execute program commands (such as ROMs, RAMs, flash memory, etc.). Examples of program instructions include machine code (e.g., code generated by a compiler) and high-level language code (e.g., code that a computer can execute using an interpreter). The hardware device may be configured to perform the operations of the embodiments as one or more software modules, and vice versa.

[0144] Software may include computer programs, code, instructions, or one or more combinations thereof, and may be configured to operate as intended by a processing unit, or may command the processing unit independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual device, computer storage medium or device, or in transmitted signal waves, so that it may be interpreted by the processing unit or used to provide instructions or data to the processing unit. This software is distributed across networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0145] Although the above embodiments have been described with limited accompanying drawings, those skilled in the art can make various modifications and alterations based on the above applications. For example, appropriate results can be obtained if the described techniques are performed in a different order than the described methods, and / or if the components of the described systems, structures, devices, circuits, etc., are combined or combined in a different manner than the described methods, or are replaced or substituted by other components or equivalents.

[0146] Therefore, other implementations, other methods of implementation, and implementations equivalent to the patent claims also fall within the scope of the following claims.

Claims

1. An automated method for selecting and requesting vintage product design, executed by a device, wherein, the automated method for selecting and requesting vintage product design comprises: collecting pictures of vintage products from vintage related online marketplaces and vintage related offline shopping centers; selecting a vintage product to be produced using the vintage product pictures collected above; and transferring the aforementioned produced vintage product to a designer's terminal and asking for its production.

2. The automated method for selecting and requesting vintage product design according to claim 1, wherein, the collecting pictures of vintage products from vintage related online marketplaces and vintage related offline shopping centers comprises: collecting pictures of vintage products from vintage related online marketplaces and vintage related offline shopping centers; the selecting a vintage product to be produced using the vintage product pictures collected above comprises: selecting a vintage product to be produced using the vintage product pictures collected above; and the transferring the aforementioned produced vintage product to a designer's terminal and asking for its production comprises: transferring

Images