AI agent method for personalized wearing and building recommendation

Abstract

The invention discloses an AI intelligent agent method for personalized wearing and fitting recommendation. A photo processing module, a human body model module, a clothes data module, an AI model module, an AI result feedback module and a personal historical data module are included in sequence. The photo processing module receives, stores and processes user photos and extracts key information such as human face, skin color and gender, the human body model module can create a 3D model according to the size of a user body, and the AI model module generates a user wearing recommendation according to the information extracted by the photo processing module, the human body 3D model, clothes data in the clothes data module and an AI algorithm. Obtaining clothing composition details recommended by wearing and overlapping; the user can check the recommended clothes and clothes composition; and an AI algorithm is optimized according to operation data such as retention time, collection and the like to obtain a user personalized model. According to the method, on the basis of the generative AI large model, the clothes wearing is enabled, and the quality and the efficiency are improved, so that people can quickly find favorite clothes, and the time and the money of people are saved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of generative AI large models, and relates to a method for making dressing recommendations according to human body information and personal preferences, in particular to an AI agent method for personalized dressing recommendations. BACKGROUND

[0002] With the popularity of GPT and other generative AI large models, a large number of companies in China are also actively involved in this competition. At present, the underlying of large models has been perfected, and in the future it will be the era of intelligent agents in various subfields. By meeting the different scene needs of people, the production tool will be qualitatively improved and leap forward, and the AI large model will truly penetrate into various industries, improving and empowering various industries, and completing the complete transformation of the production mode. At the same time, the main way for people to buy clothes at present is to search for keywords on e-commerce websites and select clothes they like in the search list. Sometimes the clothes purchased online may not fit or the clothes may not meet their demands for being handsome, beautiful, and appropriate, resulting in return and waste of valuable time. Going to the mall to try on clothes before purchasing is not only laborious but also expensive in physical stores. In addition, it is required to have a high demand for one's own aesthetics to dress in a good-looking outfit. This problem has been bothering people. Therefore, it is necessary to provide a method that combines AI large models to meet people's individualized requirements for dressing and to achieve the demands of being appropriate, beautiful, and handsome through personal self-help without the help of others. SUMMARY

[0003] The purpose of the present application is to provide an AI agent method for personalized dressing recommendations, which uses generative AI large models to empower the field of clothing dressing and enables people to obtain appropriate and beautiful dressing through algorithm recommendations.

[0004] To solve the above technical problems, the technical solution adopted by the present application is: An AI agent method for personalized dressing recommendations, which sequentially comprises a photo processing module, a human body model module, a clothing data module, an AI model module, an AI result feedback module, and a personal historical data module. The photo processing module is responsible for receiving, storing, and processing human body photos and extracting key information such as body shape, head, face, skin color, gender, and hair information. The photo processing module supports receiving multiple photos, such as front, side, and back photos, as well as 360-degree rotating human body appearance videos taken in one week, in order to enable the AI model module to better and more comprehensively understand the current personnel object. Note that it is not necessary to undress, and dressing photos can meet the information recognition requirements. The human body model module is a 3D modeling center, responsible for receiving user input human body size for human body modeling, such as height, weight, neck circumference, arm length, shoulder width, leg length, waist circumference, waist, chest circumference, waist circumference, hip circumference, thigh circumference, etc. Personnel data modeling, while the face and skin color collected by the photo processing module are rendered into 3D models, so that a sensory real human 3D model can be obtained; The clothing data module is a clothing database that stores a large amount of clothing data and records various related information of the clothing, such as clothing photos, names, brands, fabrics, types, styles, clothing sizes, length dimensions, etc. The AI model module is a generative AI large model main body, including built-in general recommendation algorithm and each user's customized and optimized personalized algorithm, which is the core tool for dressing recommendation; By assembling the photo information received by the photo processing module, the model information of the human body model module, the clothing information of the clothing data module, and the recommendation algorithm information of the AI model module, a suitable dressing recommendation is formed, and a photo or video output with a real face and suitable body is generated, which is displayed to the user through the AI result feedback module; The AI result feedback module is responsible for presenting the recommendation results of the AI model module to the user, so that the user can not only view the dressing effect, but also obtain detailed composition of dressing recommendation, such as coat, underwear, and pants collocation information, so as to facilitate ordering and purchasing on e-commerce platform; At the same time, this module is also responsible for collecting and recording user operation information, and storing it into the personal historical data module, to prepare for further optimization of personalized model; The personal historical data module is a record module that stores user preferences, collections, and dwell time traces, responsible for feeding user data to the algorithm, so as to optimize the personalized model that fits the user's dressing preferences; The data flow for dressing recommendation is as follows: the user uploads personal photos to the photo processing module, the photo processing module processes the photos and collects key information; At the same time, according to the user's setting of human body model, the user's 3D model is modeled; Combined with the clothing data module, the current recommendation parameter set by the user and other information, the recommendation algorithm of the AI model module is calculated, and the dressing recommendation is generated to the AI result feedback module to display to the user.

[0005] As an improvement, when the user uses it initially, the information in the personal history data module is less, that is, the model cannot perceive the user's preferences, and the general recommendation algorithm trained in advance is mainly used; when the user uses it for a period of time, the personal history data module collects enough data such as likes, collections, and residence time, and then the user data is fed into the model to gradually optimize the personalized model of the user, and then the personalized model of the user is used for dressing recommendation.

[0006] As an improvement, when the information in the user's personal history data module is less, some dressing recommendations can be randomly generated by the AI model module and displayed to the user, and the user scores the dressing recommendation results, and the scoring data is used as the personal history data module to train the model, so that the personalized model recommendation algorithm that meets the user's individuality can be obtained.

[0007] As an improvement, the photo processing module can receive, store and process the user's daily appearance photos, so as to be used as the training data of the AI model module, so as to extract the user's dressing preferences, and the more and richer the daily appearance photos uploaded by the user, the more accurate the personalized model trained, and the more the dressing recommended meets the user's expectations.

[0008] As an improvement, when the AI model generates the recommended dressing, in addition to considering the recommendation parameters in the recommendation algorithm, it also needs to consider the parameters set by the user, such as changing casual style to campus style, or changing the recommended spring dressing to the recommended summer dressing, or changing the recommended whole body dressing to only recommending the coat dressing, and then the recommendation result of the model needs to be dynamically adjusted.

[0009] As an improvement, in order to reduce the user's use threshold, the human body model module only needs to collect height, weight and age to create a general human body model, without setting detailed girth, three-circumference, thigh circumference and various size information; Although it cannot approach the real user 3D model, when the user does not want to provide these data or does not want to provide these data, the process will not be interrupted, and the desired recommendation result can be quickly obtained, so that the user's use experience is greatly improved.

[0010] As an improvement, the clothing data module records information including the information of merchants and prices of goods on e-commerce platforms, and when the user is satisfied with the recommended clothing dressing, the corresponding merchant and price source information can be obtained by disassembling the coat, underwear and trousers etc. obtained by the model, and jumping to the corresponding merchant store on the e-commerce platform is supported, so that the user can obtain a smooth use experience.

[0011] The beneficial effects of the present application are: The application can obtain good clothing recommendation with low cost according to the personal preferences of the user. In the case that the user does not have good aesthetics, appropriate dressing can still be obtained, and the emotional appeal of ordinary people who love beauty can be met. At the same time, the time cost caused by return after online shopping, and the time and money cost of trying on clothes in a shopping mall can be saved. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1 : The module structure diagram of the AI intelligent body method for personalized dressing recommendation of the application; Figure 2 : The data flow diagram of the AI intelligent body method for personalized dressing recommendation of the application; Figure 3 : The personalized model optimization flowchart of the AI intelligent body method for personalized dressing recommendation of the application.

[0013] The figure mark: 1-Photo processing module; 2-Human body model module; 3-Clothing data module; 4-AI model module; 5-AI result feedback module; 6-Personal history data module. DETAILED DESCRIPTION

[0014] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is described and explained below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application. Based on the examples provided in the present application, all other examples obtained by those of ordinary skill in the art without making creative efforts fall within the scope of the present application.

[0015] Obviously, the drawings in the following description are only some examples or embodiments of the present application, and for those of ordinary skill in the art, the present application can be applied to other similar scenarios without making creative efforts based on these drawings. In addition, it can be understood that although the efforts made in this development process can be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in the present application, some design, manufacture or production changes based on the technical content disclosed in the present application are only routine technical means, and should not be understood as insufficient disclosure of the content disclosed in the present application.

[0016] In the present application, the phrase "embodiment" means that the specific features, structures or characteristics described in combination with the embodiment can be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily mean the same embodiment, nor is it an independent or alternative embodiment that is not mutually exclusive with other embodiments. Those of ordinary skill in the art explicitly and implicitly understand that the embodiments described in the present application can be combined with other embodiments without conflict.

[0017] Unless otherwise defined, technical terms or scientific terms used in the present application shall have the same meaning as commonly understood by one of ordinary skill in the art to which the present application pertains. The terms "a", "an", "one", "this", and similar terms as used herein do not denote a limitation of quantity but denote the presence of at least one of the referenced item. The terms "include", "comprise", "have", and any variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, system, product, or device that comprises a list of steps or units not only comprises the steps or units in the list but can also comprise other steps or units not expressly listed or other steps or units inherent to such process, method, product, or device. The terms "connect", "connected", "coupling", and similar terms as used herein are not limited to a direct or physical connection, but can include an electrical connection, whether direct or indirect. The terms "a plurality of", "N groups of", and "N" refer to two or more.

[0018] The embodiments of the present application provide an AI agent method for personalized dressing recommendation, as shown in the method includes a photo processing module 1, a human body model module 2, a clothing data module 3, an AI model module 4, an AI result feedback module 5, and a personal history data module 6. Figures 1 to 3 As shown in the method includes a photo processing module 1, a human body model module 2, a clothing data module 3, an AI model module 4, an AI result feedback module 5, and a personal history data module 6.

[0019] As shown in the method includes a photo processing module 1, a human body model module 2, a clothing data module 3, an AI model module 4, an AI result feedback module 5, and a personal history data module 6. Figures 1-2 As shown in the method includes a photo processing module 1, a human body model module 2, a clothing data module 3, an AI model module 4, an AI result feedback module 5, and a personal history data module 6. Specifically, the received photos support multiple photos, such as front view, side view, and back view; and the received video is a 360-degree video around the human body, so as to obtain comprehensive appearance information of the user's body shape. It is particularly noted that the user does not need to take off clothes when taking photos or videos, and can wear appropriate clothes.

[0020] As shown in the method includes a photo processing module 1, a human body model module 2, a clothing data module 3, an AI model module 4, an AI result feedback module 5, and a personal history data module 6. Figures 1-2 As shown in the method includes a photo processing module 1, a human body model module 2, a clothing data module 3, an AI model module 4, an AI result feedback module 5, and a personal history data module 6. It should be noted that the more and more accurate size data provided by the user, the more the 3D model creation result fits the real body shape, and the more the finally generated dressing effect approaches the real wearing of the user. Considering the user's unfamiliarity with the size and the user's convenience, the human body model module 2 supports only entering core key data such as age, height, and weight, so as to create a common version of 3D model, and perform clothing wearing effect rendering based on the common version of 3D model.

[0021] As Figures 1-2 shown, the clothing data module 3 is a clothing database for storing a large amount of original clothing data, recording relevant information of the clothing, such as clothing photos, names, brands, fabrics, types, styles, clothing sizes, length sizes, and various information; used as a recommended clothing source library of the AI model module 3; It should be noted that the clothing data module 3 records relevant information of the clothing, which can be more detailed, such as recording information including relevant stores and prices of the clothing on e-commerce platforms, so that when the dressing and outfit is recommended to the user, the user can view the corresponding store and reference price information of each clothing.

[0022] As Figures 1-2 shown, the AI model module 4 is a generative AI large model, which has built-in general recommendation algorithms and personalized algorithms for each user; responsible for integrating the user information in the photo processing module 1, the human body 3D model in the human body model module 2, and the clothing database in the clothing data module 3 according to the AI model recommendation algorithm, and receiving the current recommendation parameters of the user in real time, and finally generating a human dressing effect picture (including output photos or videos), and displaying the recommendation results to the user through the AI result feedback module 5; Specifically, the current recommendation parameters of the user can be to adjust the recommended style from casual to formal, adjust the spring outfit to summer outfit, etc. The AI result feedback module 5 is responsible for presenting the recommendation results of the AI model module 4 to the user, and also can view the detailed composition of the dressing recommendation through the AI result feedback module 5, such as the composition information of the coat, underwear, and pants, so as to facilitate the user to place an order on the e-commerce platform; at the same time, this module is also responsible for collecting and recording the operation information of the user, and storing it in the personal historical data module 6; The personal historical data module 6 is a recording module for storing user preferences, collections, and dwell time traces, which prepares for further optimization of personalized models; Specifically, the user data in the personal historical data module 6 is fed to the algorithm, so as to optimize the personalized model that fits the user's dressing preferences; when the user's dressing preference data is less, the AI model module 4 can also randomly generate some recommended outfits, and the user can score the like degree of the recommendation results, so as to also obtain the user's preference information and store it in the personal historical data module 6.

[0023] As Figures 2-3As shown, the method realizes the data flow of user personalized dressing recommendation from collecting user photo information to completing dressing recommendation, and according to user historical data. The user uploads personal photos through the photo processing module 1, and extracts the key information of the user; the user sets the body size data through the human body model module 2, and obtains the 3D model of the user; the AI model module 4 generates a plurality of dressing recommendation effect pictures or videos with the real portrait and figure of the user according to the user key information, 3D model, clothing data, and current recommendation parameters set by the user according to the recommendation algorithm, and displays them to the user one by one; during the process of the user viewing the recommended effect pictures one by one, the user can further view the detailed dressing and collocation of the preferred recommendation, such as coat, inside, trousers, etc., and store the user viewing action, dwell time, collection, like, etc. Information to the personal historical data module 6. In order to optimize the personalized model in line with the user's preference, the photo processing module 1 is supported to receive the user's daily image photo, and the user information is extracted so as to train the AI model module, so that the AI model learns the user's dressing preference. The more daily photos the user uploads, the more the dressing recommendation will meet the user's expectations; at the same time, the stored personal historical data can be fed to the AI model module 4, so as to gradually optimize the dressing recommendation algorithm in line with the current user's individualization.

[0024] Those skilled in the art should understand that each technical feature of the above-described embodiments can be combined arbitrarily, and in order to make the description concise, each technical feature in the above-described embodiments has not been described all possible combinations, however, as long as the combination of these technical features does not exist contradictory, it should be considered as the scope of the present application.

[0025] The above-described embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be noted that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of the patent of the present application should be subject to the appended claims.

Claims

1. A personalized outfit recommendation AI agent method, comprising, in sequence, a photo processing module, a human body model module, a clothing data module, an AI model module, an AI result feedback module, and a personal historical data module; Its special feature is that the photo processing module is responsible for receiving, storing, and processing photos or 360-degree human body rotation videos uploaded by users, and extracting key information such as the user's face, skin color, and gender; The human body model module receives user-input size data and builds a 3D human body model; the clothing data module stores a large amount of clothing data; the AI ​​model module has built-in general recommendation algorithms and user-personalized recommendation algorithms. The AI ​​model module can combine user photo information, human body model information, and clothing database information with algorithms to generate recommended outfit images for the user; the AI ​​result feedback module presents the generated recommended outfit images to the user and is responsible for collecting user dwell time, likes, favorites, and other actions, storing them in the personal history data module; and feeding the AI ​​model module with data from the personal history data module to gradually optimize a personalized recommendation model unique to the user; at the same time, information extracted from the photo processing module can also be used as a data source to train the AI ​​model module, thereby optimizing the user-personalized model.

2. The AI ​​agent method for personalized outfit recommendation as described in claim 1, characterized in that: The photo processing module supports receiving multiple user photos or videos taken by rotating the human body 360 degrees, with the aim of better obtaining key information such as the user's body shape, gender, face, and skin color.

3. The AI ​​agent method for personalized outfit recommendation as described in claim 1, characterized in that: The human body model module can model a person's body shape based on data such as height, weight, neck circumference, arm length, shoulder width, leg length, waist circumference, mid-waist, chest circumference, trouser waist circumference, hip circumference, and thigh circumference input by the user. In order to lower the threshold for users, it supports modeling a normal model based only on height, weight, and age, and uses the normal model as the user's body shape to make clothing recommendations.

4. The AI ​​agent method for personalized outfit recommendation as described in claim 1, characterized in that: The AI ​​model module not only provides outfit recommendation images with real user avatars, but also provides the clothing components corresponding to the recommended outfits, which are presented to the user through the AI ​​result feedback module.

5. The AI ​​agent method for personalized outfit recommendation as described in claim 1, characterized in that: The clothing data module records not only clothing attribute information, such as photos, names, brands, fabrics, types, styles, clothing sizes, and lengths, but also product attribute information on e-commerce platforms, such as store information and reference prices. This allows users to smoothly navigate to the corresponding e-commerce platform and store to place an order when viewing the clothing composition in the AI ​​result feedback module.

6. The AI ​​agent method for personalized outfit recommendation as described in claim 1, characterized in that: The personal history data module is used to record the user's personal operation history, such as likes, favorites, and the duration of stay for recommended outfits. This information is used as training data for the AI ​​model module to optimize the current personalized recommendation model for the user. In the early stages of user use, when personal history data is not yet abundant, the AI ​​model can randomly recommend some outfits, allowing the user to rate the recommendations and thus obtain user preference information.

7. The AI ​​agent method for personalized outfit recommendation as described in claim 1, characterized in that: When the user's historical data module contains limited information, the user can upload a large number of daily photos through the photo processing module to extract the user's daily outfit preferences. These photos can then be used as a data source to train the AI ​​model module, thereby optimizing the recommendation model to match the user's personalization.

Images