A clothing product display method and system based on AI

Abstract

Multiple embodiments of this specification relate to the field of information technology, and specifically to a method and system for displaying clothing products based on AI. The method includes the following steps: receiving a three-dimensional model of a target garment and generating a virtual model; receiving product images of multiple other garments in a display page and information of a currently logged-in user, generating a target model posture according to the multiple product images and the user information, and adjusting the posture of the virtual model according to the target model posture; fusing the three-dimensional model of the target garment with the virtual model to obtain a fused model; generating a scene according to the multiple product images and the user information, and placing the fused model into the scene; creating a framing camera in the scene according to the target model posture, and obtaining a target display picture according to the framing camera.

Description

Technical Field

[0001] Multiple embodiments of this specification relate to the field of information technology, and specifically to an AI-based clothing product display method and system. Background Art

[0002] The application of AI technology in the field of clothing is becoming more and more widespread. It not only revolutionizes the traditional shopping experience, but also provides new marketing and service tools for brands and retailers. For example, applications such as image recognition, personalized recommendations, and virtual try-on. Image recognition technology enables the system to automatically analyze the style, color, texture and other characteristics of clothing in uploaded pictures, and match this information with the clothing information stored in the database. For example, a user can take a photo or upload a picture of a favorite piece of clothing, and the AI ​​system can identify similar styles of clothing and provide a purchase link. Personalized recommendation is also a highlight of AI in the field of clothing. By building user portraits, AI algorithms can predict which products may meet the user's preferences, and actively push related products. The virtual try-on function uses augmented reality (AR) and computer vision technology to allow consumers to try different clothing combinations online and see the effect of wearing clothes on themselves. This feature is very attractive to customers who want to make purchasing decisions at home because it reduces uncertainty and enhances shopping confidence. Product images have an important impact on product visits in current e-commerce, especially the main picture of the product displayed in the product list, but the current product images rely on real-life photography. Real-life photography is not only inefficient but also has limited scenes and expressiveness. Therefore, it is necessary to study technologies that can help improve the expressiveness of product images. Summary of the invention

[0003] Multiple embodiments of this specification describe an AI-based clothing product display method and system.

[0004] In a first aspect, the embodiments of this specification provide a clothing product display method based on AI, comprising the steps of:

[0005] Receiving a three-dimensional model of a target garment and generating a virtual model;

[0006] Receiving a plurality of other clothing product images and currently logged-in user information on a display page, generating a target model posture according to the plurality of product images and the user information, and adjusting the posture of the virtual model according to the target model posture;

[0007] Fusing the three-dimensional model of the target garment with the virtual model to obtain a fused model;

[0008] Generate a scene according to the plurality of product images and the user information, and place the fused model into the scene;

[0009] A framing camera is created in the scene according to the target model's posture, and a target display picture is obtained according to the framing camera.

[0010] In a second aspect, the embodiments of this specification provide an AI-based clothing product display system, including:

[0011] A first receiving module receives a three-dimensional model of a target garment and generates a virtual model;

[0012] A second receiving module receives a plurality of product pictures of other clothing in the display page and information of a currently logged-in user, generates a target model posture according to the plurality of product pictures and the user information, and adjusts the posture of the virtual model according to the target model posture;

[0013] A fusion module, which fuses the three-dimensional model of the target garment with the virtual model to obtain a fused model;

[0014] A generating module, generating a scene according to the plurality of product images and the user information, and placing the fused model into the scene;

[0015] A framing module is used to create a framing camera in the scene according to the target model's posture, and obtain a target display picture according to the framing camera.

[0016] In a third aspect, an embodiment of this specification provides an electronic device, including a processor and a memory;

[0017] The processor is connected to the memory;

[0018] The memory is used to store executable program code;

[0019] The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the method described in any one of the above aspects.

[0020] In a fourth aspect, an embodiment of the present specification provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in any of the above aspects.

[0021] In a fifth aspect, an embodiment of this specification provides a computer program product, including a computer program, which implements the method described in any of the above aspects when executed by a processor.

[0022] The beneficial effects brought by the technical solutions provided by some embodiments of this specification include at least:

[0023] In multiple embodiments of this specification, the provided AI-based clothing commodity display method can adopt the generation of target display pictures assisted by AI computer, so that clothing commodities can be quickly and automatically displayed by multiple models with different styles, and more abundant and more expressive target display pictures can be obtained at a lower cost, thereby improving the efficiency of obtaining clothing display pictures and reducing costs; by combining the user information of the currently logged-in user, different target display pictures can be generated for each user, providing personalized services, making clothing more likely to attract the attention of the currently logged-in user; by obtaining adjacent product pictures in the product display page of the e-commerce platform, targeted generation of postures and scenes that are harmonious with the style and posture of the adjacent product pictures, but different from each other, further improves the possibility of the target clothing being paid attention to by the currently logged-in user.

[0024] Other features and advantages of the various embodiments of the present specification will be further disclosed in the following detailed description and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of this specification, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 This is a schematic diagram of a product display page used in an embodiment of this specification.

[0027] Figure 2 A schematic diagram of the target display image generation process provided in the embodiments of this specification.

[0028] Figure 3 This is a schematic diagram of the architecture used in the embodiments of this specification.

[0029] Figure 4 A schematic diagram of the interactive interface provided in the embodiments of this specification.

[0030] Figure 5 A schematic diagram of a method for generating a virtual model provided in an embodiment of this specification.

[0031] Figure 6 This is a schematic diagram of the gesture library provided in the embodiments of this specification.

[0032] Figure 7 This is a schematic diagram of the target model posture acquisition process provided in the embodiments of this specification.

[0033] Figure 8 A schematic diagram of the virtual model fusion process provided in an embodiment of this specification.

[0034] Fig. 9 Schematic diagram of the AI-based clothing product display system provided in the embodiments of this specification.

[0035] Fig.10 A schematic diagram of an electronic device provided in an embodiment of this specification.

[0036] Among them, 11, target clothing, 12, target display picture, 13, virtual model, 14, fusion model, 15, product picture, 16, target model posture, 17, scene, 18, product display page, 19, posture library, 20, reference posture, 21, calibration point, 201, e-commerce platform, 202, server, 203, terminal device, 204, interactive interface, 1100, electronic device, 1101, processor, 1102, communication bus, 1103, user interface, 1104, network interface, 1105, memory. DETAILED DESCRIPTION

[0037] The following is an explanation and description of the technical solutions of the embodiments of this specification in conjunction with the drawings of the embodiments of this specification, but the following embodiments are only preferred embodiments of this specification, not all. Based on the embodiments in the implementation mode, other embodiments obtained by those skilled in the art without creative work are all within the scope of protection of this specification.

[0038] The terms "first", "second", "third", etc. in the description and claims of this specification and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0039] In the following description, terms such as "inside", "outside", "up", "down", "left", "right", etc. that indicate directions or positional relationships are only used to facilitate the description of the embodiments and simplify the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as a limitation of this specification.

[0040] The data involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection of relevant data complies with the relevant laws, regulations and standards of relevant countries and regions.

[0041] Before describing the technical solution in this specification, an introduction is given to the application scenarios and related technologies of the technical solution.

[0042] AI, or Artificial Intelligence, refers to intelligent behavior exhibited by computer systems. It is a discipline that studies and develops technologies and theories that can simulate, extend, or replace human intelligence. With the development of AI, more and more industries have achieved technological development and application innovation with the help of AI technology. For example, AI-assisted clothing design tools can generate unique clothing design solutions based on the latest fashion trends and personal style suggestions. Fashion designers can use these tools to quickly iterate their ideas while ensuring that their works meet current aesthetic standards. The intelligent recommendation system implemented by AI provides users with personalized clothing matching suggestions by analyzing their browsing history, purchase records and other data. This precision marketing method not only improves customer satisfaction, but also increases cross-selling opportunities. Virtual fitting technology combines computer vision and augmented reality, allowing consumers to "try on" clothes online, view the effects from different angles, and even simulate the dynamic performance of clothes when walking. It effectively improves the online shopping experience and reduces the return rate caused by inconsistent sizes or styles.

[0043] With the development of e-commerce, online clothing purchases have become an important way to shop for clothing. When people buy clothing online, product images15 are one of the important ways for consumers to understand and evaluate clothing. Since it is impossible to touch or try on the products directly, high-quality and informative pictures have an important impact on purchasing decisions. Please refer to the attached Figure 1 and attached Figure 2 , goods a, b, c, e, and f are all product images 15, and product images 15 specifically refer to product images 15 of goods from other merchants, and product d is a target display image 12, and target display image 12 specifically refers to the display of the image of the target clothing 11 processed by this application. This application uses AI technology to generate target display images 12 based on product images 15 of other merchants and target clothing 11. The scenes 17 of goods a and c are all street scenes. Product b is a specially shot staircase scene 17, with the model sitting on the steps, which is suitable for showing the elegant temperament of this clothing. The product images 15 in goods e and f are taken in a studio with a nearly pure color background, and the clothing itself is fully displayed. Product d is the target display image 12 generated by the target clothing 11. However, the current product images 15 are still mainly shot by manual shooting, which is not only inefficient but also has limited scenes 17. In addition, the models have different abilities to show clothing. Generally, only one model is used to shoot a piece of clothing, and the scene 17 and display richness are insufficient. The scene 17 represents the background in the product image 15 and the target display image, that is, the environment the model is in. For example, stairs, a photo studio, and a natural environment are all scenes.

[0044] This specification provides an AI-based clothing product display method, which generates clothing display information through AI assistance, mainly based on the information of other clothing product pictures 15 on the result page, to help improve the display power of the target display picture 12 and improve the efficiency of producing the target display picture 12. The process is as follows Figure 2 See attached. Figure 3 , this specification discloses an AI-based clothing product display method, which is used to generate an asynchronous target display image 12 of a clothing product at a given display position. The asynchronous position can be determined by an agreement between a clothing merchant and an e-commerce platform 201, or obtained by payment. When the e-commerce platform 201 displays the target clothing 11 for the currently logged-in user, the server 202 of the clothing merchant will be notified. The user information of the currently logged-in user and other adjacent product images 15 are provided to the server 202 of the clothing merchant. After the clothing merchant's server 202 runs this method, it generates a target display image 12 and feeds it back to the e-commerce platform 201, which is displayed to the currently logged-in user by the e-commerce platform 201.

[0045] This specification provides an AI-based clothing product display method, which is applied to a system including a server 202 and a user terminal. Figure 4 , an interactive interface 204 is deployed on the terminal device 203, wherein the interactive interface 204 can be run on the terminal device 203 in the form of a browser, or in the form of an independent application (APP), etc., and the specific presentation form of the client is not limited here. The server 202 involved in this application can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms. The terminal device 203 can be a smart phone, a tablet computer, a laptop, a PDA, a personal computer, a smart speaker, a smart TV, a smart watch, a car device, a wearable device, etc., but is not limited thereto. The terminal device 203 and the server 202 can be directly or indirectly connected by wired or wireless communication, which is not limited in this application. The number of servers 202 and terminal devices 203 is also not limited. A data connection is established between the server 202 and the e-commerce platform 201 for obtaining data from the e-commerce platform 201 and feeding back the generated target display picture 12 to the e-commerce platform 201 for displaying clothing products.

[0046] Since this application involves some professional terms, these professional terms will be introduced below.

[0047] Avatar

[0048] A virtual model is a digital human figure created through computer graphics technology, which is used to display clothing, accessories or other fashion products in a virtual environment. These virtual images can be completely fictional characters or highly realistic characters generated by scanning and modeling real people. Building a virtual model 13 requires a combination of multiple advanced technologies, including but not limited to 3D modeling, texture mapping, physical simulation, animation production, real-time rendering, etc. With the development of artificial intelligence (AI), virtual models 13 can also integrate intelligent algorithms to achieve more natural expressions and movements.

[0049] Model pose

[0050] Model posture refers to the body position and expression of the model during the shooting or presentation. It includes various aspects such as standing, walking, sitting, hand gestures and facial expressions. Different postures can convey various emotions, such as confidence, elegance, vitality, etc. At the same time, it helps to show the design highlights of the clothing, such as tailoring and decoration.

[0051] Scenario

[0052] The scene refers to the environment in which the model is located. It can be an indoor studio, an outdoor natural scenery, or a carefully arranged art installation. The selection and design of the scene is crucial to creating an atmosphere and telling a story, because it directly affects the audience's first impression of the work.

[0053] Viewfinder Camera

[0054] In 3D modeling software, the camera is a basic and crucial concept. The concept of the camera in the scene 17 is consistent with the camera in the 3D modeling software. The position, direction and viewing angle of the observer are clarified, thereby controlling the final output image content. Combining the scene 17 and the fusion model 14 therein, the target display image 12 can be obtained.

[0055] For details, please refer again to the attached Figure 2 This specification first provides an AI-based clothing product display method, including the steps of:

[0056] Receiving a three-dimensional model of a target garment 11 and generating a virtual model 13;

[0057] Receiving a plurality of product images 15 of other clothing in a display page, generating a target model posture 16 according to the plurality of product images 15, and adjusting the posture of the virtual model 13 according to the target model posture 16;

[0058] Merging the three-dimensional model of the target garment 11 with the virtual model 13 to obtain a fused model 14;

[0059] Generate a scene 17 according to the plurality of product images 15, and place the fused model 14 into the scene 17;

[0060] A framing camera is created in the scene 17 according to the target model posture 16, and a target display picture 12 is obtained according to the framing camera.

[0061] Among them, the three-dimensional model of the target clothing 11 can be realized by using 3D modeling software. For example, professional 3D modeling software such as CLO 3D, Marvelous Designer, Blender, etc. are used to create a three-dimensional model of clothing. This type of software provides a rich tool set, including functions such as cloth physics simulation and texture mapping to achieve highly realistic effects. Or a method of obtaining a three-dimensional model of clothing using scanning technology. A real clothing sample is captured by a 3D scanner and then converted into a digital three-dimensional model. This method is particularly suitable for applications such as replicating existing physical objects or protecting cultural heritage. It has high accuracy and can accurately reproduce all the features of the original object. It is fast, and the scanning process is relatively fast compared to manual modeling.

[0062] On the other hand, please see the attached Figure 5 , the method of generating the virtual model 13 includes:

[0063] Step S101: obtaining clothing information according to the three-dimensional model of the target clothing 11, wherein the clothing information includes style, size, color, pattern and design elements.

[0064] Size, color and pattern can be directly obtained from the three-dimensional model. Exemplary patterns are stripes, plaids, polka dots, and prints. When obtaining styles and design elements, a two-dimensional image of the three-dimensional model is obtained from the front of the three-dimensional model of the target garment 11. The two-dimensional image of the front is input into a pre-connected AI model, such as a large language model, and the AI ​​recognizes the style and design elements. Styles include dresses, jackets, pants, coats, skirts, etc. Design elements include puff sleeves, lantern sleeves, A-line skirts, embroidery, lace, sequins, etc.

[0065] Step S102: obtaining a description of the clothing according to the clothing information, wherein the description includes a description of wearing occasions, target audience, design emotion, and design style.

[0066] The description of the clothing is generated by AI, specifically by inputting the clothing information into AI and sending the self-heating language task of generating the description to AI.

[0067] For example, the description of the clothing is style: A-line suspender dress, color: cinnamon, pattern: no obvious pattern, simple design. Design elements: deep V-neck, suspenders, slim waist.

[0068] The description of the garment was provided to AI, and the description obtained was: This cinnamon-colored A-line suspender dress is an ideal choice tailored for modern women who pursue elegance and comfort. It is suitable for a variety of formal or semi-formal occasions, such as business meetings, evening parties or weekend dates. Its simple yet sophisticated design conveys a low-key and confident emotion, which is neither too ostentatious nor fashionable. The deep V-neck design cleverly elongates the neck line; while the slim waist cut and the gradually expanding skirt perfectly show the beauty of women's curves, giving the wearer confidence and charm.

[0069] Step S103: generating guiding words according to the description, wherein the guiding words include gender guiding words, appearance guiding words, makeup guiding words, body guiding words, expression and posture guiding words, and action and posture guiding words.

[0070] Then the description is input to AI, and AI is instructed to generate guidance words, including gender guidance words, appearance guidance words, makeup guidance words, body guidance words, expression and posture guidance words, and action and posture guidance words. The output of AI is:

[0071] Based on the above descriptions, we can generate the following guidance words for this A-line suspender dress.

[0072] Gender Guidance Words: Female;

[0073] Appearance guidelines: Fresh and natural facial features, elegant and graceful face, with soft facial features such as round cheeks and delicate chin lines;

[0074] Makeup guide: natural nude makeup, emphasizing the radiance of the skin, deep but not strong eye makeup, earth-tone eyeshadow, reddish brown or rose-colored lip gloss can be used to coordinate with the overall color tone while retaining feminine charm, and a slight blush can add a healthy complexion;

[0075] Body shape guide: Suitable for a variety of body types, especially women who want to highlight the waist curve and leg lines. The slim but not too tight design is suitable for different body types, especially recommended for women with S-shaped curves.

[0076] Facial expression and posture guidance: A gentle smile conveys a sense of confidence and friendliness, and a look that reveals calmness and composure, showing an inner strength;

[0077] Action and posture instructions: Maintain a good posture when standing, with your back straight and shoulders relaxed to show your elegance; walk with light steps and let your skirt sway gently in the wind to increase dynamic beauty; use your hands to gently adjust your skirt or belt to create a casual and elegant moment.

[0078] The aforementioned guiding words can provide a comprehensive image reference for generating the virtual model 13, ensuring that the generated virtual model 13 can show its design beauty and personal charm to the greatest extent when wearing the dress.

[0079] Step S104: Generate a virtual model 13 according to the guiding words. The aforementioned guiding words are rich enough to generate a virtual model 13 by the generating test AI. Alternatively, the most matching one is selected from the existing model library as the virtual model 13. Specifically, multiple virtual models 13 are manually established through 3D modeling software, and the descriptions of the gender, appearance, makeup, figure, facial expressions, and action postures of the virtual models 13 are recorded. Multiple such virtual models 13 and their descriptions constitute a model library. After obtaining the guiding words, the guiding words are compared with the virtual models 13 in the model library to obtain the most matching virtual model 13.

[0080] On the other hand, the method of generating the target model posture 16 according to the plurality of product images 15 and the user information includes:

[0081] identifying the poses of models in the plurality of product images 15, and grouping the product images 15 according to the poses of the models;

[0082] Obtaining the group with the largest number of product images 15, and obtaining the model posture of the corresponding group as the reference posture 20;

[0083] Reading a preset posture library 19, and calculating the similarity between each posture in the posture library 19 and the reference posture 20;

[0084] Obtaining a preferred style according to the user information, and calculating a matching degree between each posture and the preferred style;

[0085] The similarity is corrected according to the matching degree to obtain a corrected similarity, and the target model posture 16 adopts a posture whose corrected similarity with the reference posture is within a preset range.

[0086] 5 product images 15 adjacent to the image display position of the clothing item are obtained from the e-commerce platform 201, the model postures in the product images 15 are identified, and the groups are grouped according to the model postures. The group with the most product image 15 data is obtained as the group consisting of products e and f. The model postures of products e and f are used as reference postures 20.

[0087] Please see attached Figure 6 , which is a pre-established posture library 19. The posture library 19 can be established manually through 3D modeling software, or by manually collecting pictures that can show postures, cutting out pictures through AI processing, and then supplementing the postures through AI that can generate three-dimensional models. Multiple postures constitute the posture library 19.

[0088] The user information includes clothing feature descriptions extracted from clothing included in the user's historical shopping information. The clothing feature descriptions may include descriptions of wearing occasions, target audiences, design emotions, design styles, and may also include descriptions of other aspects. The user information is submitted to AI to obtain the user's preferred style. For example, the user's preferred style is elegant and comfortable.

[0089] Please see attached Figure 7 , respectively, the similarity of each posture in the posture library 19 with the reference posture 20, as well as the matching degree with the preferred style, finally obtain the modified similarity. The similarity between the posture and the reference posture 20 is determined according to the movement and position of the main limbs. Exemplarily, if the movement of the head is the same, the similarity of the head part is 1, if the movement of the upper arm is the same but the movement of the lower arm is significantly different, then the similarity of the arm part is 0.4, and so on. Finally, the weighted average of all local similarities is used as the similarity between the posture and the reference posture 20. The weighted average of the similarity and the matching degree is used as the modified similarity.

[0090] Exemplarily, the preset interval is [0.7, 0.9]. That is, a target model posture 16 that is too similar to the reference posture 20 will cause the target display image 12 to lack recognition, while being too dissimilar to the reference posture 20 will cause the target display image 12 to be in a minority category in the final product display page 18, which is inconsistent with the surrounding product images 15 and is easily ignored by the user. In combination with the user's preferred style of elegance and comfort, the final target model posture 16 is selected from the posture library 19.

[0091] On the other hand, the method of adjusting the posture of the virtual model 13 according to the target model posture 16 includes:

[0092] The posture of the virtual model 13 is adjusted so that the posture of the virtual model 13 matches the posture of the target model 16 at a certain viewing angle.

[0093] On the other hand, please see the attached Figure 8 The method for fusing the three-dimensional model of the target garment 11 with the virtual model 13 includes:

[0094] The three-dimensional model of the target garment 11 is meshed to obtain a mesh model, and calibration points 21 of the mesh model are extracted; the calibration points 21 include one or more of a neckline calibration point 21, a shoulder calibration point 21, an elbow calibration point 21, a cuff calibration point 21, a chest calibration point 21, a waist calibration point 21, a hip calibration point 21, a buttocks calibration point 21, a thigh calibration point 21, a knee calibration point 21, a trouser leg calibration point 21, and a skirt calibration point 21,

[0095] Aligning the calibration points 21 of the grid model with the virtual model 13 one by one, and reconstructing the grid model;

[0096] After all the calibration points 21 are aligned with the virtual model 13 , the fusion of the three-dimensional model of the target garment 11 and the virtual model 13 is completed.

[0097] The limb alignment points of the virtual model 13 are known, and it is possible to determine which calibration points 21 in the three-dimensional model of the target garment 11 need to be aligned with the virtual model 13 according to the style. For example, the shoulder calibration points 21 and waist calibration points 21 of an A-shaped suspender dress need to be aligned with the virtual model 13, while other calibration points 21 do not need to be aligned. In another example, for trousers, the waist calibration points 21 and the trouser leg calibration points 21 need to be aligned with the virtual model 13. The alignment of the trouser leg calibration points 21 with the virtual model 13 means that the positions of the two correspond, and it is not required to be completely overlapped, and the technology disclosed in the art can be used for processing.

[0098] On the other hand, the method for generating the scene 17 according to the plurality of product images 15 and the user information includes:

[0099] Identify the style of the product image 15 and obtain a preferred style according to the user information;

[0100] generating an extended description, wherein the extended description is a description of the style intensified in degree;

[0101] According to the extended description and the preferred style, a preset scene 17 is generated or matched, and the generation of the scene 17 is completed.

[0102] Please refer again to the attached Figure 1 , the style of product images 15 of products a and c is street scene, the style of product image 15 of product b is elegant, and the style of product images 15 of products e and f is simple. The extended description is obtained, and the extended description of product images 15 of products a and c is "vibrant city, free and uninhibited", the extended description of product image 15 of product b is "classical and elegant, exquisite and charming", and the extended description of product images 15 of products e and f is "minimal fresh, pure beauty".

[0103] The method of generating or matching the preset scene 17 according to the extended description and the preferred style includes:

[0104] Input all the extended descriptions and the preferred style to the AI, and instruct the AI ​​to generate a scene 17 description that best stands out from the existing extended descriptions and conforms to the preferred style;

[0105] Generate or match a preset scene 17 according to the scene 17 description.

[0106] For example, after the aforementioned extended description and preferred style are input to AI, the description of scene 17 is obtained as "classical garden, with elegant temperament displayed in the interweaving of light and shadow". Figure 2 Shown.

[0107] On the other hand, the method of creating a framing camera in the scene 17 according to the target model pose 16 includes:

[0108] Creating a framing camera, and adjusting the angle of the framing camera so that the posture of the virtual model 13 matches the posture of the target model 16 at the angle of the framing camera;

[0109] The size of the viewfinder frame of the viewfinder camera is adjusted so that the proportion of the virtual model 13 in the viewfinder frame exceeds a preset threshold. In 3D modeling software, the viewfinder camera is a basic concept and can be implemented using the technology disclosed in the art.

[0110] Through the technical solution of this embodiment, the generation of the target display picture 12 based on AI computer assistance can be adopted, so that the clothing products can be quickly and automatically worn by multiple models with different styles for display, and more abundant and more expressive target display pictures 12 can be obtained at a lower cost, thereby improving the efficiency of obtaining clothing display pictures and reducing costs; by combining the user information of the currently logged-in user, different target display pictures 12 can be generated for each user, providing personalized services, so that the clothing is more likely to attract the attention of the currently logged-in user; by obtaining the adjacent product pictures 15 in the product display page 18 in the e-commerce platform 201, targeted generation of postures and scenes 17 that are harmonious with the style and posture of the adjacent product pictures, but different from each other, further improves the possibility of the target clothing 11 being paid attention to by the currently logged-in user.

[0111] On the other hand, this specification provides an AI-based clothing product display system, please refer to the attached Fig. 9 ,include:

[0112] A first receiving module 100 receives a three-dimensional model of a target garment 11 and generates a virtual model 13;

[0113] The second receiving module 200 receives a plurality of product images 15 of other clothing in the display page and information of a currently logged-in user, generates a target model posture 16 according to the plurality of product images 15 and the user information, and adjusts the posture of the virtual model 13 according to the target model posture 16;

[0114] A fusion module 300 is configured to fuse the three-dimensional model of the target garment 11 with the virtual model 13 to obtain a fused model 14;

[0115] A generating module 400 generates a scene 17 according to the plurality of product images 15 and the user information, and places the fused model 14 into the scene 17;

[0116] The framing module 600 creates a framing camera in the scene 17 according to the target model posture 16, and obtains a target display picture 12 according to the framing camera.

[0117] See also Fig.10 A schematic diagram of the structure of an electronic device provided in an embodiment of this specification is shown.

[0118] like Fig.10 As shown, the electronic device 1100 may include: at least one processor 1101, at least one network interface 1104, a user interface 1103, a memory 1105 and at least one communication bus 1102. Among them, the communication bus 1102 can be used to realize the connection and communication of the above-mentioned components. Among them, the user interface 1103 may include a button, and the optional user interface may also include a standard wired interface and a wireless interface. Among them, the network interface 1104 may include but is not limited to a Bluetooth module, an NFC module, a Wi-Fi module, etc. Among them, the processor 1101 may include one or more processing cores. The processor 1101 uses various interfaces and lines to connect the various parts of the entire electronic device 1100, and executes various functions of the routing device and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 1105, and calling data stored in the memory 1105. Optionally, the processor 1101 can be implemented in at least one hardware form of DSP, FPGA, and PLA. The processor 1101 can integrate one or a combination of CPU, GPU, modem, etc. Among them, the CPU mainly handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content that needs to be displayed on the display; and the modem is used to handle wireless communications.

[0119] It is understandable that the above-mentioned modem may not be integrated into the processor 1101, but may be implemented by a separate chip.

[0120] Among them, the memory 1105 may include RAM or ROM. Optionally, the memory 1105 includes a non-transitory computer-readable medium. The memory 1105 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 1105 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 1105 may also be at least one storage device located away from the aforementioned processor 1101. The memory 1105 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application. The processor 1101 may be used to call the application stored in the memory 1105 and execute the methods in the above-mentioned multiple embodiments.

[0121] The embodiments of this specification also provide a computer-readable storage medium, which stores instructions, and when the instructions are executed on a computer or a processor, the computer or the processor executes the multiple steps in the above embodiments. If the components of the above electronic device are implemented in the form of software functional units and sold or used as independent products, they can be stored in the computer-readable storage medium.

[0122] The embodiments of this specification also provide a computer program product, including a computer program, which implements multiple steps in the above embodiments when executed by a processor.

[0123] In the absence of conflict, the technical features in this embodiment and implementation scheme can be combined arbitrarily.

[0124] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes a plurality of computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of this specification is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions may be transmitted from a website site, a computer, a server or a data center to another website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server 202, a data center, etc. that includes multiple available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)), etc.

[0125] When implemented by hardware or firmware, the aforementioned method flow is programmed into the hardware circuit to obtain the corresponding hardware circuit structure and realize the corresponding function. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit, and its logic function is determined by the user programming the device. The designer programs by himself to "integrate" a digital system on a PLD, without asking a chip manufacturer to design and make a dedicated integrated circuit chip. Moreover, nowadays, instead of manually making integrated circuit chips, this programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when writing program development, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL), and HDL is not just one, but many. Those skilled in the art should also be aware that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit to easily obtain the hardware circuit that implements the logic method flow.

[0126] The embodiments described above are merely preferred embodiments of this specification and are not intended to limit the scope of this specification. Without departing from the design spirit of this specification, various modifications and improvements made to the technical solutions of this specification by ordinary technicians in this field should fall within the scope of protection determined by the claims of this specification.

Claims

1. A clothing product display method based on AI, characterized in that: Includes steps: Receiving a three-dimensional model of a target garment and generating a virtual model; Receiving a plurality of other clothing product images and currently logged-in user information on a display page, generating a target model posture according to the plurality of product images and the user information, and adjusting the posture of the virtual model according to the target model posture; Fusing the three-dimensional model of the target garment with the virtual model to obtain a fused model; Generate a scene according to the plurality of product images and the user information, and place the fused model into the scene; Creating a framing camera in the scene according to the target model's posture, and obtaining a target display picture according to the framing camera; The method for generating a target model posture according to the plurality of product images and the user information comprises: identifying the poses of models in a plurality of the product images, and grouping the product images according to the poses of the models; Obtain the group with the largest number of product images, and obtain the model posture of the corresponding group as a reference posture; Reading a preset posture library, and calculating the similarity between each posture in the posture library and the reference posture; Obtaining a preferred style according to the user information, and calculating a matching degree between each posture and the preferred style; The similarity is corrected according to the matching degree to obtain a corrected similarity, and the target model posture adopts a posture whose corrected similarity with the reference model posture is within a preset range.

2. The AI-based clothing product display method according to claim 1, characterized in that: Methods for generating avatars include: Obtaining clothing information according to the three-dimensional model of the target clothing, wherein the clothing information includes style, size, color, pattern and design elements; Obtaining a description of the clothing according to the clothing information, wherein the description includes a description of wearing occasions, target audiences, design emotions, and design styles; Generating guidance words according to the description, the guidance words including gender guidance words, appearance guidance words, makeup guidance words, body guidance words, expression and posture guidance words, and action and posture guidance words; A virtual model is generated according to the guiding words.

3. The AI-based clothing product display method according to claim 1 or 2, characterized in that: The method for adjusting the posture of the virtual model according to the posture of the target model includes: Adjusting the posture of the virtual model so that the posture of the virtual model matches the posture of the target model at a certain viewing angle; The method for fusing the three-dimensional model of the target garment with the virtual model comprises: The three-dimensional model of the target garment is meshed to obtain a mesh model, and calibration points of the mesh model are extracted; the calibration points include one or more of a neckline calibration point, a shoulder calibration point, an elbow calibration point, a cuff calibration point, a chest calibration point, a waist calibration point, a hip calibration point, a buttocks calibration point, a thigh calibration point, a knee calibration point, a trouser leg calibration point, and a skirt calibration point, Aligning the calibration points of the grid model with the virtual model one by one, and reconstructing the grid model; After all the calibration points are aligned with the virtual model, the fusion of the three-dimensional model of the target garment and the virtual model is completed.

4. The AI-based clothing product display method according to claim 1 or 2, characterized in that: The method for generating a scene according to the plurality of product images and the user information includes: Identify the style of the product image and obtain the preferred style according to the user information; generating an extended description, wherein the extended description is a description of the style intensified in degree; According to the extended description and the preferred style, a pre-set scene is generated or matched to complete the scene generation.

5. The AI-based clothing product display method according to claim 3, characterized in that: The method for creating a framing camera in the scene according to the target model's posture comprises: Creating a framing camera, and adjusting the angle of the framing camera so that the posture of the virtual model matches the posture of the target model at the angle of the framing camera; The size of the viewfinder frame of the viewfinder camera is adjusted so that the proportion of the virtual model in the viewfinder frame exceeds a preset threshold.

6. An AI-based clothing product display system, characterized in that: include: A first receiving module receives a three-dimensional model of a target garment and generates a virtual model; A second receiving module receives a plurality of product pictures of other clothing in the display page and information of a currently logged-in user, generates a target model posture according to the plurality of product pictures and the user information, and adjusts the posture of the virtual model according to the target model posture; A fusion module, which fuses the three-dimensional model of the target garment with the virtual model to obtain a fused model; A generating module, generating a scene according to the plurality of product images and the user information, and placing the fused model into the scene; A framing module, which creates a framing camera in the scene according to the target model's posture, and obtains a target display picture according to the framing camera; The method for generating a target model posture according to the plurality of product images and the user information comprises: identifying the poses of models in a plurality of the product images, and grouping the product images according to the poses of the models; Obtain the group with the largest number of product images, and obtain the model posture of the corresponding group as a reference posture; Reading a preset posture library, and calculating the similarity between each posture in the posture library and the reference posture; Obtaining a preferred style according to the user information, and calculating a matching degree between each posture and the preferred style; The similarity is corrected according to the matching degree to obtain a corrected similarity, and the target model posture adopts a posture whose corrected similarity with the reference model posture is within a preset range.

7. An electronic device, characterized in that including a processor and a memory; The processor is connected to the memory; The memory is used to store executable program code; The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

9. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images