System for providing immersive marketing content through generative ai-based spatial computing web service

Abstract

The present invention relates to a system for providing immersive marketing content through a generative AI-based spatial computing web service, which enables a brand unique customized style to be generated, on the basis of generative AI, for a keyword input through a user terminal, regardless of the size of a brand or the presence or absence of professional personnel. In addition, the present invention has the advantage of enabling a user to feel a high-quality virtual space experience with little effort by generating a desired image with only a simple command statement and providing a service in a web environment that does not require installation.

Description

An immersive marketing content delivery system using generative AI-based spatial computing web services.

[0001] The present invention relates to an immersive marketing content provision system through a generative AI-based spatial computing web service, and more specifically, to an immersive marketing content provision system through a generative AI-based spatial computing web service that enables the creation of a brand's unique customized style based on generative AI for keywords input through a user terminal, regardless of the brand's size or the presence of specialized personnel.

[0002] In recent years, the e-commerce landscape has been constantly evolving to capture customer attention. Rapid advancements in technology capable of displaying 3D content in web and mobile applications have laid the foundation for a three-dimensional experience for customers, enabling them to experience brands and products. Moving beyond the standardized format of listing product images and prices, it's now possible to present products and brand stories in a three-dimensional virtual environment, actively engaging customers and allowing them to engage with the experience more actively.

[0003] However, producing and distributing this type of content requires a variety of expertise. Because it requires significant human, time, and financial resources, including 3D design, engineering, and development, it has been difficult for smaller brands to attempt.

[0004] The present invention is intended to solve the aforementioned problems, and provides an immersive marketing content provision system through a generative AI-based spatial computing web service that enables the creation of a brand's unique customized style based on generative AI for keywords input through a user terminal, regardless of the brand's size or the presence of specialized personnel.

[0005] A system (100) for providing immersive marketing content through a generative AI-based spatial computing web service according to one embodiment of the present invention may include a user account management unit (110) for creating and managing a user account, an image generation unit (120) for creating a generative AI-based virtual content image corresponding to a text keyword input from a user terminal linked to the created user account, based on the text keyword, a sample style recommendation unit (130) for storing a plurality of sample styles that can be reflected in the created virtual content image and recommending a sample style corresponding to the virtual content image created through the image generation unit (120) based on a recommendation algorithm, an image rendering unit (140) for performing image rendering based on the result of reflecting a sample style selected through the user terminal to the previously created virtual content image, and a result provision unit (150) for converting the format of virtual content data for which rendering has been completed into a format usable in a third-party web service environment and providing the converted data to a cloud server linked to a third-party web service.

[0006] In one embodiment, the user account management unit (110) may create a corresponding user account for each user terminal, and may create the user account by reflecting SNS account information previously created based on the user terminal.

[0007] In one embodiment, the user account management unit (110) stores parameter information, style information, and text keywords input through a user terminal in association with a user account, generates a result asset based on the stored parameter information, style information, and text keywords, and stores a virtual content image generated through the image generation unit (120) and virtual content data rendered through the image rendering unit (140) in association with a user account.

[0008] In one embodiment, the image generation unit (120) provides a dedicated application including a TTS AI prompt input module capable of inputting text keywords, and can generate images for spatial images and marketing content based on text keywords input through the TTS AI prompt input module.

[0009] In one embodiment, the image generation unit (120) generates images for spatial images and marketing content based on input text keywords, and may use a dataset that has learned multiple image patterns and features.

[0010] According to the present invention, regardless of the size of the brand or the presence of specialized personnel, it has the advantage of being able to create a brand-specific customized style based on generative AI for keywords input through a user terminal.

[0011] In particular, according to the present invention, a desired image is created with only a simple command and a service is provided in a web environment that does not require installation, thereby providing an advantage of allowing a user to experience a high-quality virtual space experience with little effort.

[0012] In addition, according to the present invention, the user can personalize and set the virtual space in the direction he / she needs through functions such as removing background images, adding multimedia (video, music, images, 3D assets, link connection, etc.), thereby having the advantage of allowing the user to edit freely and flexibly according to his / her taste or inclination.

[0013] In addition, according to the present invention, new technology integration such as virtual humans or immersive ads can be integrated with existing services in a simple add-on manner, and when providing images and information reflecting brand identity, it has the advantage of being able to generate a unique brand style by precisely learning a generative AI model.

[0014] FIG. 1 is a drawing schematically showing the overall configuration of an immersive marketing content provision system (100) through a generative AI-based spatial computing web service according to one embodiment of the present invention.

[0015] FIG. 2 is a diagram illustrating a concept of creating and providing an optimal customized style using an immersive marketing content provision system (100) through a generative AI-based spatial computing web service illustrated in FIG. 1.

[0016] 100: An immersive marketing content delivery system using generative AI-based spatial computing web services.

[0017] 110: User Account Management

[0018] 120: Image generation unit

[0019] 130: Sample Style Recommendation

[0020] 140: Image rendering unit

[0021] 150: Results Provider

[0022] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings so that those skilled in the art can easily implement the present invention. The present invention may be implemented in various different forms and is not limited to the embodiments described herein.

[0023] In order to clearly explain the present invention, parts that are not related to the description are omitted, and the same reference numerals are used for identical or similar components throughout the specification.

[0024] In addition, in various embodiments, components having the same configuration are described only in representative embodiments using the same symbols, and in other embodiments, only configurations different from the representative embodiments are described.

[0025] Throughout the specification, when a part is said to be "connected" to another part, this includes not only "directly connected" but also "indirectly connected" with other elements intervening. Furthermore, when a part is said to "include" a component, this may mean that the other component is included, rather than excluded, unless otherwise specifically stated.

[0026]

[0027] FIG. 1 is a drawing schematically showing the overall configuration of an immersive marketing content provision system (100) through a generative AI-based spatial computing web service according to one embodiment of the present invention.

[0028] Referring to FIG. 1, an immersive marketing content provision system (100) through a generative AI-based spatial computing web service according to one embodiment of the present invention can be largely configured to include a user account management unit (110), an image generation unit (120), a sample style recommendation unit (130), an image rendering unit (140), and a result provision unit (150).

[0029] First, the user account management unit (110) is responsible for creating and managing user accounts. More specifically, the user account management unit (110) ensures that a corresponding user account is created for each user terminal. For example, if the user is a general user or a business, one account is created for each general user or business.

[0030] Additionally, the user account management unit (110) can retrieve information on other SNS accounts previously created based on the user terminal and then reflect this information in the user account to create the account.

[0031] To this end, the user account management unit (110) may, during the process of creating a user account, send a notification to the user terminal requesting consent to inquire whether there are any SNS accounts registered based on the user terminal's phone number. If a consent feedback for inquiry is received from the user terminal, the user account management unit (110) will inquire whether there are any SNS accounts registered based on the user terminal's phone number.

[0032] If there are SNS accounts registered based on the phone number of the user terminal, the user account management unit (110) obtains and stores keywords, parameters, styles, etc. entered or searched by the user in the SNS account. The information stored in this way can be utilized in the image generation unit (120) described below.

[0033] In one embodiment, when SNS accounts linked to a user terminal are not searched, the user account management unit (110) may present sample keywords to the user terminal to determine the user's taste, tendency, or style, and then determine the user's taste, tendency, or style based on the keywords selected through the user terminal.

[0034] In addition, the user account management unit (110) can store parameter information, style information, and text keywords input through a user terminal in connection with a user account, and generate result assets based on the stored parameter information, style information, and text keywords.

[0035] In addition, the user account management unit (110) can store virtual content images generated through the image generation unit (120) described below and virtual content data that has been rendered through the image rendering unit (140) described below in conjunction with the user account. The user account management unit (110) can separately store the generation and storage history, details, and history of virtual content images and virtual content data, and can then proceed with image learning after collecting metadata about the results (virtual content data that has been rendered) that the user is satisfied with. The results of the image learning process can be stored in conjunction with the corresponding user account, and when converted into a format usable in a third-party web service environment through the result provision unit (150) described below, can be linked to a cloud server linked to the third-party web service through SSO linkage.

[0036]

[0037] The image generation unit (120) plays a role in generating a generative AI-based virtual content image corresponding to a text keyword input from a user terminal linked to a generated user account created through the user account management unit (110).

[0038] More specifically, the image generation unit (120) can generate a virtual content image based on the tastes, tendencies, styles, parameters, etc. previously identified through the user account management unit (110). To this end, the image generation unit (120) provides a dedicated application including a TTS AI prompt input module capable of inputting text keywords, and can generate images for spatial images and marketing content based on the text keywords input through the TTS AI prompt input module. At this time, the image generation unit (120) can analyze the text keywords input through the user terminal using natural language processing (NLP) technology and perform preprocessing accordingly. The preprocessed data is input to the AI ​​model GANs (generative adversarial network) for generating spatial images.

[0039] In addition, the image generation unit (120) provides 7 major classification styles (Realistic / Ilustrative / Fantasy / Futuristic / Surreal / Render / ArtTechniques) as a basic feature and 38 sub-classification styles in detail. In addition, the image generation unit (120) can learn continuously generated images and periodically update new styles. In addition, the image generation unit (120) can refer to the 'Spatial' dataset that has learned various image patterns and characteristics in the process of creating spatial images and images for marketing content based on input text keywords.

[0040] In addition, the image generation unit (120) can provide convenience for easier service use by providing sample result assets and templates so that the user can predict the result value in advance. The image generated through the image generation unit (120) automatically goes through a verification process. In this process, the image generation unit (120) checks whether the generated image matches the text keyword entered through the user terminal and satisfies the quality standard. The generated image can be generated in high resolution without any sense of incongruity in the process of being provided to a cloud server linked to a third-party web service through the result provision unit (150), and by providing an AI editing tool, the user can edit the generated spatial image, such as adding multimedia or inserting graphic elements.

[0041]

[0042] The sample style recommendation unit (130) stores a plurality of sample styles that can be reflected in the virtual content image generated through the image generation unit (120), and can recommend a sample style corresponding to the virtual content image generated through the image generation unit (120) based on a recommendation algorithm. More specifically, the sample style recommendation unit (130) can provide a variety of pre-standardized plurality of sample styles because the user may not know which text keyword to input through the TTS AI prompt input module.

[0043] To this end, the sample style recommendation unit (130) first provides a good quality sample through a template, thereby reducing the user's learning curve, and helps to easily select a style by checking the sample results of each style as a pop-up when each prompt is input. When a prompt is input, a recommendation algorithm is provided to recommend keywords that match the input prompt so that the user can easily select them, and with the guidance of an AI human, it can provide results that meet the user's needs through a storyline, such as template selection, style sample confirmation, and prompt recommendation.

[0044]

[0045] The image rendering unit (140) performs image rendering based on the result of reflecting the nautical style on the previously generated virtual content image when a specific style is selected through the user terminal from among the sample styles recommended through the sample style recommendation unit (130). More specifically, the image rendering unit (140) renders the resulting asset based on the server-based GPU according to the text keywords, styles, and parameters input through the user terminal. Here, the server-based GPU can perform rendering based on the A10 Nvidia GPU.

[0046] In addition, the image rendering unit (140) utilizes an AI acceleration algorithm to maximize GPU utilization and can achieve an image extraction effect that is approximately twice that of general rendering. Here, the AI ​​acceleration algorithm refers to an AI learning and inference-specific algorithm that dramatically reduces the time required for image inference. In addition, the image rendering unit (140) also utilizes the vGPU technology of the GPU platform to cost-effectively utilize the infrastructure, and all servers are cloud-native to ensure both stability and scalability.

[0047]

[0048] The result providing unit (150) can convert the format of the virtual content data rendered through the image rendering unit (140) into a format usable in a third-party web service environment and then provide it to a cloud server linked to the third-party web service. More specifically, the result providing unit (150) can provide integrated data so that the extracted result images and result assets can be easily integrated into a third-party web service. In addition, the result providing unit (150) can provide integrated data using HTML-based code, and at this time, Java, Python, and Node.js-based API integration can be enabled.

[0049] In addition, in one embodiment, the result provider (150) can provide a productive panoramic 3D background and content to users providing game / metaverse services by providing integration based on Unity and Unreal Engine. Here, the result provider (150) provides two types of integration models: one that integrates only 3D images and the other that integrates VoidX services, and can increase the usability of VoidX services depending on the needs of the user.

[0050]

[0051] Referring to FIG. 2, the immersive marketing content provision system (100) through a generative AI-based spatial computing web service according to the present invention can generate a brand-specific customized style and provide marketing content through various steps.

[0052] More specifically, in the user input collection step, the user can create spatial images and marketing content and input necessary instructions in text through the TTS AI prompt input module provided to the user terminal.

[0053] Next, in the data preprocessing stage, the instructions entered through the user terminal are analyzed and processed by the image generation unit (120), and in this stage, natural language processing (NLP) technology is used to understand the user's instructions and preprocess them into necessary data.

[0054] Next, in the AI ​​model activation stage, the preprocessed data from the data preprocessing stage is input into an AI model called GANs (generative adversarial networks) for spatial image generation. In the image generation stage, the AI ​​model generates new spatial images based on the input data. This process references the "Spatial" dataset, which internally learns various image patterns and features.

[0055] Next, in the result verification stage, the pre-generated images (virtual content images) are automatically verified. In this stage, AI verifies that the images match the user's instructions and meet quality standards.

[0056] Next, in the image rendering step, the generated virtual content image is provided to the user terminal, and in the image editing step, an editing tool within the web service interface is provided to the user terminal, through which the user can edit the generated spatial image or virtual content image, such as adding multimedia or inserting graphic elements.

[0057] Next, in the step of reflecting the edited content, all edits made through the user terminal are reflected in real time on the space image or virtual content image, and in the step of requesting publishing, when the user requests publishing through the user terminal for an image for which editing has been completed, an HTML file is generated through the code generator.

[0058]

[0059] Meanwhile, the voidX AI model and voidX Studio used in the present invention will be examined in more detail as follows.

[0060] The voidX AI model specializes in generating spatial images through precision learning of a large-scale visual model (LVM). The voidX AI model excels at providing detailed descriptions of spatial backgrounds by expressing local coordinates within an image in the form of natural language numbers.

[0061] The voidX AI model builds a spatial dataset to learn spatial knowledge in various styles. The voidX AI model includes images, graphic designs, and text data covering various forms and styles, including outdoor, indoor, architectural, natural, science fiction, and animation. Additionally, it may include negative data that can degrade the quality of the results or cause object hallucinations.

[0062] Furthermore, the voidX AI model can analyze the 'Spatial' dataset to recognize specific patterns, styles, relationships between components, etc., and generate new styles. The voidX AI model can then fine-tune and optimize the results to meet specific style requirements. Furthermore, the voidX AI model can understand the learned data and collected user instructions through natural language processing, and generate new spatial images based on them. This process is handled by generative adversarial networks (GANs).

[0063]

[0064] voidX Studio is a browser-based generative AI studio that allows users to generate 360-degree panoramic spatial and 3D rendered spatial images using simple text prompts. This technology is optimized for creating immersive content and can be used in a variety of fields, including design, games, immersive web, virtual studios, and marketing.

[0065] One of voidX Studio's key features is its "Style Selection" feature, which allows users to select from several available creation styles and generate spatial images using that style, ensuring the most appropriate result for their intended use and industry. Furthermore, editing features, including the ability to insert additional graphic elements and multimedia, allow users to customize as needed. After editing, users can publish their work in various formats, including HTML code, EXR, HDR, and cube maps.

[0066] Additionally, voidX Studio also provides user account options to allow users to access their creation history, allowing them to review their previous creations, make further edits, or re-export the results.

[0067]

[0068] Although the present invention has been described above with reference to preferred embodiments thereof, it will be understood by those skilled in the art that various modifications and changes may be made to the present invention without departing from the spirit and scope of the present invention as set forth in the claims below.

Claims

1. User account management unit (110) that creates and manages user accounts; An image generation unit (120) that generates a generative AI-based virtual content image corresponding to a text keyword input from a user terminal linked to a generated user account; A sample style recommendation unit (130) that stores multiple sample styles that can be reflected in the generated virtual content image and recommends a sample style corresponding to the virtual content image generated through the image generation unit (120) based on a recommendation algorithm; An image rendering unit (140) that performs image rendering based on the result of reflecting a sample style selected through a user terminal into a pre-generated virtual content image; and It includes a result providing unit (150) that converts the format of the virtual content data for which rendering is completed into a format usable in a third-party web service environment and then provides the converted data to a cloud server linked to a third-party web service. The above user account management unit (110) creates a corresponding user account for each user terminal, and creates a user account by reflecting the SNS account information previously created based on the user terminal. The user account management unit (110) above, in the process of creating the user account, sends a notification to the user terminal to obtain consent to check whether there are SNS accounts registered based on the phone number of the user terminal, and when a feedback of consent to check is received, checks whether there are SNS accounts registered based on the phone number of the user terminal, If the SNS accounts linked to the user terminal are not found, The above user account management unit (110) is characterized in that it presents sample keywords to identify the user's taste, tendency or style, and then identifies the user's taste, tendency or style based on keywords selected through the user terminal. An immersive marketing content delivery system using generative AI-based spatial computing web services.

2. In paragraph 1, The above user account management unit (110) Parameter information, style information, and text keywords entered through the user terminal are stored in connection with the user account, and result assets are created based on the stored parameter information, style information, and text keywords. It is characterized in that the virtual content image generated through the image generation unit (120) and the virtual content data rendered through the image rendering unit (140) are stored in connection with a user account. An immersive marketing content delivery system using generative AI-based spatial computing web services.

3. In paragraph 1, The above image generation unit (120) Providing a dedicated application including a TTS AI prompt input module capable of inputting text keywords, characterized in that it generates images for spatial images and marketing content based on text keywords input through the TTS AI prompt input module. An immersive marketing content delivery system using generative AI-based spatial computing web services.

4. In paragraph 3, The above image generation unit (120) It is characterized by generating images for spatial images and marketing content based on input text keywords, using a dataset that has learned multiple image patterns and features. An immersive marketing content delivery system using generative AI-based spatial computing web services.

Images