Electronic device and content generation method of electronic device

Abstract

An electronic device, according to one embodiment of the present disclosure, may: display a first image through a display; receive a first user input for executing an editing function of the first image; and generate, on the basis of the presence of personal metadata related to the first image, a second image obtained by editing at least a part of the first image by using an artificial intelligence model trained on the basis of the personal metadata. Various other embodiments understood through the specification are also possible.

Description

Electronic device and method for creating content of the electronic device

[0001] The embodiments disclosed in this document relate to a technique for creating new content by editing content.

[0002] An electronic device may acquire images using a camera or provide editing functions for acquired images. For example, the electronic device may generate a new image based on user input, extract at least a portion of an image, or add new objects to an image. Images acquired by the electronic device may include metadata related to the image. For example, metadata related to the image may include the date (capture) of the image, the time of acquisition, information related to camera settings, and / or information representing the image (e.g., tags). For example, metadata related to the image may be used to recognize, search, and / or classify the image.

[0003] The information described above may be provided as related art for the purpose of aiding understanding of the present disclosure. No claim or determination is made as to whether any of the foregoing may be applied as prior art related to the present disclosure.

[0004] An electronic device according to one embodiment disclosed in this document may include a display, a memory for storing instructions, and at least one processor. When the instructions are executed individually or collectively by the at least one processor, the electronic device may display a first image through the display, receive a first user input for executing an editing function of the first image, and generate a second image in which at least a portion of the first image is edited using an artificial intelligence model learned based on said personal metadata, based on the existence of said personal metadata associated with the first image.

[0005] Additionally, a method according to one embodiment disclosed in this document may include an operation of displaying a first image, an operation of receiving a first user input for executing an editing function of the first image, and an operation of generating a second image in which at least a portion of the first image is edited using an artificial intelligence model learned based on said personal metadata, based on the existence of said personal metadata associated with the first image.

[0006] Additionally, a storage medium according to one embodiment disclosed in this document may store a program and / or instructions that, when executed individually or collectively by at least one processor of an electronic device, cause the electronic device to display a first image, receive a first user input for executing an editing function of the first image, and, based on the existence of personal metadata associated with the first image, generate a second image in which at least a portion of the first image is edited using an artificial intelligence model learned based on said personal metadata.

[0007] FIG. 1 is a block diagram of an electronic device according to one embodiment.

[0008] FIG. 2 is a block diagram of an electronic device according to one embodiment.

[0009] FIG. 3 is a diagram illustrating the content creation operation of an electronic device according to one embodiment.

[0010] FIG. 4 is a diagram illustrating the content creation operation of an electronic device according to one embodiment.

[0011] FIG. 5 is a diagram illustrating the content creation operation of an electronic device according to one embodiment.

[0012] FIG. 6 is a diagram illustrating the content creation operation of an electronic device according to one embodiment.

[0013] FIG. 7 is a flowchart of a method for creating content of an electronic device according to one embodiment.

[0014] FIG. 8a is a flowchart of a method for creating content of an electronic device according to one embodiment.

[0015] FIG. 8b is a flowchart of a method for creating content of an electronic device according to one embodiment.

[0016] FIG. 9 is a block diagram of an exemplary electronic device capable of performing the operations described in this document.

[0017] FIG. 10 illustrates a generative artificial intelligence system according to one embodiment.

[0018] In relation to the description of the drawings, the same or similar reference numerals may be used for identical or similar components.

[0019] FIG. 1 is a block diagram of an electronic device according to one embodiment.

[0020] According to one embodiment, an electronic device (100) (e.g., electronic device (200; 900; 1000)) may include a display (110) (e.g., user interface (240), display (940), or user question / response interface (1010)), a memory (120) (e.g., database (217), common metadata (218), private metadata (219), memory (920), or database (1030)), and a processor (130) (e.g., content analysis module (210), processor (910), or artificial intelligence framework (1020)).

[0021] According to one embodiment, the display (110) can visually display information. For example, the display (110) can display at least one image (e.g., original image, edited image, and / or image generated by the electronic device (100). The display (110) can display information related to at least one image (e.g., metadata related to the image (e.g., common metadata and / or private metadata)). According to one embodiment, the display (110) can be integrally formed with an input device (e.g., a touch panel). For example, the display (110) may include a touchscreen display (110).

[0022] According to one embodiment, the memory (120) may store instructions that control the operation of the electronic device (100) when executed individually or collectively by at least one processor (130). For example, the instructions may be stored in one memory (120) or multiple memories (120). The memory (120) may store at least temporarily information and / or data related to the operations of the electronic device (100). The memory (120) may store at least one image (e.g., original image, edited image, and / or image generated by the electronic device (100)). The memory (120) may store information related to the image (e.g., metadata related to the image). For example, the metadata of the image may include common metadata (which may be referred to in this disclosure as 'non-context-based metadata' or 'common tag') and private metadata (which may be referred to in this disclosure as 'context-based metadata' or 'private tag'). Common metadata may include, but is not limited to, the time of image acquisition, the location of acquisition, information on objects included in the image, the name of the image, storage space, and / or capacity. Personal metadata may include, but is not limited to, at least one of: information related to an application in use at the time of acquiring the first image, the acquisition path of the first image, the sharing history of the first image, the conversation history related to the first image, images similar to the first image, user emotion information related to the first image, information about functions and / or applications related to the first image, information related to objects included in the first image, context information related to common metadata, or user-related information stored in the electronic device (100) (e.g., contact information, email information, setting information of the electronic device (100)).

[0023] For example, memory (120) may store at least one artificial intelligence model (e.g., a generative AI model (220; 1050) or an LVM (230)). The artificial intelligence model may be trained based on at least one of user input, the original image, an image edited from the original image, a reference image input by the user, or metadata related to the original image (e.g., tags included in the metadata). For example, the artificial intelligence model may include a generative AI model including a language model (e.g., large language model (LLM)), a vision model (e.g., large vision model (LVM)), or a multimodal model (e.g., large multi-modal model (LMM)), but is not limited thereto.

[0024] For example, the memory (120) may include at least one learning data. For example, the memory (120) may include at least one general learned data including an image learned based on specified common metadata using an AI model and / or at least one personal learned data including an image learned based on specified personal metadata using an AI model.

[0025] According to one embodiment, a processor (130) may control the operations of an electronic device (100) by executing instructions stored in memory (120) individually or collectively. For example, the operations described in this disclosure as being performed by the 'processor (130)' may be understood as being performed individually or collectively by at least one processor (130). For example, at least one processor (130) may control each of the operations of the electronic device (100) described below independently or collectively. According to one embodiment, at least one processor (130) may include circuits such as a central processing unit (CPU), a micro processor unit (MPU), an application processor (AP), a communication processor (CP), a System On Chip (SoC), and / or an Integrated Circuit (IC).

[0026] According to one embodiment, the processor (130) may display a first image through a display (110). For example, the first image may include an image stored in an electronic device (100) and / or an image obtained from an external source (e.g., an external device, an external server, and / or an external network). The first image may include a photograph, a drawing, an image, and / or a video (e.g., at least one frame of a video). For example, the processor (130) may display the first image through an image viewer application.

[0027] According to one embodiment, the processor (130) may receive a first user input for executing an editing function of a first image. For example, the processor (130) may enable an image editing function provided by a viewer application based on the first user input. For example, the processor (130) may receive a first user input for selecting a user interface (e.g., an indication, a button, a menu, and / or an affordance item) for executing an image editing function.

[0028] According to one embodiment, the processor (130) may generate a second image in which at least a portion of the first image is edited using an artificial intelligence (AI) model learned based on personal metadata, based on the existence of personal metadata associated with the first image. For example, the processor (130) may determine whether personal metadata and / or personal learning data associated with the first image exist. The electronic device may generate a second image in which at least a portion of the first image is edited based on personal learning data, using an AI model.

[0029] For example, while displaying the first image, the processor (130) may display, through the display (110), a user interface (e.g., an indication, a button, a menu, and / or affordance item) indicating that the first image is editable based on personal metadata along with the first image. For example, the processor (130) may display, through the display (110), a user interface indicating whether each image is editable based on personal metadata along with the images included in the electronic device. For example, the processor (130) may generate a second image based on user input selecting the user interface indicating that the first image is editable based on personal metadata. The processor (130) may also generate the second image without additional user input when the image editing function is enabled.

[0030] For example, the processor (130) may use a trained AI model to generate a second image in which at least a portion of the first image is edited based on personal metadata. For example, editing of the first image may include in-painting, out-painting, changing the background of the first image, adding objects to the first image, deleting at least a portion of the first image, replacing or changing objects in the first image, and / or applying visual effects to the first image.

[0031] For example, the processor (130) can generate a second image by adding another person related to the user to the first image based on personal metadata. The processor (130) can recognize the relationships of the people included in the first image (e.g., people within a designated group of contacts) based on personal metadata and add a person not included in the first image (e.g., a person within a designated group of contacts not included in the first image) to the first image. The processor (130) can generate a second image by adding an object to the first image representing the person with whom the user was having a call or conversation (e.g., a conversation through a chat window) via an electronic device at the time the first image was acquired. The processor (130) can add an object to the first image representing the person with whom the conversation related to the first image and / or the person with whom the first image was shared, based on information regarding the conversation history and / or the sharing history of the first image included in the personal metadata.

[0032] For example, the processor (130) may add a background to the first image corresponding to the location of the electronic device (user) at the time of acquiring the first image, or change the background included in the first image to a background corresponding to the location of the electronic device (user) at the time of acquiring the first image. The processor (130) may change the background of the first image based on application and / or function information that the user was using at the time of acquiring the first image. For example, the processor (130) may add a background representing a gym to the first image or modify the background of the first image to a background representing a gym based on personal metadata indicating that the user was using a health (or exercise) related application or was at a gym at the time of acquiring the first image.

[0033] For example, the processor (130) may add an object representing a travel route to the first image based on the determination that the user was traveling when the first image was acquired based on personal metadata, or add an object and / or background related to the travel destination.

[0034] For example, the processor (130) may add objects and / or visual effects (e.g., applying specific weather effects) to the first image based on personal metadata that represent the user's emotional state when the first image is acquired and / or when the first image is edited.

[0035] For example, the processor (130) can generate a second image modified based on private metadata (and / or common metadata) so that the weather (e.g., clear, rain, snow) or time (e.g., day or night) represented by the first image appears differently.

[0036] For example, the processor (130) may display at least one indication indicating at least one available editing method through the display (110). For example, the processor (130) may determine the availability of each of the at least one editing method based on the degree of learning of the AI ​​model for each of the at least one editing method. The processor (130) may display a first image and at least one indication through the display (110). For example, the processor (130) may receive user input selecting at least one of the at least one indications. The processor (130) may generate a second image based on the editing method corresponding to the selected at least one indication. For example, the at least one editing method may include at least one of in-painting, out-painting, an editing method for adding at least one new object, an editing method for replacing a background or at least one object, an editing method for deleting a background or at least one object, an editing method for applying at least one visual effect, or an editing method for adding at least one item.

[0037] For example, the processor (130) can generate at least one image by editing at least a portion of the first image based on personal metadata using a trained AI model. The processor (130) can provide the generated at least one image as a recommended image. The processor (130) can determine the recommended image selected based on user input as the second image.

[0038] For example, the processor (130) may provide a recommendation item that includes information representing at least one piece of personal learning data. For example, the recommendation item may include an image representing the personal learning data and / or text describing the personal learning data. Based on user input selecting at least one of the recommendation items, the processor (130) may generate a second image by editing at least a portion of the first image based on the personal learning data corresponding to the selected recommendation item.

[0039] For example, when executing an editing function of the first image, the processor (130) may recognize current context information based on at least one of the state of the electronic device (100) (e.g., communication state, battery state, and / or setting value of the electronic device (100), location, function running on the electronic device (100), or application running on the electronic device (100). The processor (130) may generate a second image based on at least some of the personal metadata and / or current context information.

[0040] For example, the processor (130) may receive a second user input containing information related to editing a first image based on the absence of personal metadata. For example, the information related to editing a first image may include at least one of a method for editing the first image, a function related to image editing, or information provided to a trained AI model (e.g., a prompt). The processor (130) may generate a third image in which at least a portion of the first image has been edited based on at least a portion of the second user input.

[0041] For example, when the processor (130) executes an editing function, it may receive a third user input to select whether to use personal metadata associated with the first image for editing the first image. For example, the processor (130) may provide a user interface for selecting whether to use common metadata and / or personal metadata for image editing. The processor (130) may determine the type of metadata to use for image editing based on the third user input received through the user interface for selecting whether to use common metadata and / or personal metadata. For example, if the processor (130) decides to use personal metadata based on the third user input, it may generate a second image based at least partially on the personal metadata. If the processor (130) decides not to use personal metadata based on the third user input, it may generate a fourth image that edits at least a part of the first image based on the common metadata associated with the first image.

[0042] For example, when the processor (130) executes an editing function, it may receive a fourth user input to select whether to use personal learning data or general learning data for editing the first image. For example, the processor (130) may provide a user interface for selecting whether to use personal learning data or general learning data for image editing. The processor (130) may determine the type of learning data to use for image editing based on the fourth user input received through the user interface. For example, if the processor (130) decides to use personal learning data based on the fourth user input, it may generate a second image based on at least a portion of the personal learning data associated with the first image. If the processor (130) decides not to use personal learning data based on the fourth user input, it may generate a fourth image in which at least a portion of the first image is edited based on general learning data associated with the first image.

[0043] For example, the processor (130) may display a first image and / or a second image through the display (110). For example, if the display (100) includes a plurality of displays or includes a foldable display and / or a flexible display, the processor (130) may display the first image and the second image simultaneously through each of the plurality of displays or simultaneously display them in different areas of the foldable display and / or the flexible display.

[0044] According to various embodiments, the configuration of the electronic device (100) is not limited to that shown in FIG. 1, and at least some configurations may be omitted, or at least one configuration (e.g., at least one of the components of the electronic device (200; 900; 1000)) may be added. In this disclosure, an image is described as an example of content, but the features described in this disclosure based on an image are not limited to images and may be applied to other types of content including text and / or video.

[0045] An electronic device according to one embodiment can generate and / or recommend an edited image that modifies an original image based on user-specific personal metadata without user input, based on the situation, context, emotional state, relationships of objects within the original image, and / or the state of the electronic device at the time the user acquired the original image. According to one embodiment, the electronic device can edit the original image by reflecting place information, people, and / or situations not included in the original image, based on personal metadata.

[0046]

[0047] FIG. 2 is a block diagram of an electronic device according to one embodiment.

[0048] According to one embodiment, an electronic device (200) (e.g., electronic device (100; 900; 1000)) may include a content analysis module (210), a generative AI model (220) (e.g., generative artificial intelligence model (1050)), a vision model (large vision model, LVM (230)) (e.g., generative artificial intelligence model (1050)), and a user interface (240).

[0049] According to one embodiment, the content analysis module (210) may include a context recognition engine (211), a context recognition module (213), an object analysis engine (215), a database (217), common metadata (218), and private metadata (219). Hereinafter, the content may be understood to include a still image (e.g., a photograph, a drawing, an image) or a video (e.g., at least one frame included in a video).

[0050] The context recognition engine (211) can recognize information used to obtain context information. For example, the context recognition engine (211) can recognize the state of the electronic device (200), the location of the electronic device (200), time, date, day of the week, the function running on the electronic device (200), the application running, the user's usage history of the electronic device (200), the user's communication history (e.g., conversation history performed through a specific application), the user profile (e.g., information related to the user including the user's nationality, gender, age, address, and / or biometric information), and / or information stored in the electronic device (200) (e.g., contact information). The information recognized by the context recognition engine (211) is not limited to those listed above.

[0051] The context recognition module (213) can obtain context information of the user based on information recognized by the context recognition engine (211). For example, the context recognition module (213) can obtain context information based on information recognized by the context recognition engine (211) using generative AI. For example, context information may include information recognized by the context recognition engine (211), user preference information, user's human relationships (e.g., family, friends, or colleagues), the place where the user is at the time of content acquisition, location, time, the user's surrounding circumstances, the state of the electronic device (200), and / or information related to the function and / or application being performed by the electronic device (200). The context recognition module (213) can obtain or generate at least some of the personal metadata (219) based on the context information. For example, the context recognition module (213) can generate a personal tag (hereinafter referred to as a 'context tag') related to the content. For example, personal metadata (219) (e.g., common tags) may mean metadata related to context information (i.e., different for each user).

[0052] The object analysis engine (215) can recognize objects included in the content and extract information related to the objects. For example, the object analysis engine (215) can obtain the results of analyzing objects within the content using a generative AI model (220). The object analysis engine (215) can obtain or generate at least some of the common metadata (218) based on the results of analyzing objects within the content. For example, the object analysis engine (215) can generate common tags related to the content. For example, the common metadata (218) (e.g., common tags) may refer to metadata related to the content itself, regardless of context information (i.e., not different for each user).

[0053] The database (217) may store at least one piece of content. For example, the database (217) may include at least one piece of personal learning data (which may be referred to as 'vector DB' in this disclosure) that includes an image learned based on specified personal metadata (219) using an AI model. The database (217) may include at least one piece of general learning data that includes an image learned based on specified common metadata (218) using an AI model. For example, if multiple pieces of personal learning data exist, each of the multiple pieces of personal learning data may include an image learned based on different metadata and / or an image learned for different editing methods.

[0054] The database (217) may contain original content and / or created (edited) content. The database (217) may store information about common metadata (218) and / or private metadata (219) that map to the stored content.

[0055] Common metadata (218) may include, but is not limited to, the time of content acquisition, the place of acquisition, information about objects included in the content, the name of the content, storage space, and / or capacity.

[0056] Personal metadata (219) may include at least one of the following: information related to an application used when acquiring the first content, the acquisition path of the first content, the sharing history of the first content, the conversation history related to the first content, content similar to the first content, user emotion information related to the first content, information about functions and / or applications related to the first content, information related to objects included in the first content (e.g., names of objects included in the content, location names, and / or place names), context information related to common metadata (218), or user-related information stored in the electronic device (200) (100) (e.g., contact information, email information, setting information of the electronic device (200) (100)), but is not limited thereto.

[0057] According to one embodiment, the generative AI model (220) may include a language model, a vision model, a multimodal model, or a combination thereof. The generative AI model (220) may be used to recognize objects included in content or to extract features of recognized objects. For example, the generative AI model (220) may be used to obtain context information based on at least one of user input, the state of the electronic device (200), functions and / or applications in use, the user's usage history of the electronic device (200), a user profile, or the user's communication history (e.g., conversation history through a messaging application, etc.).

[0058] According to one embodiment, LVM (230) can create edited content (250) (250) by modifying at least a portion of the original content. For example, LVM (230) can create edited content (250) based on common metadata (218) and / or private metadata (219). For example, LVM (230) can in-paint or out-paint the original content, add objects to the original content, replace objects, delete objects from the original content, add or change the background of the original content, and / or apply visual effects to the original content.

[0059] For example, if private metadata (219) for the original content exists, LVM (230) may modify at least part of the original content based at least part of the private metadata (219). For example, if private metadata (219) for the original content does not exist, LVM (230) may modify at least part of the original content based at least part of user input and / or common metadata (218) received through the user interface (240).

[0060] For example, LVM (230) can learn visual information about metadata collected through low-rank adaptation (LoRA) techniques (e.g., common metadata (218) (e.g., common tags) and / or private metadata (219) (e.g., private tags)).

[0061] According to one embodiment, the user interface (240) may provide content and / or metadata to the user and / or receive user input. For example, the user interface (240) may provide visual information through a display (not shown) (e.g., display (110; 940). For example, user input may include at least one of an input for selecting content to display or edit, an input for enabling an editing function, an input for selecting metadata and / or training data to use for editing, an input for specifying an editing function or method, or an input for saving and / or sharing the generated edited content. The user interface (240) may provide user input to the LVM (230) including content to be provided to the LVM (230) (e.g., original content, edited content (250), and / or reference content for training the LVM (230), prompts to be provided to the LVM (230), and / or setting values ​​of the LVM (230).

[0062] According to various embodiments, the configuration of the electronic device (200) is not limited to that shown in FIG. 2, and at least some configurations may be omitted, or at least one configuration (e.g., at least one of the components of the electronic device (100; 900; 1000)) may be added. According to one embodiment, at least some of the configurations of the electronic device (200) may be implemented as an integrated configuration. For example, the context awareness engine (211), the context awareness module (213), and / or the object analysis engine (215) may be implemented as (e.g., at least one processor (130; 910; 1020). For example, the generative AI model (220) and the LVM (230) may be implemented as an integrated artificial intelligence model (e.g., a language model, a vision model, a multimodal model, or a combination thereof).

[0063]

[0064] FIG. 3 is a diagram illustrating the content creation operation of an electronic device according to one embodiment.

[0065] For example, 310 represents a first user interface (UI) (310) that displays the first content (original content). For example, an electronic device (e.g., electronic device (100; 200; 900; 1000)) may display the first UI (310) containing the first content (original content) on a display. The electronic device may execute a viewer application based on user input and display the first UI (310) containing the first content through the viewer application. For example, the first screen may include a control area (311) for managing the first content. The control area (311) may include an item for setting the first content as preferred content, an item (315) for editing the first content, an item for checking information related to the first content, an item for sharing the first content, and an item for deleting the first content, and the items included in the control area (311) are not limited to those listed above.

[0066] For example, the electronic device may recognize content analysis-based metadata (which may be referred to in the present disclosure as 'common metadata', 'context-non-based metadata', or 'common tag') and / or context-based metadata (which may be referred to in the present disclosure as 'personal metadata' or 'personal tag') related to the first content during the display of the first UI (310). The electronic device may determine whether context-based metadata related to the first content (and / or personal learning data learned based on context-based metadata) exists.

[0067] The electronic device may enable an editing function for the first content in response to receiving user input for an item (315) for editing the first content. For example, the electronic device may provide a second UI (320) in response to receiving user input for an item (315) for editing the first content.

[0068] For example, 320 represents a second UI (320) for editing the first content. For example, an electronic device may display the second UI (320) including the first content and at least one editing tool. According to various embodiments, the second UI (320) may be provided through a viewer application and may also be provided through an editing application different from the viewer application.

[0069] For example, the second UI (320) may include an AI editing item (325) to enable a generative AI-based content editing function in the first UI (310). For example, the electronic device may display the AI ​​editing item (325) in the second UI (320) based on the existence of context-based metadata (and / or personal learning data learned based on context-based metadata). As another example, the electronic device may display the AI ​​editing item (325) included in the second UI (320) in a different way (e.g., a different color) based on whether context-based metadata (and / or personal learning data learned based on context-based metadata) exists.

[0070] For example, the electronic device may generate at least one recommended content (330) based on receiving user input for an AI edited item (325). For example, the electronic device may generate the recommended content (330) using a generative AI model. For example, the electronic device may generate at least one recommended content (330) based on personal learning data and / or general learning data using an AI model. For example, the generative AI model may be trained based on content analysis-based metadata and / or context-based metadata.

[0071] For example, an electronic device may generate recommended content (330) based on content analysis-based metadata (and / or general learning data) and / or context-based metadata (and / or personal learning data) using a trained generative AI model. For example, if a health-related application was running at the time the user acquired the first content, the context-based metadata associated with the first content may include health-related information (e.g., a “#Health” tag). In this case, the electronic device may generate the first recommended content (331) by using a generative AI model to change the background of the first content to a gym based on the context-based metadata, and / or by adding exercise-related objects to the first content. For example, if the electronic device (user) has a history of sharing the first content externally through a social media application with tags such as “#Tokyo” and “Market,” the context-based metadata associated with the first content may include information related to the sharing history of the first content (e.g., travel, Japan, “#Tokyo”, and / or “#Market”). In this case, the electronic device may use a generative AI model to generate a second recommended content (333) by replacing the background of the first content with a background related to a travel destination (e.g., Japan, Tokyo) based on context-based metadata, and / or by replacing the background of the first content with a market background. According to various embodiments, the information and editing method of the context-based metadata used to generate the recommended content (330) are not limited to those described above.

[0072] The electronic device may provide the generated recommended content (330) to the user. Based on user input selecting at least one of the recommended content (330), the electronic device may store the selected recommended content (330) in the electronic device or / or transmit (or share) it externally.

[0073] According to various embodiments, the method by which the electronic device generates recommended content (330) (e.g., the time of generation of recommended content (330), the generation stage, and / or trigger action) may be configured in various ways. For example, the electronic device may generate at least one recommended content (330) in response to receiving user input regarding an AI edited item (325). In another example, the electronic device may identify metadata of content being displayed on a display in response to receiving user input regarding an AI edited item (325) and provide user-selectable objects related to the identified metadata. For example, if the metadata of the content is 'Health' and 'Tokyo', the electronic device may display objects related to 'Health' and objects related to 'Tokyo' on the display. Based on receiving user input selecting at least one of the provided objects, the electronic device may provide edited content based on metadata related to the selected object as recommended content (330).

[0074]

[0075] FIG. 4 is a diagram illustrating the content generation operation of an electronic device according to one embodiment. In the following, descriptions that overlap with the description of FIG. 3 are omitted or briefly described.

[0076] According to one embodiment, an electronic device (e.g., electronic device (100; 200; 900; 1000)) may display a first UI (410) including a first image. The first UI (410) may include an item (411) for enabling an image editing function.

[0077] For example, the electronic device may include metadata (e.g., tags) (490) associated with the first image. For example, the tags (490) associated with the first image may include common tags obtained through an object analysis engine and / or context-based personal tags obtained through a context recognition engine. Hereinafter, the description assumes that the common tags associated with the first image include “#Kyoto” and “#Hayoung”, and the personal tags include “#Travel”, “#Market”, “#FamousRestaurant”, and “#Cat”.

[0078] The electronic device may provide a second UI (420) for editing a first image based on receiving user input for an item (411) for activating an editing function.

[0079] For example, the second UI (420) may include an AI editing item (421) for executing an AI-based image editing function.

[0080] For example, in operation 430, the electronic device can determine whether a context-based personal tag exists based on receiving user input for the AI ​​edit item (421). For example, if a personal tag associated with the first image exists, the electronic device can perform operation 440.

[0081] For example, in operation 440, the electronic device may determine whether there exists personal training data containing images trained based on personal tags. For example, the AI ​​model may be trained based on a reference image for training the AI ​​model, an image stored on the electronic device (e.g., a first image), a previously edited image, common tags associated with images stored on the electronic device, and / or personal tags. For example, the electronic device may use the AI ​​model to generate personal training data containing images trained based on specified personal tags and / or general training data containing images trained based on specified common tags.

[0082] For example, an AI model may be trained for each of at least one editing method. For example, the AI ​​model may include at least one AI model trained for each of at least one editing method (e.g., trained via the LoRA technique). For example, at least one editing method may include, but is not limited to, at least one of in-painting, out-painting, an editing method for adding at least one new object, an editing method for replacing a background or at least one object, an editing method for deleting a background or at least one object, an editing method for applying at least one visual effect, or an editing method for adding at least one item. According to one embodiment, the AI ​​model may be trained for each of at least one editing method, or may be trained for a plurality of editing methods as a whole. For example, the AI ​​model may include training data (e.g., personal training data and / or general training data) learned from images having metadata (e.g., personal tags or common tags). For example, the AI ​​model (e.g., training data) may be utilized commonly across at least one editing method. For example, each individual learning data may include data learned based on different individual tags and / or data learned about different editing methods.

[0083] For example, if personal learning data exists, the electronic device may provide a fourth UI (460) including at least one indication representing at least one personal learning data. For example, the electronic device may determine the availability of each of at least one editing method based on the degree of learning of the AI ​​model (and / or learning data) for each of at least one editing method. For example, the electronic device may determine the availability of each of at least one learning data based on the degree of learning for each of at least one learning data.

[0084] For example, the fourth UI (460) may include a recommendation area (480) indicating a plurality of available editing methods. The recommendation area (480) may include a first indication (481) indicating a first available editing method, a second indication (483) indicating a second available editing method, and a third indication (485) indicating a third available editing method. For example, the first editing method may include an editing method that adds a background to the first image or changes the background. The second editing method may include an editing method that adds a new object (e.g., a person) to the first image. The third editing method may include an editing method that adds an object (e.g., a sticker, text, and / or an item) to the first image or / or provides a visual effect. According to various embodiments, the number, type, and / or editing methods of the indications included in the recommendation area (480) and each indication are not limited to those described in FIG. 4.

[0085] For example, the recommendation area (480) of the fourth UI (460) may include an indication representing an available AI model (and / or, personal learning data (e.g., a vector database (vector DB))). For example, the electronic device may include at least one AI model trained based on images having personal metadata and / or at least one personal learning data including trained images. For example, the electronic device may classify images having specified personal metadata, train an AI model using the classified images, and generate and store personal learning data. For example, image and / or text information related to at least one personal learning data (e.g., an image and / or text representing the personal learning data) may be displayed in the recommendation area (480).

[0086] For example, if the electronic device includes a plurality of trained AI models, the recommendation area (480) may include a first indication (481) representing a first trained AI model, a second indication (483) representing a second trained AI model, and a third indication (485) representing a third trained AI model. For example, if the electronic device includes personal learning data including a plurality of trained images, the recommendation area (480) may include a first indication (481) representing a first representative image representing the first personal learning data, a second indication (483) representing a second representative image representing the second personal learning data, and a third indication (485) representing a third representative image representing the third personal learning data. For example, each of the plurality of personal learning data (e.g., first personal learning data to third personal learning data) may include images learned based on different metadata. According to one embodiment, a first AI model and / or first individual learning data may be associated with a first editing method, a second AI model and / or second individual learning data may be associated with a second editing method, and a third AI model and / or third individual learning data may be associated with a third editing method.

[0087] For example, the electronic device may receive user input selecting at least one of at least one indication. For example, the electronic device may provide a fifth UI (470) that displays at least one selected indication based on the user input. For example, assuming the first indication (481) is selected, the electronic device may display a fifth UI (470) in which the color of the selected first indication (481) is displayed differently from the colors of the unselected second indication (483) and third indication (485).

[0088] For example, the electronic device may use a trained AI model to edit a first image based on an editing method, an AI model, and / or training data (e.g., personal training data) corresponding to a selected indication. For example, assuming a first indication (481) representing a first editing method is selected, the electronic device may generate an edited image by editing the background of the first image into a background of a travel destination (e.g., Kyoto) and / or a background of a place (e.g., a market) based on common tags and / or personal tags. As another example, assuming a second indication (483) representing a second editing method is selected, the electronic device may generate an edited image by adding an object (e.g., a person who traveled with you or a cat) to the first image based on personal tags. As yet another example, assuming a third indication (485) representing a third editing method is selected, the electronic device may generate an edited image by adding stickers or decorative objects related to travel, markets, and / or restaurants based on personal tags.

[0089] For example, assuming that a first indication (481) representing a first AI model is selected, the electronic device can generate an edited image by using the first AI model to edit the first image based on common tags and / or private tags. As another example, assuming that a second indication (483) representing a second AI model is selected, the electronic device can generate an edited image by using the second AI model to edit the first image based on common tags and / or private tags. As another example, assuming that a third indication (485) representing a third AI model is selected, the electronic device can generate an edited image by using the third AI model to edit the first image based on common tags and / or private tags.

[0090] For example, assuming that a first indication (481) representing a first representative image representing first individual learning data is selected, the electronic device can generate an edited image by editing the first image based on the first individual learning data. As another example, assuming that a second indication (483) representing a second representative image representing second individual learning data is selected, the electronic device can generate an edited image by editing the first image based on the second individual learning data. As another example, assuming that a third indication (485) representing a third representative image representing third individual learning data is selected, the electronic device can generate an edited image by editing the first image based on the third individual learning data.

[0091] For example, if no personal tag associated with the first image exists (No of action 430), or if no personal training data exists (No of action 440), the electronic device may provide a third UI (450) that requests information related to editing the first image. For example, the electronic device may obtain user input for manually editing the first image or information for editing the first image using an AI model (e.g., voice input to provide as a prompt to the AI ​​model) through the third UI (450). Based on the information obtained through the third UI (450), the electronic device may modify at least a portion of the first image to obtain an edited image. For example, the electronic device may edit at least a portion of the first image based on general training data.

[0092]

[0093] FIG. 5 is a diagram illustrating the content generation operation of an electronic device according to one embodiment. In the following, descriptions that overlap with the descriptions of FIG. 3 and 4 are omitted or briefly described.

[0094] According to one embodiment, an electronic device (e.g., electronic device (100; 200; 900; 1000)) may provide a first UI (510) including a first image. For example, the first UI (510) may include an item (511) for enabling an editing function of the first image. The electronic device may include a tag (590) associated with the first image. Hereinafter, it is assumed that common tags associated with the first image include “#Person1”, “#Person2”, and “#Person3”, and personal tags include “#Person1”, “#Person2”, “#Person3”, “Person4_Messensger”, and “#Birthday_Calendar”. For example, Person1, Person2, and Person3 may correspond to the persons included in the first image, respectively. The electronic device may recognize Person4 not included in the first image and information associated with the first image (e.g., birthday) through context-based analysis. For example, the electronic device can recognize Person 4 by searching for persons included in the same group as Persons 1 through 3 in the contacts. The electronic device can recognize Person 4 related to Persons 1 through 3 based on messenger conversation history. The electronic device can recognize Person 4 based on calendar event content (e.g., attendee information). The electronic device can recognize that the first image is an image related to a birthday based on calendar event content. The electronic device can register tags related to "Person 4" and "Birthday" recognized based on context as personal tags.

[0095] For example, in operation 520, the electronic device may determine whether a context-based personal tag exists based on receiving user input for an item (511) for enabling an editing function of the first image. For example, if a personal tag exists, the electronic device may perform operation 530.

[0096] For example, in operation 530, the electronic device can determine whether personal learning data exists. For example, the electronic device can determine whether at least one piece of personal learning data exists, including an image learned based on a specified personal tag, using an AI model.

[0097] For example, the electronic device may provide a second UI (540) for editing a first image if personal learning data exists. For example, based on the existence of personal learning data, the electronic device may display an AI editing item (541) on the second UI (540) indicating that image editing is possible using the personal learning data.

[0098] For example, based on receiving user input for an AI editing item (541), the electronic device may generate an edited image based on personal tags and / or personal learning data using a trained AI model. For example, the electronic device may add an object corresponding to Person 4 that is not included in the first image based on personal tags and / or personal learning data. For example, the electronic device may identify an object corresponding to Person 4 based on images stored in the electronic device, tags of the images, and / or personal learning data. The electronic device may generate an edited image by adding an object corresponding to Person 4 identified in the first image using an AI model. For example, the electronic device may provide a third UI (550) containing the edited image.

[0099] According to one embodiment, the electronic device may provide at least one recommendation item related to content (e.g., an edited image) that can be generated based on a personal tag (and / or personal learning data). For example, each recommendation item may represent at least one of a personal tag, a learned AI model, personal learning data, an image editing method, or an edited image. In response to receiving user input selecting at least one of the recommendation items, the electronic device may generate an edited image corresponding to the selected recommendation item. The electronic device may provide the edited image corresponding to the selected recommendation item through a third UI (550).

[0100]

[0101] FIG. 6 is a diagram illustrating the content generation operation of an electronic device according to one embodiment. In the following, descriptions that overlap with the descriptions of FIG. 3 to 5 are omitted or briefly described.

[0102] According to one embodiment, an electronic device (e.g., electronic device (100; 200; 900; 1000)) may provide a first UI (610) including a first image. The first UI (610) may include an item (611) for enabling an editing function of the first image.

[0103] For example, in operation 620, the electronic device may determine whether a context-based personal tag exists based on receiving user input for an item (611) for enabling the editing function of the first image. For example, if a personal tag exists, the electronic device may perform operation 630.

[0104] For example, in operation 630, the electronic device may determine whether there is personal learning data including an image learned based on a personal tag. For example, if personal learning data exists, the electronic device may provide a second UI (640) for editing the first image. For example, based on the existence of personal learning data, the electronic device may display an AI editing item (641) on the second UI (640) indicating that image editing is possible using the personal learning data.

[0105] For example, the electronic device may provide a third UI (650) based on receiving a specified user input (643) while displaying the second UI (640). For example, the third UI (650) may provide a selection interface (651) for selecting training data to be used to edit the first image using an AI model. For example, the electronic device may provide a selection interface (651) for selecting whether to use general training data (and / or common tags) or personal training data (and / or personal tags) to edit the first image.

[0106] According to one embodiment, the electronic device may receive user input for selecting an area to edit a first image before or after displaying a selection interface (651). For example, the user input may include a pinch-in input (pinch zoom-out), and the pinch-in input may be an input for performing out-painting on the image.

[0107] For example, the electronic device may display a region designation interface (not shown) for designating a region to be edited together with a first image. The electronic device may designate an editing region associated with the first image based on user input received through the region designation interface. For example, the region designation interface may be configured to change the size and / or shape of the region to be edited based on user input, and may designate a specific object (e.g., the outline of a specific object) according to user input. For example, the region designation interface may include an indication that the designated region can be adjusted, and may include visual information (e.g., text, symbols, indications, outlines, colors, and / or shading) representing the designated region and / or the designated editing method.

[0108] According to various embodiments, FIG. 6 illustrates a case where outpainting is performed, but the editing method is not limited thereto. For example, the electronic device may provide a selection interface (651) for selecting whether to use a common tag or a private tag to edit the first image using the editing method before editing the first image through various editing methods (e.g., inpainting, adding objects, deleting objects, changing objects, decorating the image, and / or applying visual effects).

[0109] For example, if the electronic device receives user input selecting to use personal learning data (and / or personal tags) for editing the first image through the selection interface (651), it may generate a first edited image (660) by editing the first image based on the personal learning data (and / or personal tags) using a trained AI model. For example, the electronic device may generate the first edited image (660) by modifying the background of the first image to a different background based on the personal learning data (and / or personal tags).

[0110] For example, if the electronic device receives user input selecting to use general training data (and / or common tags) for editing the first image through the selection interface (651), it can generate a second edited image (670) by editing the first image based on the general training data (and / or common tags) using a trained AI model. For example, the electronic device can generate the second edited image (670) by expanding (e.g., outpainting) the background image included in the first image.

[0111]

[0112] An electronic device according to one embodiment of the present disclosure may include a display, a memory for storing instructions, and at least one processor.

[0113] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may display a first image through the display.

[0114] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may receive a first user input for executing an editing function of the first image.

[0115] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate a second image in which at least a portion of the first image is edited using an artificial intelligence (AI) model learned based on said personal metadata, based on the existence of said personal metadata associated with the first image.

[0116] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may display at least one indication representing at least one editing method available through the display.

[0117] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may receive a user input selecting at least one of the at least one indication.

[0118] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate the second image based on an editing method corresponding to the selected at least one indication.

[0119] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may determine whether each of the at least one editing method is available based on the degree of learning of the AI ​​model for each of the at least one editing method.

[0120] According to one embodiment, the at least one editing method may include at least one of in-painting, out-painting, an editing method for adding at least one new object, an editing method for replacing a background or at least one object, an editing method for deleting a background or at least one object, an editing method for applying at least one visual effect, or an editing method for adding at least one item.

[0121] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate at least one recommended image by editing at least a portion of the first image based on the personal metadata using the learned AI model.

[0122] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may display the at least one recommended image through the display.

[0123] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may determine a selected recommended image among the at least one recommended images as the second image based on user input.

[0124] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate a second image in which at least a portion of the first image is edited based on the at least one individual learning data using the AI ​​model, based on the existence of at least one individual learning data including an image learned based on specified individual metadata associated with the first image.

[0125] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may display at least one recommendation item through the display, which represents at least one personal learning data including an image learned based on specified personal metadata associated with the first image.

[0126] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may receive user input selecting at least one of the at least one recommendation item.

[0127] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate the second image based on personal learning data corresponding to the selected at least one recommendation item.

[0128] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may receive a second user input containing information related to editing the first image based on the absence of the personal metadata.

[0129] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate a third image in which at least a portion of the first image is edited based on the second user input.

[0130] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may display an indication with the first image that the first image is editable based on the personal metadata while the first image is being displayed.

[0131] According to one embodiment, the personal metadata may include at least one of information related to an application that was in use at the time of acquiring the first image, a sharing history of the first image, a conversation history related to the first image, an image similar to the first image, user emotion information related to the first image, or information related to objects included in the first image.

[0132] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may recognize current context information based on at least one of the state, location, or application running on the electronic device when executing the editing function of the first image.

[0133] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate the second image based on at least some of the current context information.

[0134] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may receive a third user input selecting whether to use personal metadata associated with the first image for editing the first image when executing the editing function.

[0135] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate the second image based at least partially on the personal metadata if it determines to use the personal metadata based on the third user input.

[0136] According to one embodiment, when the instructions are executed individually or collectively by the at least one processor, the electronic device may generate a fourth image in which at least a portion of the first image is edited based on common metadata associated with the first image, if it is determined not to use the private metadata based on the third user input.

[0137]

[0138] FIG. 7 is a flowchart of a method for generating content of an electronic device according to one embodiment. In the following description, the content is described based on an image as an example, but the content is not limited to an image and may include other types of content, including video.

[0139] According to one embodiment, in operation 710, an electronic device (e.g., electronic device (100; 200; 900; 1000)) may display a first image. For example, the first image may include an image stored in the electronic device and / or an image obtained from an external source (e.g., an external device, an external server, and / or an external network). The first image may include a photograph, a drawing, an image, and / or a video (e.g., at least one frame of a video). For example, the electronic device may display the first image through an image viewer application.

[0140] According to one embodiment, in operation 720, the electronic device may receive a first user input for executing an editing function of a first image. For example, the electronic device may enable an image editing function provided by a viewer application based on the first user input. For example, the electronic device may receive a first user input for selecting a user interface (e.g., an indication, a button, a menu, and / or an affordance item) for executing an image editing function.

[0141] According to one embodiment, an electronic device may receive user input for specifying an editing area associated with a first image. For example, the electronic device may receive user input for specifying at least a portion of the area included in the first image, at least one object, and / or the area outside the image (e.g., when extending the background of the image) as the editing area. For example, user input for editing the background (area) of the image may include pinch-in input (e.g., pinch-zoom-out input). For example, the electronic device may perform editing on the editing area based on the user input. For example, when the electronic device receives user input specifying a specific object, the electronic device may perform editing using a generative AI model for the specific object. For example, when the electronic device receives pinch-in (pinch-zoom-out) input, it may reduce the image displayed on the screen and, depending on the reduction of the image, add the background of the image (e.g., outpainting) using an AI model to the area where the image is not displayed on the display.

[0142] For example, the electronic device may provide a region designation interface for designating an editing region. The electronic device may determine an editing region associated with a first image based on user input received through the region designation interface. For example, the region designation interface may include an indication that the designated region can be adjusted, and may include visual information (e.g., text, symbols, indications, outlines, colors, and / or shading) representing the designated region and / or the designated editing method.

[0143] According to one embodiment, in operation 730, the electronic device may generate a second image in which at least a portion of the first image is edited using an artificial intelligence (AI) model learned based on personal metadata, based on the existence of personal metadata associated with the first image.

[0144] For example, an electronic device may determine whether personal metadata related to a first image exists. The personal metadata may include at least one of information related to an application that was in use at the time of acquiring the first image, a sharing history of the first image, a conversation history related to the first image, images similar to the first image, user sentiment information related to the first image, information about functions and / or applications related to the first image, or information related to objects contained in the first image. For example, the electronic device may determine whether personal training data exists that includes an image learned based at least partially on specified personal metadata (e.g., an image having personal metadata) using an AI model.

[0145] For example, the electronic device may display a user interface (e.g., indication, button, menu, and / or affordance item) indicating that the first image is editable based on personal learning data (and / or personal metadata) while the first image is being displayed. For example, the electronic device may generate a second image in which at least a portion of the first image has been edited based on personal learning data (and / or personal metadata), based on user input selecting the user interface indicating that the first image is editable based on personal learning data (and / or personal metadata). The electronic device may also generate the second image without additional user input when the image editing function is enabled.

[0146] For example, the trained artificial intelligence model may include, but is not limited to, a generative AI model including a language model (e.g., large language model (LMM)), a vision model (e.g., large vision model (LVM)), and / or a multimodal model (e.g., large multi-modal model (LMM)).

[0147] The electronic device may generate a second image by editing at least a portion of the first image based on personal learning data (and / or personal metadata) using a trained AI model. For example, editing the first image may include in-painting, out-painting, changing the background of the first image, adding objects to the first image, deleting at least a portion of the first image, replacing or changing objects in the first image, and / or applying visual effects to the first image.

[0148] According to one embodiment, the electronic device may provide a selection interface for selecting whether to use general training data (and / or common metadata (e.g., common tags)) or private training data (and / or private metadata (e.g., private tags)) to edit the first image before editing the first image. The electronic device may edit the first image using the training data (and / or metadata (e.g., tags)) selected through the selection interface. For example, the electronic device may edit the first image based on the private training data (and / or private metadata) when the user selects the private training data (and / or private metadata) through the selection interface, and edit the first image based on the general training data (and / or common metadata) when the user selects the general training data (and / or common metadata) through the selection interface.

[0149] For example, the electronic device may display at least one indication indicating at least one available editing method. For example, the electronic device may determine the availability of each of the at least one editing method based on the degree of learning of an AI model for each of the at least one editing method. For example, the electronic device may receive user input selecting at least one of the at least one indication. The electronic device may generate a second image based on the editing method corresponding to the selected at least one indication. For example, the at least one editing method may include at least one of in-painting, out-painting, an editing method for adding at least one new object, an editing method for replacing a background or at least one object, an editing method for deleting a background or at least one object, an editing method for applying at least one visual effect, or an editing method for adding at least one item.

[0150] For example, the electronic device may provide indications representing each of the learned AI models and / or training data (e.g., personal training data). For example, the electronic device may provide an indication including an image representing the learned AI models and / or training data. For example, the image representing the learned AI models and / or training data may represent the form and / or type of an edited image that can be generated using the said AI models and / or training data. The electronic device may edit a first image based on the AI ​​models and / or training data corresponding to the indication selected based on user input.

[0151] For example, the electronic device may generate at least one image by editing at least a portion of a first image based on personal learning data (and / or personal metadata) using a trained AI model. The electronic device may provide the generated at least one image as a recommended image. The electronic device may determine a selected image among the recommended images as a second image based on user input.

[0152] For example, when executing an editing function of a first image, the electronic device may recognize current context information based on at least one of the state of the electronic device (e.g., communication status, battery status, and / or setting value of the electronic device), location, a function running on the electronic device, or an application running on the electronic device. The electronic device may generate a second image based on at least some of personal metadata and / or current context information.

[0153] For example, the electronic device may receive a second user input containing information related to editing a first image, based on the absence of personal learning data (and / or personal metadata). For example, the information related to editing the first image may include at least one of a method for editing the first image, a function related to image editing, or information provided to a trained AI model (e.g., a prompt). The electronic device may generate a third image in which at least a portion of the first image has been edited, based on at least a portion of the second user input. The electronic device may generate a third image in which at least a portion of the first image has been edited, based on general learning data (and / or common metadata), based on the absence of personal learning data (and / or personal metadata).

[0154] For example, when an electronic device executes an editing function, it may receive a third user input selecting whether to use personal training data (and / or personal metadata) associated with a first image for editing the first image. For example, the electronic device may provide a user interface for selecting whether to use general training data (and / or common metadata) and / or personal training data (and / or personal metadata) for image editing. The electronic device may determine the type of training data (and / or metadata) to be used for image editing based on the third user input received through the user interface for selecting whether to use general training data (and / or common metadata) and / or personal training data (and / or personal metadata). For example, if the electronic device decides to use personal training data (and / or personal metadata) based on the third user input, it may generate a second image based at least partially on the personal training data (and / or personal metadata). If the electronic device decides not to use personal learning data (and / or personal metadata) based on a third user input, it may generate a fourth image by editing at least a portion of the first image based on general learning data (and / or common metadata) associated with the first image.

[0155] According to various embodiments, the order of the operations of FIG. 7 may be changed, or at least some of the operations may be performed simultaneously. At least some of the operations of FIG. 7 may be omitted, or at least one operation (e.g., at least one of the operations of FIG. 8a and / or 8b) may be added. For example, the operations of FIG. 7 may be performed independently of the operations of FIG. 8a and / or 8b, performed in conjunction with each other, or at least some of the operations may be integrated.

[0156]

[0157] FIG. 8a is a flowchart of an image generation method of an electronic device according to one embodiment. In the following, descriptions that overlap with the description of FIG. 7 are omitted or briefly described. In the following, the content is described based on an image as an example, but the content is not limited to an image and may include other types of content, including video.

[0158] According to one embodiment, in operation 810a, an electronic device (e.g., electronic device (100; 200; 900; 1000)) can display an image (e.g., original image).

[0159] According to one embodiment, in operation 820a, the electronic device can execute an image editing function. For example, the electronic device can enable the image editing function based on specified user input.

[0160] According to one embodiment, in operation 830a, the electronic device may determine whether personal learning data exists. For example, the electronic device may determine whether at least one piece of personal learning data exists, comprising an image learned based on specified context-based metadata (e.g., an image having specified context-based metadata) using an AI model. In the present disclosure, 'context-based metadata' may be referred to as 'personal metadata'. The electronic device may perform operation 840a if personal learning data exists. The electronic device may perform operation 870a if personal learning data does not exist.

[0161] According to one embodiment, in operation 840a, the electronic device may generate at least one recommendation item based on personal learning data. For example, the electronic device may generate information (e.g., images and / or text) representing personal learning data (which may be referred to as 'vector DB' in this disclosure) as a recommendation item. For example, the electronic device may generate at least one recommendation image as a recommendation item by editing at least a portion of an original image using a trained AI model. For example, the electronic device may display the generated at least one recommendation item through a display.

[0162] According to one embodiment, in operation 850a, the electronic device may determine whether a first user input selecting at least one of the recommended items is received. If the first user input selecting the recommended item is received, the electronic device may perform operation 860a. If the first user input is not received, the electronic device may perform operation 870a.

[0163] According to one embodiment, in operation 860a, the electronic device may generate an image (e.g., a recommended image and / or an edited image) corresponding to a selected recommended item based on a first user input. For example, the electronic device may generate a recommended image corresponding to a selected recommended item, and / or generate an edited image by editing an image based on personal learning data corresponding to a selected recommended item. The recommended image may also be generated at the time of generating the recommended item prior to operation 860a. The electronic device may store and / or share the recommended image and / or the edited image. For example, the electronic device may store the selected recommended image and / or the edited image in an electronic device (e.g., memory) and / or in an external storage space (e.g., an external electronic device or an external server). The electronic device may transmit the selected recommended image and / or the edited image to a network (e.g., a cloud), social media, and / or an external electronic device.

[0164] According to one embodiment, in operation 870a, the electronic device may determine whether a second user input for image editing is received. For example, the second user input may include at least one of an input specifying a region or object to be edited in the original image, an input selecting an editing method, or an input including parameters for editing (e.g., a prompt to be input to a generative AI model). The second user input may be referred to as an input for manually editing the image. If the second user input is received, the electronic device may perform operation 880a. If the second user input is not received, the electronic device may perform operation 890a.

[0165] According to one embodiment, in operation 880a, the electronic device may generate an edited image based on a second user input. For example, the electronic device may generate an edited image in which at least a portion of the image is edited based on general training data (and / or common metadata). The electronic device may store the generated edited image and / or share it externally (e.g., an external device, an external server, and / or an external network).

[0166] According to one embodiment, in operation 890a, the electronic device may terminate image editing. For example, the electronic device may disable the image editing function when there is no second user input.

[0167] According to various embodiments, the order of the operations of FIG. 8a may be changed, or at least some of the operations may be performed simultaneously. At least some of the operations of FIG. 8a may be omitted, or at least one operation (e.g., at least one of the operations of FIG. 7 and / or FIG. 8b) may be added. For example, the operations of FIG. 8a may be performed independently of the operations of FIG. 7 and / or FIG. 8b, performed in conjunction with each other, or at least some of the operations may be integrated.

[0168]

[0169] FIG. 8b is a flowchart of an image generation method of an electronic device according to one embodiment. In the following, descriptions that overlap with the descriptions of FIG. 7 and 8a are omitted or briefly described. In the following, the content is described based on images as examples, but the content is not limited to images and may include other types of content, including videos.

[0170] According to one embodiment, in operation 810b, the electronic device (e.g., electronic device (100; 200; 900; 1000)) can display an image (e.g., original image).

[0171] According to one embodiment, in operation 820b, the electronic device can execute an image editing function. For example, the electronic device can enable an image editing function based on specified user input.

[0172] According to one embodiment, in operation 830b, the electronic device may determine whether there is personal learning data including an image learned based on context-based metadata (e.g., an image having context-based metadata). If personal learning data is present, the electronic device may perform operation 840b. If personal learning data is not present, the electronic device may perform operation 890b.

[0173] According to one embodiment, in operation 840b, the electronic device may receive a first user input for selecting training data to be used for image editing. For example, the electronic device may provide a selection interface for selecting whether to use personal training data for image editing or general training data including images trained based on non-context-based metadata (e.g., images having non-context-based metadata). The electronic device may select training data to be used for image editing based on the first user input received through the selection interface. That is, the electronic device may select whether to use data trained based on which metadata (context-based metadata or non-context-based metadata) based on the first user input.

[0174] According to one embodiment, in operation 850b, the electronic device can determine whether personal training data has been selected. For example, the electronic device can identify whether training data selected by a first user input received through a selection interface is personal training data. The electronic device can perform operation 860b if personal training data has been selected, and perform operation 890b if personal training data has not been selected.

[0175] According to one embodiment, in the operation of 860b, the electronic device may provide at least one recommendation item. For example, each of the at least one recommendation item may correspond to at least one image editing method, a trained AI model, and / or training data (e.g., personal training data). For example, each of the at least one recommendation item may include an image representing the image editing method, an image representing the trained AI model, and / or an image representing the training data. For example, the trained AI model may include a model trained based on a specified image editing method, a specified training image, and / or specified metadata. For example, the training data may include data trained based on a specified image editing method, a specified training image, and / or specified metadata (e.g., an image).

[0176] According to one embodiment, in operation 870b, the electronic device may receive a second user input selecting at least one of the recommended items.

[0177] According to one embodiment, in the operation of 880b, the electronic device can generate an edited image corresponding to a selected recommendation item based on personal learning data. For example, the electronic device can generate an edited image using an image editing method corresponding to a selected recommendation item based on personal learning data. The electronic device can generate an edited image using an AI model corresponding to a selected recommendation item based on personal learning data. The electronic device can generate an edited image based on personal learning data corresponding to a selected recommendation item.

[0178] According to one embodiment, in operation 890b, the electronic device may generate an edited image based on general learning data. According to one embodiment, in operation 890b, the electronic device may display at least one recommended item based on general learning data in the same or similar manner as in operations 860b and 870b, and receive user input selecting at least one of the recommended items. The electronic device may generate an edited image corresponding to the selected recommended item based on general learning data.

[0179] According to various embodiments, the order of the operations of FIG. 8b may be changed, or at least some of the operations may be performed simultaneously. At least some of the operations of FIG. 8b may be omitted, or at least one operation (e.g., at least one of the operations of FIG. 7 and / or FIG. 8a) may be added. For example, the operations of FIG. 8b may be performed independently of the operations of FIG. 7 and / or FIG. 8a, performed in conjunction with each other, or at least some of the operations may be integrated.

[0180]

[0181] A method according to one embodiment of the present disclosure may include an operation of displaying a first image. According to one embodiment, the method may include an operation of receiving a first user input for executing an editing function of the first image. According to one embodiment, the method may include an operation of generating a second image in which at least a portion of the first image is edited using an artificial intelligence (AI) model learned based on said personal metadata, based on the existence of personal metadata associated with the first image.

[0182] According to one embodiment, the operation of generating the second image may include the operation of displaying at least one indication representing at least one available editing method.

[0183] According to one embodiment, the operation of generating the second image may include receiving a user input selecting at least one of the at least one indication.

[0184] According to one embodiment, the operation of generating the second image may include the operation of generating the second image based on an editing method corresponding to at least one selected indication.

[0185] According to one embodiment, the method may include an operation of determining whether each of the at least one editing method is available based on the degree of learning of the AI ​​model for each of the at least one editing method.

[0186] According to one embodiment, the at least one editing method may include at least one of in-painting, out-painting, an editing method for adding at least one new object, an editing method for replacing a background or at least one object, an editing method for deleting a background or at least one object, an editing method for applying at least one visual effect, or an editing method for adding at least one item.

[0187] According to one embodiment, the operation of generating the second image may include the operation of generating at least one recommended image by editing at least a portion of the first image based on the personal metadata using the learned AI model.

[0188] According to one embodiment, the operation of generating the second image may include the operation of providing the at least one recommended image.

[0189] According to one embodiment, the operation of generating the second image may include the operation of determining a selected recommended image among the at least one recommended image as the second image based on user input.

[0190] According to one embodiment, the method may include an operation of displaying, together with the first image, an indication that the first image is editable based on the personal metadata while the first image is being displayed.

[0191] According to one embodiment, the operation of generating the second image may include, based on the existence of at least one personal learning data including an image learned based on specified personal metadata related to the first image, using the AI ​​model to generate a second image in which at least a portion of the first image is edited based on the at least one personal learning data.

[0192] According to one embodiment, the operation of generating the second image may include the operation of providing at least one recommendation item representing at least one personal learning data including an image learned based on specified personal metadata associated with the first image.

[0193] According to one embodiment, the operation of generating the second image may include receiving user input selecting at least one of the at least one recommended item.

[0194] According to one embodiment, the operation of generating the second image may include the operation of generating the second image based on personal learning data corresponding to at least one selected recommendation item.

[0195] According to one embodiment, the personal metadata may include at least one of information related to an application that was in use at the time of acquiring the first image, a sharing history of the first image, a conversation history related to the first image, an image similar to the first image, user emotion information related to the first image, or information related to objects included in the first image.

[0196] According to one embodiment, the method may include an operation of recognizing current context information based on at least one of the state, location, or application running on the electronic device when executing an editing function of the first image.

[0197] According to one embodiment, the operation of generating the second image may include the operation of generating the second image based on at least some of the current context information.

[0198] According to one embodiment, the method may include receiving a third user input that selects whether to use personal metadata associated with the first image for editing the first image when executing the editing function.

[0199] According to one embodiment, the operation of generating the second image may include, when it is determined to use the personal metadata based on the third user input, the operation of generating the second image based at least partially on the personal metadata.

[0200] According to one embodiment, the operation of generating the second image may include, when it is determined not to use the private metadata based on the third user input, the operation of generating a fourth image by editing at least a portion of the first image based on common metadata associated with the first image.

[0201] A storage medium according to one embodiment of the present disclosure may store instructions that cause the electronic device to display a first image when executed individually or collectively by at least one processor of the electronic device.

[0202] According to one embodiment, when the instructions are executed by the at least one processor, the electronic device may receive a first user input for executing an editing function of the first image.

[0203] According to one embodiment, when the instructions are executed by the at least one processor, the electronic device may generate a second image in which at least a portion of the first image is edited using an artificial intelligence model learned based on said personal metadata, based on the existence of said personal metadata associated with the first image.

[0204]

[0205] FIG. 9 is a block diagram of an exemplary electronic device (900) capable of performing the operations described in this document.

[0206] Referring to FIG. 9, the electronic device (900) may be one of various forms of electronic devices, such as a notebook (990), smartphones (991) having various form factors (e.g., a bar-type smartphone (991-1), a foldable-type smartphone (991-2), or a sliderable (or rollable)-type smartphone (991-3)), a tablet (992), a cellular phone (not shown), and other similar computing devices (not shown). The components, their relationships, and their functions illustrated in FIG. 9 are illustrative only and are not intended to limit the implementations described or claimed herein. The electronic device (900) may be referred to as a mobile device, a user device, a multifunction device, a portable device, or a server.

[0207] The electronic device (900) may include components comprising at least one processor (910) (hereinafter referred to as processor (910)), at least one memory (920) (hereinafter referred to as memory (920)), at least one display (940) (hereinafter referred to as display (940)), at least one image sensor (950) (hereinafter referred to as image sensor (950)), at least one communication circuit (960) (hereinafter referred to as communication circuit (960)), and / or at least one sensor (970) (hereinafter referred to as sensor (970)). The components are merely exemplary. For example, the electronic device (900) may include other components (e.g., power management integrated circuitry (PMIC), audio processing circuit, antenna, rechargeable battery, or input / output interface). For example, some components may be omitted from the electronic device (900). For example, some components may be integrated into a single component.

[0208] The processor (910) may be implemented as one or more IC (integrated circuit (or circuitry)) chips and may perform various data processing operations. The processor (910) may include at least one electrical circuit and may process instructions (or programs, data, etc.) stored in memory (920) individually or collectively in a distributed manner. The processor (910) may include a processor assembly comprising one or more processing circuits. The processor (910) may include any processing circuit that is operative to control the performance and operations of one or more components of the electronic device (900) (e.g., memory (920), display (940), image sensor (950), communication circuit (960), and / or sensor (970)). For example, the processor (910) (e.g., application processor (AP)) may be implemented as a system on chip (SoC) (e.g., a single chip or chipset). For example, the processor (910) may be implemented with a plurality of cores (or at least one core circuit), a plurality of chips, or a plurality of chipsets. For example, the processor (910) may include one or more processing circuits. For example, the processor (910) may include one or more processing circuits configured to perform the various functions of the present disclosure individually and / or collectively. As an example without limitation, at least a portion of the processor (910) may be included in a first chip of the electronic device (900), and at least another portion of the processor (910) may be included in a second chip of the electronic device (900) different from the first chip of the electronic device (900).

[0209] For example, the processor (910) may include a central processing unit (CPU) (911), a graphics processing unit (GPU) (912), a neural processing unit (NPU) (913), an image signal processor (ISP) (914), a display controller (915), a memory controller (916), a storage controller (917), a communication processor (CP) (918), and / or a sensor interface (919). These components of the processor (910) are merely exemplary. For example, the processor (910) may include other components. For example, some components of the processor (910) may be omitted from the processor (910). For example, some components of the processor (910) may be included as separate components of the electronic device (900) outside of the processor (910). For example, some components of the processor (910) (e.g., memory controller (916)) may be included in other components (e.g., at least part of memory (920), an interface (e.g. available for connection to at least one component of the electronic device (100)), a display (940) and / or an image sensor (950)).

[0210] The processor (910) may cause other components of the electronic device (900) to perform various operations by executing instructions stored in memory (920). The CPU (911) (or central processing circuit) may be configured to control the components of the processor (910) based on the execution of instructions stored in memory (920) (e.g., volatile memory (921) and / or non-volatile memory (922)). The GPU (912) (or graphics processing circuit) may be configured to execute parallel operations (e.g., rendering). The NPU (913) (or neural processing circuit, or AI (artificial intelligence) chip) may be configured to execute operations for an artificial intelligence model (e.g., convolution computation). An ISP (914) (or image signal processing circuit) may be configured to process a raw image acquired through an image sensor (950) into a format suitable for a component within the electronic device (900) or a component of the processor (910). A display controller (915) (or display control circuit, or DPU (display processing unit)) may be configured to process an image acquired from a CPU (911), GPU (912), ISP (914), or memory (920) (e.g., volatile memory (921)) into a format suitable for a display (940). A memory controller (916) (or memory control circuit) may be configured to control reading data from volatile memory (921) and writing data to volatile memory (921). A storage controller (917) (or storage control circuit) may be configured to control reading data from non-volatile memory (922) and writing data to non-volatile memory (922).The CP (918) (communication processing circuit) may be configured to process data obtained from a component of the processor (910) into a format suitable for transmitting to another electronic device via the communication circuit (960), or to process data obtained from another electronic device via the communication circuit (960) into a format suitable for processing by the component of the processor (910). For example, the communication circuit (960) may include one or more communication circuits. The sensor interface (919) (or sensing data processing circuit, sensor hub) may be configured to process data regarding the state of the electronic device (900) and / or the state around the electronic device (900), obtained through the sensor (970), into a format suitable for the component of the processor (910).

[0211] Memory (920) may include one or more storage media (or one or more storage devices). For example, memory (920) may include a memory assembly comprising one or more storage media. For example, the one or more storage media may include a hard drive, a permanent memory such as flash memory, read-only memory (ROM) (e.g., non-volatile memory (922)), a semi-permanent memory such as random access memory (RAM) (e.g., volatile memory (921)), any other suitable type of storage (or storage assembly), or any combination thereof. Memory (920) may include a cache memory, which is one or more different types of memory used to temporarily store data for a function or feature of the electronic device (900). As an example not limited to, the cache memory may be included within the processor (910). The memory (920) may be fixedly embedded within the electronic device (900) or incorporated into one or more suitable types of components (e.g., a SIM (subscriber identity module) card and / or an SD (secure digital) card) that can be repeatedly inserted into and removed from the electronic device (900).

[0212] For example, memory (920) may store one or more software applications, such as operating system (or system) software applications, firmware software applications, driver software applications, plugin (e.g., add-in, add-on, and / or applet) software applications, and / or any other suitable software applications. For example, the one or more software applications may include instructions executable by the processor (910). For example, memory (920) may store instructions that can be called by an application programming interface (API). For example, memory (920) may store instructions within a library.

[0213]

[0214] FIG. 10 illustrates a generative artificial intelligence system according to one embodiment.

[0215] Referring to FIG. 10, in a generative artificial intelligence system (1000), a user question / response interface (1010) can receive user input. The user input may be in the form of natural language, images, and / or videos. Additionally, context information may be transmitted along with the user input. The context information may include various additional information at the time of user input. For example, information about the application currently being used by the user or the user's location information. Furthermore, the user input may be in a mixed form of the aforementioned natural language, images, sounds, and context information. Additionally, the user input may be in a non-natural language form, such as selecting a menu.

[0216] According to one embodiment, the user question / response interface (1010) can output results of a generative artificial intelligence system to the user. The output may be in the form of natural language or specific content, and may also be provided in the form of actions requested by the user. The user question / response interface (1010) can output results of a generative artificial intelligence system (1000) to the user. The output may be in the form of natural language or specific content, and may also be provided in the form of actions requested by the user.

[0217] According to one embodiment, the artificial intelligence framework (1020) can receive user input and coordinate and control each component or module necessary to perform the user's intention based on the user's query.

[0218] According to one embodiment, user input received from a user question / response interface (1010) may be transmitted to a prompt design module (1021). The prompt design module (1021) may be used to generate prompts suitable for inputting user input into a large language model (LMM) or a large multi-modal model (LMM). The prompt design module (1021) may be an artificial intelligence component that uses machine learning algorithms or neural networks to develop better prompts over time. The prompt design module (1021) may generate prompts by accessing a database (1030) containing user preference data, a prompt library, and prompt examples based on user input, and transmit the generated prompts to the LLM or LMM.

[0219] According to one embodiment, the API / plugin management module (1022) can perform the role of communicating with external information when there is a request for additional information when transmitting user input as input to the generative artificial intelligence model (1050). The API / plugin management module (1022) establishes a channel to communicate with the outside of the artificial intelligence interface through an application programming interface (API), and can enable access to various data sources through the established channel. Additionally, the API / plugin management module (1022) can request the action through the API if the application / service module (1040) needs to perform an action that executes the user input as a final step, rather than an intermediate result. The information obtained from the outside may be used to generate a prompt in the prompt design module (1021) along with the user input, or it may be transmitted as input to the generative artificial intelligence model (1050).

[0220] According to one embodiment, the conversion module (1023) can fine-tune the output from the generative artificial intelligence model (1050). For example, the conversion module (1023) can verify whether the content generated through the LLM and / or LMM is irrelevant, contains biased content, or contains harmful content. Additionally, the conversion module (1023) can determine the extent to which the output matches what the user wants and, if additional processing is required, proceed with that process. Furthermore, the conversion module (1023) can configure and provide hints to the user to avoid unwanted output.

[0221] According to one embodiment, a generative artificial intelligence model (1050) may generally refer to an artificial intelligence neural network that generates new forms of data based on user input information. The generative artificial intelligence model (1050) may include a model that generates images and / or a model that generates language. Models that generate images include, but are not limited to, GANs (generative adversarial networks) and VAEs (variational autoencoders), and examples include diffusion-based generative models that use VAEs and Transformer structures. Models that generate language are models trained to output the most statistically appropriate output value based on input values, and examples include models such as CHAT-GPT 3 and CHAT-GPT 4. There are also LMMs that can recognize various forms of data input, such as text, images, and voice, and generate new data corresponding to them.

[0222]

[0223] The various embodiments of this document and the terms used therein are not intended to limit the technical features described in this document to specific embodiments, and should be understood to include various modifications, equivalents, or substitutions of said embodiments. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of said items unless the relevant context clearly indicates otherwise. In this document, phrases such as "A or B," "at least one of A and B," "at least one of A or B," "A, B or C," "at least one of A, B and C," and "at least one of A, B, or C" may each include any one of the items listed together in the corresponding phrase, or all possible combinations thereof. Terms such as "first," "second," or "first" or "second" may be used simply to distinguish said components from other said components and do not limit said components in any other aspect (e.g., importance or order). Where any (e.g., 1st) component is referred to as "coupled" or "connected" to another (e.g., 2nd) component, with or without the terms "functionally" or "communicationly," it means that said any component may be connected to said other component directly (e.g., via a wire), wirelessly, or through a third component.

[0224] As used in the various embodiments of this document, the term “module” may include a unit implemented in hardware, software, or firmware, and may be used interchangeably with terms such as logic, logic block, component, or circuit, for example. A module may be a component formed integrally, or a minimum unit of said component or a part thereof that performs one or more functions.

[0225] According to various embodiments, each component (e.g., module or program) of the components described above may include a singular or multiple entities, and some of the multiple entities may be separated and placed in other components. According to various embodiments, one or more of the components or operations of the aforementioned components may be omitted, or one or more other components or operations may be added. Generally or additionally, multiple components (e.g., module or program) may be integrated into a single component. In this case, the integrated component may perform one or more functions of each of the multiple components in the same or similar manner as those performed by the corresponding component among the multiple components prior to integration. According to various embodiments, operations performed by the module, program, or other components may be executed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be executed in a different order, omitted, or one or more other operations may be added.

Claims

1. In an electronic device, display; Memory for storing instructions; and It includes at least one processor, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, A first image is displayed through the above display, and Receiving a first user input to execute an editing function of the first image, and An electronic device that generates a second image by editing at least a portion of the first image using an artificial intelligence (AI) model learned based on said personal metadata, based on the existence of said personal metadata associated with the first image.

2. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, Displaying at least one indication indicating at least one editing method available through the above display, and Receiving user input to select at least one of the above at least one indication, and An electronic device that generates the second image based on an editing method corresponding to at least one selected indication.

3. In Claim 2, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that determines whether each of the at least one editing method is available based on the degree of learning of the AI ​​model for each of the at least one editing method.

4. In Claim 2, The above-mentioned at least one editing method comprises at least one of in-painting, out-painting, an editing method for adding at least one new object, an editing method for replacing a background or at least one object, an editing method for deleting a background or at least one object, an editing method for applying at least one visual effect, or an editing method for adding at least one item.

5. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, Using the above-mentioned trained AI model, at least one recommended image is generated by editing at least a portion of the above-mentioned first image based on the above-mentioned personal metadata, and Through the above display, at least one recommended image is displayed, and An electronic device that determines a selected recommended image among at least one recommended image as the second image based on user input.

6. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that generates a second image by editing at least a portion of the first image based on the at least one personal learning data, using the AI ​​model, based on the existence of at least one personal learning data including an image learned based on specified personal metadata related to the first image.

7. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, Through the above display, at least one recommendation item representing at least one personal learning data including an image learned based on specified personal metadata related to the first image is displayed, and Receiving user input to select at least one of the above at least one recommendation item, and An electronic device that generates the second image based on personal learning data corresponding to at least one selected recommendation item.

8. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, Based on the fact that the above personal metadata does not exist, a second user input containing information related to the editing of the first image is received, and An electronic device that generates a third image by editing at least a portion of the first image based on the second user input.

9. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that displays, together with the first image, an indication that the first image is editable based on the personal metadata while displaying the first image.

10. In Claim 1, The above personal metadata comprises at least one of the following: information related to an application being used at the time of acquiring the first image, a sharing history of the first image, a conversation history related to the first image, an image similar to the first image, user emotion information related to the first image, or information related to objects included in the first image.

11. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, When executing the editing function of the first image, current context information is recognized based on at least one of the state, location, or application running on the electronic device, and An electronic device that generates the second image based on at least some of the current context information.

12. In Claim 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, When the above editing function is executed, a third user input is received to select whether to use personal metadata related to the first image for editing the first image, and If it is decided to use the personal metadata based on the third user input above, the second image is generated based on at least part of the personal metadata, and An electronic device that generates a fourth image by editing at least a portion of the first image based on common metadata associated with the first image, when it is decided not to use the private metadata based on the third user input.

13. In a method for generating an image of an electronic device, Action of displaying the first image; An operation of receiving a first user input for executing an editing function of the first image; and A method comprising the operation of generating a second image by editing at least a portion of the first image using an artificial intelligence (AI) model learned based on said personal metadata, based on the existence of said personal metadata associated with the first image.

14. In Claim 13, The operation of generating the second image above is, An action of displaying at least one indication representing at least one available editing method; The operation of receiving user input selecting at least one of the above at least one indication; and A method comprising the operation of generating the second image based on an editing method corresponding to at least one selected indication.

15. In a storage medium for storing computer-readable instructions, When the above instructions are executed individually or collectively by at least one processor of the electronic device, the electronic device, Display the first image, and Receiving a first user input to execute an editing function of the first image, and A storage medium that generates a second image by editing at least a portion of the first image using an artificial intelligence (AI) model learned based on said personal metadata, based on the existence of said personal metadata associated with the first image.

Images