Role visual data generation method and device, equipment and medium

Abstract

The embodiment of the invention discloses a role visual data generation method and device, equipment and a medium, and relates to the technical field of artificial intelligence. The method comprises the steps of obtaining description information of a to-be-created target role; the description information is used for indicating setting requirements of the target role; based on the description information, generating first text information which corresponds to the target role and is used for describing role setting; based on the first text information, generating second text information which corresponds to the target role and is used for describing visual data of the role; generating role visual data of the target role based on the second text information and the role style; the role style is used for indicating the design theme of the target role. According to the embodiment of the invention, the user can easily create complex and specific role setting, and the user can generate the visual data with consistent styles and features by using the visual data of the role again in the subsequent creation process, so that the consistency of the visual data and the continuity of creation are ensured, and the visual effect of the visual data is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to a method, device, equipment and medium for generating character visual data. Background Art

[0002] Wenshengtu technology is an artificial intelligence technology that converts text descriptions into images. Through Wenshengtu technology, the demand for character images can be directly sent to the program, allowing the program to generate the required character images. During the user's creation process, images containing a certain character may be generated multiple times. In multiple creations, the user's requirements for the character's posture and action in each generated character image may be different, resulting in problems such as poor consistency of character images and different styles in the multiple generated character images, affecting the coherence of the creation and the visual effects of the character images, resulting in a poor user experience. In addition, the specific application scenarios and application designs of large artificial intelligence models are also hot topics and difficulties that everyone is currently discussing. Summary of the invention

[0003] In view of the above problems in the prior art, the embodiments of the present disclosure provide a method, device, equipment and medium for generating character visual data. The embodiments of the present disclosure enable users to easily create complex and specific character settings, and users can reuse the character visual data in subsequent creation processes to generate character visual data with consistent style and characteristics, thereby ensuring the consistency of the character visual data and the coherence of the creation, and improving the visual effect of the character visual data.

[0004] To achieve the above objectives, the present disclosure provides, in a first aspect, a method for generating character visual data, comprising:

[0005] Obtaining description information of the target role to be created; wherein the description information is used to indicate setting requirements of the target role;

[0006] Based on the description information, generating first text information corresponding to the target role and used to describe the role setting;

[0007] Based on the first text information, generating second text information corresponding to the target character and used to describe the character's visual data;

[0008] Based on the second text information and the character style, character visual data of the target character is generated; wherein the character style is used to indicate a design theme of the target character.

[0009] As a possible implementation of the first aspect, the method further includes:

[0010] In response to a first processing request of a user for character visual data, obtaining an updated character style;

[0011] First updated character visual data of the target character is generated according to the second text information and the updated character style.

[0012] As a possible implementation of the first aspect, the method further includes:

[0013] In response to a second processing request of the user for the character visual data, obtaining updated description information;

[0014] Based on the updated description information, generating updated first text information corresponding to the target character and used to describe the character setting;

[0015] Based on the updated first text information, generating updated second text information corresponding to the target character for describing the character visual data;

[0016] Second updated character visual data of the target character is generated according to the updated second text information and the character style.

[0017] As a possible implementation of the first aspect, a process of generating first text information corresponding to a target role and used to describe role settings based on the description information includes:

[0018] Get the character style;

[0019] According to the description information and the character style, first text information corresponding to the target character and used to describe the character setting is generated.

[0020] As a possible implementation manner of the first aspect, obtaining description information of a target role to be created includes:

[0021] In response to a user's request to create visual data of a character, a preset control is displayed to the user; wherein the preset control includes a description information input control and a description information selection control, and the description information selection control includes at least one recommended description information of the character for the user to select;

[0022] Based on the preset controls displayed to the user, obtain description information of the target role to be created.

[0023] As a possible implementation of the first aspect, the method further includes:

[0024] The character visual data is processed to obtain the visual avatar of the target character.

[0025] As a possible implementation of the first aspect, the method further includes:

[0026] In response to a user's operation request for the first text information and / or the character visual data, the first text information, the second text information, the character style, and the character visual data are stored correspondingly; and / or,

[0027] The first text information and the character visual data are added to the user's community list.

[0028] A second aspect of the present disclosure provides a device for generating character visual data, comprising:

[0029] The acquisition unit is used to: acquire description information of the target role to be created; wherein the description information is used to indicate the setting requirements of the target role;

[0030] A first generating unit is used to: generate first text information corresponding to the target role and used to describe the role setting based on the description information;

[0031] A second generating unit is used to: generate second text information corresponding to the target character and used to describe the character visual data based on the first text information;

[0032] The third generating unit is used to generate character visual data of the target character based on the second text information and the character style; wherein the character style is used to indicate a design theme of the target character.

[0033] As a possible implementation of the second aspect,

[0034] The acquisition unit is further used to: acquire an updated character style in response to a first processing request of a user for character visual data;

[0035] The third generating unit is further used to generate first updated character visual data of the target character according to the second text information and the updated character style.

[0036] As a possible implementation of the second aspect,

[0037] The acquisition unit is further used to: in response to a second processing request of the user for the character visual data, acquire updated description information;

[0038] The first generating unit is further used to: generate updated first text information corresponding to the target character and used to describe the character setting based on the updated description information;

[0039] The second generating unit is further used to: generate updated second text information corresponding to the target character and used to describe the character visual data based on the updated first text information;

[0040] The third generating unit is further used to generate second updated character visual data of the target character according to the updated second text information and the character style.

[0041] As a possible implementation manner of the first aspect, the first generating unit is configured to:

[0042] Get the character style;

[0043] According to the description information and the character style, first text information corresponding to the target character and used to describe the character setting is generated.

[0044] As a possible implementation manner of the second aspect, the acquiring unit is configured to:

[0045] In response to a user's request to create visual data of a character, a preset control is displayed to the user; wherein the preset control includes a description information input control and a description information selection control, and the description information selection control includes at least one recommended description information of the character for the user to select;

[0046] Based on the preset controls displayed to the user, obtain description information of the target role to be created.

[0047] As a possible implementation manner of the second aspect, the apparatus further includes a processing unit, configured to:

[0048] The character visual data is processed to obtain the visual avatar of the target character.

[0049] As a possible implementation manner of the second aspect, the apparatus further includes a storage unit, configured to:

[0050] In response to a user's operation request for the first text information and / or the character visual data, the first text information, the second text information, the character style, and the character visual data are stored correspondingly; and / or,

[0051] The first text information and the character visual data are added to the user's community list.

[0052] A third aspect of the present disclosure provides an electronic device, including:

[0053] A memory for storing a computer program product;

[0054] The processor is used to execute the computer program product stored in the memory, and when the computer program product is executed, the method described in any one of the first aspects above is implemented.

[0055] A fourth aspect of the present disclosure provides a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the method described in any one of the first aspects above is implemented.

[0056] The fifth aspect of the present disclosure provides a computer program product, including computer program instructions, which, when executed by a processor, implement any of the methods described in the first aspect. The technical solution of the present disclosure is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0058] The present disclosure may be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

[0059] Figure 1 A flowchart of an embodiment of a method for generating character visual data disclosed herein;

[0060] Figure 2 A character creation flow chart for one embodiment of a method for generating character visual data disclosed herein;

[0061] Figure 3 A character creation flow chart for one embodiment of a method for generating character visual data disclosed herein;

[0062] Figure 4 It is a structural schematic diagram of an embodiment of a device for generating character visual data disclosed in the present invention;

[0063] Figure 5 It is a structural schematic diagram of an embodiment of a device for generating character visual data disclosed in the present invention;

[0064] Figure 6 A block diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0065] Although the embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0066] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0067] It should be understood that although the terms "first", "second", "third", etc. may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0068] The following first introduces the existing method, and then introduces the technical solution of the present invention in detail.

[0069] In an example of image generation, during the user creation process, for a certain character, images containing the character may be generated multiple times. For example, in a game production project, it is necessary to generate images with different postures and actions for the same character in multiple different scenes. Since the character image generated each time may have different postures and actions, the image consistency of multiple character images is poor and the styles are different, resulting in large differences in the visual effects of the characters and a poor gaming experience.

[0070] In the disclosed embodiment, through the multi-stage text generation of the artificial intelligence AI big model, it is possible to generate a role setting and role visual data that meet user needs and have high effects based on simple user input, and then based on the role setting and role visual data, multiple visual data with high image consistency of the role can be generated. Among them, the visual data includes static image data and dynamic video data. Therefore, the disclosed embodiment not only solves the problem of poor consistency of multiple generation of role images in the prior art, but also fills the gap that the prior art cannot generate role videos with good role consistency, so that users have a better experience in use. Specifically, firstly, a professional and story-telling role setting is generated based on the simple description information of the target role to be created by using a large language big model. The professional role setting provides a more detailed description of the requirements, has better interpretability and accessibility, and can generate output that better meets user needs on this basis. Then, the large language big model is used to generate text information that is more conducive to the understanding of the multimodal big model based on the role setting, so that the multimodal big model that can generate images and videos can quickly generate high-quality role visual data based on the text information. Furthermore, the generated character-related text information and character visual data are stored correspondingly, so that in the subsequent creation process, multiple high-quality visual data of the character in different scenes can be generated based on the stored high-quality character-related text information and character visual data. In addition, since the multiple different visual data of the character are generated based on the same character-related text information and character visual data and have the same image characteristics, stable and highly consistent arbitrary visual data of the character can be generated, thereby solving the problem of poor consistency and different styles of character images mentioned in the prior art.

[0071] Figure 1 FIG. 1 is a flow chart of an embodiment of a method for generating character visual data of the present disclosure. Figure 1 As shown, it may specifically include:

[0072] Step S110, obtaining description information of the target role to be created, wherein the description information is used to indicate the setting requirements of the target role. The target role to be created includes human images, animal images, static object images (such as tables, ships, buses, etc.), etc.

[0073] In one example, a user can input descriptive information for a target character in an interactive interface of a creation tool for generating a character image. The data types of the descriptive information include text and voice, that is, the descriptive information input by the user can include text or voice information for indicating the character setting requirements. For example, a user inputs text information in an input box of the interactive interface: "I want a righteous hero", and then the creation tool receives the user's descriptive information for the target character through the input box. For another example, the user can click the microphone icon next to the input box to switch to voice input mode, and then start inputting the descriptive information for the target character in the form of voice information. The creation tool converts the received voice information into text information, thereby obtaining the descriptive information of the target character.

[0074] Step S120, based on the description information, generates first text information corresponding to the target character for describing the character setting. The first text information includes character story information and basic information. The character story information is used to describe the social environment in which the character is located, the character's social relations, and the character's personality, skills, etc. The character basic information is used to describe the character's species, appearance, and other basic information. Optionally, the character story information includes at least one of the following: character prop information, character personality information, character skill information, or character social relationship information, and the basic information includes at least one of the following: character attribute information such as character name, character gender, character age, character height, or character weight.

[0075] Optionally, in order to facilitate the understanding of the large model when subsequently generating image description information based on the first text information, the format of the first text information is a preset standard format. At the same time, the preset standard format is also conducive to the user's understanding of the first text information, so that the user can quickly determine whether the content of the first text information meets the needs, and the matching degree between the content of the first text information and the user's needs is the basis for the subsequent generation of the current character image, so the first text information in the preset standard format can make the subsequently generated character image better serve the user's expectations. Exemplarily, the first part of the first text information is the character story information, the second part is the basic information, and the third part is the appearance description, and the basic information of the second part includes the character name, character age, and character gender in sequence, and the appearance description of the third part includes the character height or character weight, etc.

[0076] Optionally, based on the description information of the voice type or text type input by the user, the first text information is generated by the multimodal large model. The first text information processed by the multimodal large model has better interpretability and accessibility, can better reflect the user's needs, and can generate output that better meets the user's needs on this basis in the subsequent steps. For example, the first text information is generated by using LLM (Large Language Model) technology. In view of the strong text processing ability of the large language model, in the optional embodiment, the large language model is combined with the user's description information for the target role to be created. The generated first text information can be more in line with the semantics of the description information input by the user, and the user's experience in the process of generating the role image is improved. Specifically, the description information of the text type is input into the large language model, and the description information is processed by the large language model to generate the first text information for the role setting. The large language model can deeply understand the context and semantics of the language, ensuring that a professional and story-telling role setting, i.e., the first text information, can be generated based on the simple input of the user.

[0077] Optionally, the process of generating first text information corresponding to the target character for describing the character setting based on the description information further includes: obtaining a character style, wherein the character style is used to indicate a design theme of the target character. Based on the description information and the character style, the first text information corresponding to the target character for describing the character setting is generated. Through this process, the theme desired by the user can be preliminarily determined based on the obtained character style, and then under this theme, the first text information further generated based on the description information will better meet the user's expectations.

[0078] For the above optional embodiment, a specific example is as follows: after the user inputs "I want a decent hero", the large language model generates the first text information corresponding to the role based on the description information, and the generated first text information for role setting is:

[0079] A) Character Story Information: In pursuit of the "Key of Light" that can rebuild world order, Monsoon crosses the wasteland and fights alien creatures and greedy predators.

[0080] B) Basic Information

[0081] Character name: Monsoon; Age: 27; Gender: Male.

[0082] C) Appearance Description

[0083] Height: 182cm; Weight: 78kg.

[0084] Step S130: generating second text information corresponding to the target character and used to describe the character's visual data based on the first text information.

[0085] The first text information is data that is easy for users to understand, has a strong story, and is consistent with the user's expectations for descriptive information. However, the first text information is relatively weak in information content for the multimodal large model that can generate character visual data. Therefore, through step S130, the first text information is converted into second text information that can be directly used by the multimodal large model, which greatly simplifies the user's operation in the character visualization process, so that the user does not need to understand the complex writing skills of data that can be directly used by the multimodal large model to obtain high-quality character visual data, which not only lowers the technical threshold, but also expands the acceptability and scope of use of creation tools.

[0086] Optionally, the large language model technology is used to convert the first text information describing the character setting into the second text information that can be directly used by the multimodal large model. Furthermore, the second text information can be directly input into the multimodal large model as a prompt word in the subsequent steps, and the character visual data can be generated after being processed by the multimodal large model. In view of the excellent performance of the large language model in application scenarios such as understanding context and generating text, the second text information generated by the large language model is used as a prompt word, which can directly guide the multimodal large model to generate content and ensure that the generated content is more in line with user needs.

[0087] Step S140, based on the second text information and the character style, character visual data of the target character is generated. The character style is used to indicate the design theme of the target character. The character visual data includes static image data and dynamic video data. In one example, the user can input the character style for the target character in the interactive interface of the creation tool that generates the character visual data. In another example, the character style can be determined by default, that is, no user input is required. The data types of the input character style include text and voice, that is, the character style input by the user may include text information or voice information for indicating the design theme of the target character. For example, the character style may include cartoon style, realistic style, Japanese and Korean style, European and American style, Pixar style, Hayao Miyazaki style, etc. Further, the theme of the character design may be pre-set in the creation tool, or it may be automatically set by the user according to the needs. The user can input or select the character style of the target character in the interactive interface of the creation tool, and the creation tool obtains the character style of the target character in response to the user's selection operation.

[0088] Optionally, the second text information and the character style are input into the multimodal large model, and after being processed by the multimodal large model, character visual data with high-quality effects required by the user is generated.

[0089] Among them, the character visual data may include a character picture or a character video. The generated character visual data can be used as a typical and representative image of the target character, and can be reused to generate multiple visual data with consistent characteristics of the character in different scenes based on the character visual data. In the embodiment of the present disclosure, the first text information generated in step S120 for describing the character setting and the character visual data generated in step S140 can be stored correspondingly. In the subsequent creation process, the multiple visual data of the character are generated based on the stored character visual data, so the consistency of the character image and the coherence of the creation can be maintained.

[0090] In summary, the disclosed embodiments provide a simplified interactive method, which generates character settings and character visual data that meet user needs and have high quality based on simple user input through multi-stage text generation of an artificial intelligence AI big model, allowing users to easily create complex and specific characters, improve the quality and depth of character settings, and increase user satisfaction during the creation process. In addition, the multimodal big model is further used to generate stable and highly consistent character visual data based on the character settings, so that users can reuse the character visual data in the subsequent creation process to generate multiple high-quality visual data with consistent style and characteristics, ensuring the consistency of the character image and the coherence of the creation, thereby improving the visual effect of the character visual data.

[0091] In one embodiment, Figure 1 Step S110 in the step of obtaining description information of the target role to be created includes:

[0092] In response to a user's request to create visual data of a character, a preset control is displayed to the user; wherein the preset control includes a description information input control and a description information selection control, and the description information selection control includes at least one recommended description information of the character for the user to select;

[0093] Based on the preset controls displayed to the user, obtain description information of the target role to be created.

[0094] After the user submits a request to create visual data of a character in the interactive interface of the creation tool, the creation tool receives the description information and character style of the target character to be created indicated by the user through the preset control. The preset control includes a description information input control, which is set in the interactive interface and is used to receive the description information of the target character input by the user. Specifically, in response to the user's input operation in the description information input control, the user's input information is received; and the user's input information is used as the description information of the target character. The preset control also includes a character style input control, which is set in the interactive interface and is used to receive the character style of the target character input by the user. Specifically, in response to the user's input operation in the character style input control, the user's character style is received; and the user's input information is used as the character style of the target character.

[0095] Among them, the description information input control may include one of a single-line input box, a multi-line input box, and a text field. In the setting of the description information input control, a specific attribute may be added to enable the voice recognition function, so that the description information input control can receive not only text information but also voice information, creating a more user-friendly experience. Similarly, the character style input control may also include one of a single-line input box, a multi-line input box, and a text field, and a specific attribute may be added to enable the voice recognition function.

[0096] The preset control also includes a description information selection control, which is set in the interactive interface and is used to receive the recommended description information of the target role selected by the user. Specifically, in response to the user's request to create the role visual data, the description information selection control is displayed to the user, and the recommended description information of at least one role for the user to select is specifically displayed; in response to the user's selection operation on the recommended description information, the recommended description information selected by the user is used as the description information for the target role.

[0097] The preset controls also include a character style selection control, which is set in the interactive interface and is used to receive the character style of the target character selected by the user. Specifically, in response to the user's request to create character visual data, the character style selection control is displayed to the user, specifically displaying at least one character style for the user to select; in response to the user's selection operation on the character style, the character style selected by the user is used as the character style of the target character.

[0098] The description information selection control and the character style selection control may include a drop-down list or a selection pop-up window, which is used to respond to user requests and display pre-set options to users. With the description information selection control and the character style selection control, users do not need to enter text information or voice information, but directly select options that meet their needs from the list or pop-up window. This interactive method is more convenient and further improves the user experience.

[0099] In one embodiment, Figure 1 In step S120, based on the description information, first text information corresponding to the target role and used to describe the role setting is generated, including:

[0100] The description information is input into a first processing module based on a multimodal large model, and the first text information is output through the first processing module.

[0101] Among them, a first processing module can be constructed based on the multimodal large model, so that it can generate professional and story-telling character settings based on simple user input. In the first processing module, the above user description information is converted into the first text information for describing the character setting, which includes character story information, basic information, and appearance description.

[0102] In one example, the user's description is: "I want a decent hero". The description is input into the first processing module, and the first text information in the standard format output is as follows:

[0103] A) Character story information

[0104] In the wasteland where the end of the world has come in 2091, Ji Feng, who calls himself the "Doomsday Knight", wears armor made from discarded materials and holds a long sword rescued from an antique shop. He used to be an engineer and knows a thing or two about future technology, but now he chooses to use the ancient way to protect the remaining justice. Ji Feng's appearance gives people a cold and unsmiling atmosphere, but in fact, he is full of deep nostalgia for the past civilization in his heart. In order to pursue the "Key of Light" that can rebuild the world order, Ji Feng crosses the wasteland and fights against alien creatures and greedy predators. His figure flickers in the dusk dust, just like the gradually shattered but tenacious hope.

[0105] B) Basic Information

[0106] Name: Monsoon

[0107] Age: 37

[0108] Gender: male

[0109] Birthday: 11 / 14

[0110] Zodiac Sign: Scorpio

[0111] C) Appearance Description

[0112] Height: 182cm

[0113] Weight: 78kg

[0114] In the above example, the user only needs to input simple role description information, and the first processing module can automatically generate the role setting corresponding to the role, converting the user's simple role setting requirements into concrete and professional first text information. The entire interactive process is extremely simple, which reduces the difficulty of creation and improves the efficiency of creation.

[0115] In one embodiment, Figure 1 Step S130 in the method of generating second text information corresponding to the target character and used to describe the visual data of the character based on the first text information includes:

[0116] The first text information is input into a second processing module based on a large language model, and the second text information is output through the second processing module.

[0117] In the previous step, the first processing module generates the first text information for describing the character setting. In the creation tool, a second processing module can be constructed based on the large language model, so that it can generate the second text information for describing the character's visual data based on the first text information. In the subsequent steps, the second text information can be used as a prompt word to generate the character's visual data.

[0118] Due to their limited professional capabilities, ordinary users cannot convert role settings into prompt words that can be used by the multimodal large model. The disclosed embodiment generates prompt words that can be used for the multimodal large model through the large language model technology, which reduces the user's operation difficulty.

[0119] In one embodiment, Figure 1 Step S140 in the step of generating character visual data of the target character based on the second text information and the character style includes:

[0120] The second text information and the character style are input into a third processing module based on the multimodal large model, and the character visual data of the target character is output through the third processing module.

[0121] In the previous step, the second processing module generates a second text information for describing the character's visual data. On this basis, a third processing module can be constructed based on the multimodal large model, so that it can generate character visual data based on the second text information and the character style input or selected by the user. Among them, the multimodal large model can process and understand a variety of different forms of information, including text, images, videos, and audio. This ability enables the multimodal large model to perform well in generating content that is highly consistent with the prompt words. Whether it is generating images and videos with rich details and diverse content, or maintaining the consistency of characters and identity subjects, the multimodal model can effectively achieve these goals. Based on the multimodal large model, realistic character visual data with a high degree of consistency with the prompt words can be generated, so that the creation tool can more accurately generate corresponding images or videos based on the description information input by the user.

[0122] In summary, in the implementation process of generating character visual data provided by the present disclosure, the user only needs to input simple character description information, and the character setting corresponding to the character can be automatically generated after processing by the first processing module. The prompt words are automatically generated based on the character setting by the second processing module, and then the character visual data is generated by the third processing module based on the prompt words. The entire interactive process is extremely simple. Furthermore, a storage system containing character text descriptions and image information is established to store the character settings and character visual data corresponding to the character. In the subsequent creation process, the user generates multiple high-quality visual data of the target character in different scenes based on the above stored character information, and a series of stable image or video data with high consistency can be obtained.

[0123] In one embodiment, the above method further comprises:

[0124] In response to a user's operation request for the first text information and / or the character visual data, the first text information, the second text information, the character style, and the character visual data are stored correspondingly; and / or,

[0125] The first text information and the character visual data are added to the user's community (studio) list.

[0126] The operation request is a user's confirmation operation on the first text information and / or the character visual data.

[0127] exist Figure 1In the process shown, the first text information and the character visual data are generated in step S120 and step S140 respectively. After the character visual data is generated, the target character required by the user can be created based on the first text information and the character visual data. Specifically, a storage system for storing character text descriptions and image information is pre-established. In one implementation, after the first text information is generated and the user's confirmation operation on the first text information is received, a storage unit for storing the target character information is allocated in the storage system, and the generated first text information is written into the storage unit. After the character visual data of the character is subsequently generated and confirmed by the user, the second text information, character style, and character visual data of the character are correspondingly stored in the storage unit. In another implementation, after the character visual data is generated and the user's confirmation operation on the character visual data is received, a storage unit for storing the target character information is allocated in the storage system, and the generated first text information, second text information, character style, and character visual data are written into the storage unit.

[0128] In the creation tool, each user can create his or her own studio and add the created characters to the user's own studio list, which is convenient for users to search, continue to create, and share characters, thereby improving the user experience. Specifically, the first text information and character visual data are added to the user's community (studio) list. In an example of a creation tool that supports story making and game development, each character or story created by each user using the creation tool can be used as a separate studio. The characters or stories created by the user are added to the user's own studio list by default. All users can arbitrarily select characters or stories created by others and add them to their own studio list. Subsequent users can search for the added characters in the studio list and continue to create based on the added characters.

[0129] In one embodiment, the method further includes: processing the character visual data to obtain a visual avatar of the target character. Based on the character visual data, a visual avatar corresponding to the target character can be cropped. The visual avatar is a representative picture of the target character. Displaying and using the visual avatar facilitates subsequent search and character sharing by users, thereby improving user experience.

[0130] In one embodiment, the method further includes: generating keywords for the target character based on at least one of the description information of the target character, the first text information, the second text information, and the character visual data of the target character. Specifically, at least one of the description information of the target character, the first text information, the second text information, and the character visual data of the target character can be input into a multimodal large model, and the multimodal large model processes the input information to generate keywords for the target character. The multimodal large model can deeply understand and integrate data from different modalities (such as text, images, audio, video, etc.), and can generate highly relevant and representative character keywords for subsequent user retrieval.

[0131] Figure 2 This is a character creation flow chart of an embodiment of the method for generating character visual data of the present disclosure. Figure 2 As shown in the figure, when the process starts running, the description information of the target role input by the user is received. Figure 2 In "Step 001: Receive user input information", the user enters a simple command, such as "I want a decent hero". Each time the user input information is received, the following three agents are triggered in sequence:

[0132] (1) role_profile (role setting) Agent

[0133] The Agent is the first processing module, which is built based on the multimodal large model. Its function is to expand the user's description information into a plain text role_profile (role setting) in a standard format, that is, the first text information. In an example, the process can support users to modify role settings through multiple rounds of dialogue.

[0134] (2) role_design_description (role visual data description) Agent

[0135] The Agent is the second processing module, built on the Large Language Model (LLM). Its function is to generate a plain text role_design_description (role visual data description), i.e. the second text information, based on the role_profile (role setting). Every time the role_profile changes, it will be sent as input to the role_design_descriptionAgent.

[0136] (3) roledesign (character visual data) Agent:

[0137] The Agent is the third processing module, built on the multimodal large model. Its function is to generate role_design (role visual data) based on role_design_description (role visual data description). Every time role_design_description changes, it will be sent as input to role_designAgent.

[0138] In one example, when any Agent is running, the user is not allowed to enter new instructions. After the three Agents have completed their execution, the user can enter new instructions and modify the role settings through multiple rounds of dialogue. Based on the above reasonable process design, the process can run smoothly, while providing a user-friendly interaction method to improve the user experience.

[0139] like Figure 3 As shown, an exemplary role creation process includes the following steps:

[0140] Step S1, receiving the description information and character style of the target character from the user. Specifically, the creation tool can receive the description information of the target character input by the user in the input box of the interactive interface, allowing the user to describe the character he wants to create with simple text. For example, the user inputs: "I want a supermodel". See also Figure 2 In actual application, one method is to receive the description information input by the user in step 001, and another method is to not require the user to input specific text as the description information in the interactive interface, but to Figure 2 "Step 002: Associative Bubble" in the example above displays at least one recommended description information of a role for the user to select, so that the user can select the desired character description information from multiple recommended description information. The recommended description information displayed in the "Associative Bubble" can include the basic description and scene of the role, such as: a supermodel, an old man reading in a library, a beggar begging on the street, etc. After the user selects a recommended description information, the subsequent character setting generation process will be triggered. The character setting generation process is built based on a multimodal big model. The multimodal big model technically supports a variety of generation results. After the user selects any description information each time, the generated result can change randomly. For example, if a supermodel is selected, the character design of the supermodel generated each time can be different, and the final character visual data can also be different. Based on the multimodal big model, richer and more diverse creative content can be generated to enhance the user's creative experience.

[0141] See also Figure 2, the user also needs to determine the role style of the target role. The interactive interface of the creation tool displays multiple preset role styles, and the user can select the role style of the target role from the preset multiple styles. The user can use the default role style, or modify the default role style to the role style he wants in "Step 003: Manually modify role style", and then input the role style confirmed by the user into the roledesignAgent module.

[0142] Step S2, generate role settings. Figure 2 In step 001, the description information input by the user is received, or in step 002, the description information selected by the user is received. The description information input or selected by the user is input into role_profileAgent as the user requirement. In role_profileAgent, a role setting (role_profile) in a standard format is generated through the multimodal large model technology. The role setting may also include the role name to facilitate the user's subsequent search and role sharing. Then the role_profile is input into role_design_descriptionAgent.

[0143] The character setting in the standard format is convenient for users to obtain character information on the one hand, and it is also convenient to generate prompt words consistent with the character information through the large language model technology on the other hand. In addition, after the character is successfully created, the character setting can also be used as a reference when generating images of various postures and actions of the character again in the subsequent creation process. The character setting in the standard format is more conducive to model understanding. The generated results after model processing have a consistent style, and the generated character image is more consistent.

[0144] Step S3, generating a character visual data description. Figure 2 Based on the role setting, the role visual data description (role_design_description) that can be used as the input of the multimodal large model is generated in role_design_descriptionAgent through the large language model technology, that is, the prompt word of the multimodal large model. Then role_design_description is input into role_design_Agent as the prompt word.

[0145] Step S4, generating character visual data. Figure 2 ,In roledesignAgent, the character visual data (roledesign) is generated through a multimodal large ,model based on the character visual data description and the ,character style selected or input by the user.

[0146] Step S5, creating a character and cropping the character portrait. Create a character based on the character setting and the character visual data, and then crop according to the character visual data. After cropping, a representative image of the target character is generated, that is, the visual portrait of the target character. Figure 2 After roledesignAgent generates the character visual data, execute "Step 004: Create character and crop avatar".

[0147] Step S6, adding the created role to the user's studio (community). Figure 3 After the character is generated in step S5, the user can choose whether to continue to adjust the character and switch the historical version in steps S52 and S54. If the user chooses not to continue to adjust the character and not to switch the historical version, that is, the user confirms the current version, the character is successfully created, and the user continues to step S6 to add the created character to his studio list (address book). The information stored in the studio list may include character settings, visual avatars, and character visual data.

[0148] See also Figure 2 After "Step 004: Create a role, crop the avatar", "complete a round of process", if the user confirms the role, execute "Step 005: Save the role version information based on the final profile, design_description, and design", and "Step 006: Enter role_studio accordingly" to add the created role to the user's studio (community). The user's studio list can be displayed in the interactive interface of the creation tool. In the studio list, the role settings and corresponding visual avatars can be displayed to the user, which is convenient for the user to search, continue to create, and share roles, thereby improving the user experience.

[0149] Step S7, adjusting the character settings through multiple rounds of dialogue and reselecting the style.

[0150] In one embodiment, the method for generating the character visual data further includes: obtaining the user's redescription information of the target character, adjusting the generated first text information based on the redescription information; and / or generating a new version of the character visual data based on the second text information and the character style reselected by the user.

[0151] See also Figure 3In the process shown, in step S52, the user can choose whether to continue adjusting the character. If the user chooses to continue adjusting the character, step S7 is executed to adjust the character settings in multiple rounds through dialogue, and to reselect the style, and then return to step S1 to receive the user's redescription information for the target character and the reselected character style, and repeat steps S1 to S5. Specifically, in step S2, the character settings are regenerated based on the redescription information of the target character, that is, the first text information is adjusted. In step S3, a new character visual data description is generated based on the adjusted character settings. In step S4, a new version of the character visual data is generated based on the new character visual data description and the character style reselected by the user.

[0152] In one embodiment, the above method further comprises:

[0153] In response to a second processing request of the user for the character visual data, obtaining updated description information;

[0154] Based on the updated description information, generating updated first text information corresponding to the target character and used to describe the character setting;

[0155] Based on the updated first text information, generating updated second text information corresponding to the target character for describing the character visual data;

[0156] Second updated character visual data of the target character is generated according to the updated second text information and the character style.

[0157] The second processing request includes a request from a user to modify the description information.

[0158] See also Figure 2 In the process shown, after "completing one round of the process", multiple rounds of dialogue modification are allowed, and the process returns to "Step 001: Receive user input information". The user can adjust the role settings in multiple rounds through dialogue, that is, modify the role settings through role_profileAgent and adjust the generated first text information. After the role settings are changed, the subsequent steps will continue to be executed. The prompt words of the multimodal large model generated by role_design_descriptionAgent will also change, and finally the character visual data generated by roledesignAgent will also change. Every time the user modifies the role settings, the three agents in the flowchart will be re-executed. After the three agents are executed, if the user needs to continue adjusting the role, the next round of user dialogue can be started. Alternatively, if the user does not need to continue adjusting the role, confirm the role and add it to his own studio.

[0159] In one example, the description information input by the user is: "Chang'e", and the first text information in a standard format generated by the large language model based on the description information is:

[0160] A) Role Setting: Chang'e is a legendary goddess of the moon, often depicted as beautiful and gentle, and living in the Moon Palace. She has an otherworldly temperament, wearing a silver-white dress with hair flowing down her shoulders. Chang'e was originally a mortal, but after drinking the elixir of immortality, she was forced to separate from her beloved Hou Yi and live alone in the cold Moon Palace.

[0161] B) Basic Information

[0162] Name: Chang'e

[0163] Age:3000

[0164] Gender: female

[0165] Birthday: 09 / 15

[0166] Zodiac Sign: Virgo

[0167] C) Appearance Description

[0168] Height: 168CM

[0169] Weight:50KG

[0170] After the first text information is generated, if the user wants to further adjust the role settings, he can submit a request to modify the description information. In response to the request, the next round of dialogue is started in the interactive interface to modify it. The user re-enters the modified description information as "Chang'e on the moon", and the large language model generates the modified first text information in the standard format based on the description information as follows:

[0171] A) Role Setting: Chang'e is a goddess living in the Moon Palace in Chinese mythology. She was isolated from the world after accidentally drinking the elixir of immortality. Her story is full of mystery and sorrow, often reminiscent of the contemplation of eternal loneliness. She lives in the vast, cold and cold Moon Palace, lonely and elegant. Chang'e wears a gorgeous long dress, her face is gentle, and her eyes hide a deep nostalgia for the past.

[0172] B) Basic Information

[0173] Name: Chang'e

[0174] Age:3001

[0175] Gender: female

[0176] Birthday: 09 / 15

[0177] Zodiac Sign: Virgo

[0178] C) Appearance Description

[0179] Height: 168CM

[0180] Weight:50KG

[0181] After the modified first text information is generated, the second text information subsequently regenerated based on the modified first text information and the final regenerated character visual data will change adaptively, so that the final generation result is more adapted to user needs in terms of content, words and visual perception.

[0182] In one embodiment, the above method further comprises:

[0183] In response to a first processing request of a user for character visual data, obtaining an updated character style;

[0184] First updated character visual data of the target character is generated according to the second text information and the updated character style.

[0185] The first processing request includes a request from a user to modify a style.

[0186] See also Figure 2 In the process shown, after "Completing a round of process", the user can also reselect the character style in "Step 007: Modify character style". The creation tool can provide users with many character styles and will continue to update the character style according to needs, so that users can create based on the selected character style. Figure 2 In "New style to role_design_description", the user's newly selected character style and the character visual data description are spliced ​​together. The name of the new style and the character visual data description can be spliced ​​into a description, and the spliced ​​description is input into the role_designAgent model. The role_designAgent model generates a new version of the character visual data based on the new input information. Then after processing in step 004, a round of the process is completed. If the user does not need to continue to modify the style, he can confirm the character and add it to his studio.

[0187] Step S8: Switch to the historical version. Figure 3 In the process shown, if the user chooses not to continue adjusting the role in step S52, he can choose whether to switch the historical version in the subsequent step S54. If the user chooses "yes", it switches back to the historical version the user wants. If the user chooses not to switch the historical version in step S54, step S6 is executed to add the created role to the user's studio (community).

[0188] See also Figure 2 In the process shown, after "Completing a round of processes", the user can choose to switch back to the intermediate version in "Step 008: Switching historical versions". After switching, the character settings and character visual data are switched back to the previously generated version, and the user can confirm the character and add it to his studio.

[0189] Step S9, storing the created character. Establish a storage system containing character text description and image information, and store the character settings, generated character visual data description, character visual data, and character style as the relevant text information and character visual data of the created character. Figure 3 In the process shown, after adding the created character to the user's studio (community) in step S6, step S9 is executed to store the relevant text information and character visual data of the created character in correspondence.

[0190] In the subsequent creation process, the user can directly select the desired designated role from the created characters. The creation tool obtains the relevant text information and character visual data of the designated role from the storage system based on the user's selection operation, and generates more creative content based on the relevant text information and character visual data of the character, for example, the story name and story data of the story to be created can be generated. The story data includes text description data and story visual data. In addition, multiple character visual data corresponding to the designated role can also be generated based on the relevant text information and character visual data of the character. The above-mentioned subsequently generated creative content is generated based on the relevant text information and character visual data of the created character, and has a consistent style and characteristics, so that the subsequent generation results have coherent content and consistent character image.

[0191] In summary, the disclosed embodiments can achieve the conversion from a simple description such as "I want a decent hero" to a complete character setting (including basic information, appearance description and character story information) by using multimodal big model technology, so that ordinary users can easily create complex and specific character settings without having to learn in-depth professional knowledge of character creation.

[0192] In addition, the disclosed embodiment greatly simplifies the user's operation in the character visualization process by automatically converting the character settings into prompt words that can be directly used by the multimodal large model. The user can obtain high-quality character visual data without having to understand complex prompt word writing skills, which lowers the technical threshold and expands the acceptability and scope of use of the creation tool.

[0193] Furthermore, by establishing a storage system that contains character text descriptions and image information, the creative tools can generate stable and highly consistent arbitrary visual data of the characters based on the stored created characters, ensuring the consistency and diversity of the character visual data, supporting a wider range of creative activities, and can be applied to story production or game development material production, thereby improving the practical value of the character visual data and the coherence of creation.

[0194] In the creation tool, a storage system including a character text description and image information can be established, and the first text information, the second text information, the first character visual data, and the character style can be stored as the character information of the created character. Based on the above pre-stored character information, a story name and story data can be further generated, wherein the story data includes text description data and story visual data. In addition, a corresponding story can also be created based on the character information and story data of the character, specifically including writing the generated story name and story data into the storage system, and adding the created story to the user's own studio list.

[0195] Optionally, the process of generating a story includes:

[0196] (1) Obtain at least one target role of the story to be created. The target role may be at least one role selected by the user from the community list for the story to be created.

[0197] A role list of created roles can be displayed to the user on the interactive interface of the creation tool. The role list includes at least one role in the pre-created studio list. The user can select at least one target role for the story to be created on the interactive interface of the creation tool.

[0198] In one example, the role list includes the following roles: upright hero, supermodel, beggar, magician, etc. The user selects "upright hero" in the role list, and the authoring tool obtains the target role selected by the user as "upright hero" through the role list.

[0199] (2) Obtaining first text information corresponding to at least one target role and used to describe role settings. Specifically, the target role may be searched in a storage system to obtain the first text information of the target role.

[0200] (3) Generate a story name for the story to be created based on the first text information. For example, a large language model technology can be used to input the first text information into the large language model, and the story name for the story to be created is generated after being processed by the large language model. The large language model can deeply understand the context and semantics of the language, and can generate a professional story name that can express the theme and emotion based on the first text information. For example, the target character selected by the user is "upright hero", and the story name of the story to be created generated by the large language model based on the target character is "Battle of Justice". Based on this step, a story name with a story overview can be generated, which not only allows users to access the story information more intuitively, but also facilitates subsequent story searches and multiple users to choose to join the story based on the story name.

[0201] (4) Generate story data of the story to be created according to the first text information and the story name. The types of story data used for story description setting include: text type, voice type, and visual type such as image or video.

[0202] Optionally, a multimodal large model technology can be used to input the first text information of at least one target character selected by the user and the story name generated in the above steps into the multimodal large model. The multimodal large model can not only understand the context and semantics of the language and the visual content of the image, but also express these contents in an appropriate text form. After being processed by the multimodal large model, story data with coherence and accuracy can be generated. In view of the strong text processing ability of the large language model, in the optional present embodiment, the combined application of the large language model with the first text information and the story name can make the generated story more in line with the user's expectations, and enhance the user's experience in the process of generating the story. Specifically, the first text information and the story name of the text type are input into the large language model, and the first text information and the story name are processed by the large language model to generate story data for the story description setting. The large language model can deeply understand the context and semantics of the language, ensuring that professional and story-like story data can be generated based on simple text information.

[0203] In summary, the disclosed embodiments provide a simplified interactive method, which only requires the user to select at least one target character, and does not require the difficult writing of prompt words. Through the multi-stage text generation technology of the artificial intelligence AI large model, a story with a specific character setting can be easily created. This method reduces the difficulty of participating in the creation and improves the user's satisfaction in the creation process. In addition, the relevant information corresponding to each character has been pre-stored in the storage system of the creation tool. The user can generate character visual data with consistent style and characteristics based on the relevant information of the same character in the subsequent creation process, ensuring the consistency of the character visual data and the coherence of the creation, and improving the visual effect of the character visual data.

[0204] The storage system including the created characters and the created stories established in the creation tool can store the first text information, the second text information, the first character visual data, and the character style of the created characters as the character information of the created characters, and can store the story name, story data, and the first text information, the second text information, the first character visual data of the characters included in the story. Based on the above pre-stored character information and story information, visual data of one or more characters can be further generated.

[0205] The process of generating corresponding visual data based on the created characters and stories includes:

[0206] (1) Obtaining character information of at least one designated character and instruction information corresponding to each designated character; wherein the character information includes at least one of the following: first text information for describing character settings, second text information for describing visual data of the first character, a character style for indicating a character design theme, and visual data of the first character; the instruction information is used to indicate an action of the designated character.

[0207] A character list may be displayed to the user on the interactive interface of the creation tool. The character list includes at least one character in the pre-created studio list. The user may select at least one designated character on the interactive interface of the creation tool. The character information of the designated character may include at least one of the following: first text information for describing the character setting, second text information for describing the visual data of the first character, a character style for indicating a character design theme, the visual data of the first character, etc.

[0208] In addition, a text input box can be set on the interactive interface of the creation tool to receive instruction information input by the user for indicating the action of the specified character. The user can enter simple text in the text input box to indicate the action of the specified character. For example, the user enters: "Running in the wilderness" in the input box corresponding to the specified character "The Honest Warrior". The creation tool then receives the user's instruction through the input box that the action corresponding to the "Honest Warrior" is "Running in the wilderness".

[0209] In another example, the instruction information input by the user may also include voice information. The user may click the microphone icon next to the input box to switch to voice input mode, and then start inputting instruction information in the form of voice information. The authoring tool converts the received voice information into text information, thereby obtaining the instruction information.

[0210] (2) Based on the character information of the designated character and the instruction information, third text information for describing the second character visual data corresponding to the designated character is generated.

[0211] The second character visual data includes an action image corresponding to the specified character. In order to generate an action image that matches the user's instruction information, a third text information for describing the action image may be first generated, and then the third text information may be used as a prompt word for the multimodal large model to guide the multimodal large model to finally generate an action image that matches the user's instruction information.

[0212] Specifically, the multimodal large model technology can be used to generate the third text information based on the role information of the specified role and the instruction information input by the user. The multimodal large model can not only understand the context and semantics of the language and the visual content of the image, but also express these contents in an appropriate text form. After processing by the multimodal large model, the third text information with clarity, relevance, specificity and guidance can be generated. Then the third text information is directly input into the multimodal large model as a prompt word, and after processing by the multimodal large model, the second role visual data corresponding to the specified role can be generated.

[0213] In this step, the role information and third text information of the specified role are directly converted into professional prompt words that can be directly input into the multimodal large model, thereby not requiring users to learn how to write prompt words, avoiding the complicated work of learning model prompt words, and not requiring users to have strong professional capabilities, greatly reducing the difficulty of user operation.

[0214] (3) Based on the role information and the third text information of the specified role, the second role visual data of the specified role is generated. The role information and the third text information of the specified role are input into the multimodal large model. The multimodal large model can deeply understand the context and semantics of the language and the visual content of the image. After being processed by the multimodal large model, the second role visual data corresponding to the specified role can be generated.

[0215] In the creation process of the above example, for the same designated character, the second character visual data with different postures and actions generated each time is generated based on the character information of the designated character. Therefore, different images of the same character contain the characteristics of the character setting of the character, thereby generating stable and highly consistent arbitrary visual data of the character, thereby maintaining the consistency of the character image and the coherence of the creation.

[0216] In summary, the disclosed embodiments provide a simplified interactive method that does not require the difficult writing of prompt words. The multimodal large model technology can generate corresponding second character visual data based on the user-specified role, increasing the user's satisfaction in the creation process. In addition, the second character visual data of the same role in different scenes are all generated based on the role information of the specified role, which can ensure the consistency of the visual data and the coherence of the creation, and improve the visual effect of the visual data.

[0217] like Figure 4As shown, the present disclosure also provides an embodiment of a corresponding device for generating character visual data. For the beneficial effects or technical problems solved by the device, please refer to the description of the method corresponding to each device, or refer to the description in the content of the invention, which will not be repeated here.

[0218] In an embodiment of the device for generating character visual data, the device comprises:

[0219] The acquisition unit 100 is used to: acquire description information of the target role to be created; wherein the description information is used to indicate the setting requirements of the target role;

[0220] The first generating unit 200 is used to: generate first text information corresponding to the target role and used to describe the role setting based on the description information;

[0221] The second generating unit 300 is used to: generate second text information corresponding to the target character and used to describe the character visual data based on the first text information;

[0222] The third generating unit 400 is used to generate character visual data of a target character based on the second text information and the character style; wherein the character style is used to indicate a design theme of the target character.

[0223] In one embodiment, the acquisition unit 100 is further used to: acquire an updated character style in response to a first processing request of a user for character visual data;

[0224] The third generating unit 400 is further used to generate first updated character visual data of the target character according to the second text information and the updated character style.

[0225] In one embodiment, the acquisition unit 100 is further used to: in response to a second processing request of the user for the character visual data, acquire updated description information;

[0226] The first generating unit 200 is further used to: generate updated first text information corresponding to the target character and used to describe the character setting based on the updated description information;

[0227] The second generating unit 300 is further used to: generate updated second text information corresponding to the target character and used to describe the character visual data based on the updated first text information;

[0228] The third generating unit 400 is further used to generate second updated character visual data of the target character according to the updated second text information and the character style.

[0229] In one implementation, the first generating unit 200 is further configured to:

[0230] Get the character style;

[0231] According to the description information and the character style, first text information corresponding to the target character and used to describe the character setting is generated.

[0232] In one implementation, the acquisition unit 100 is used to:

[0233] In response to a user's request to create visual data of a character, a preset control is displayed to the user; wherein the preset control includes a description information input control and a description information selection control, and the description information selection control includes at least one recommended description information of the character for the user to select;

[0234] Based on the preset controls displayed to the user, obtain description information of the target role to be created.

[0235] See also Figure 5 In one embodiment, the apparatus further comprises a processing unit 500, configured to:

[0236] The character visual data is processed to obtain the visual avatar of the target character.

[0237] See also Figure 5 In one embodiment, the above device further includes a storage unit 600, which is used to:

[0238] In response to a user's operation request for the first text information and / or the character visual data, the first text information, the second text information, the character style, and the character visual data are stored correspondingly; and / or,

[0239] The first text information and the character visual data are added to the user's community list.

[0240] Below, reference Figure 6 The electronic device according to the embodiment of the present disclosure is described. The electronic device may be any one or both of the first device and the second device, or a stand-alone device independent of them, and the stand-alone device may communicate with the first device and the second device to receive the collected input signals from them.

[0241] Figure 6 A block diagram of an electronic device according to an embodiment of the present disclosure is illustrated.

[0242] like Figure 6 As shown, the electronic device includes one or more processors and memory.

[0243] The processor may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0244] The memory may store one or more computer program products, and the memory may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc. One or more computer program products may be stored on the computer-readable storage medium, and the processor may run the computer program product to implement the various embodiments of the present disclosure described above and / or other desired functions.

[0245] In one example, the electronic device may further include: an input device and an output device, and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0246] In addition, the input device may also include, for example, a keyboard, a mouse, and the like.

[0247] The output device can output various information to the outside, including determined distance information, direction information, etc. The output device can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0248] Of course, to simplify, Figure 6 Only some of the components related to the present disclosure in the electronic device are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device may further include any other appropriate components.

[0249] In addition to the above methods and devices, an embodiment of the present disclosure may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the method according to various embodiments of the present disclosure described in the above part of this specification.

[0250] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0251] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, enables the processor to execute the steps of the method according to various embodiments of the present disclosure described in the above part of this specification.

[0252] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0253] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, and are not limitations. The above details do not limit the present disclosure to the necessity of adopting the above specific details to be implemented.

[0254] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0255] The block diagrams of the devices, apparatuses, equipment, and systems involved in this disclosure are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including," "comprising," "having," and the like are open words, referring to "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or," and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0256] The method and apparatus of the present disclosure may be implemented in many ways. For example, the method and apparatus of the present disclosure may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present disclosure are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present disclosure may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present disclosure. Therefore, the present disclosure also covers a recording medium storing a program for executing the method according to the present disclosure.

[0257] It should also be noted that in the apparatus, device and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0258] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0259] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A method for generating character visual data, characterized in that: include: Obtaining description information of a target role to be created; wherein the description information is used to indicate setting requirements of the target role; Based on the description information, generating first text information corresponding to the target character and used to describe the character setting; Based on the first text information, generating second text information corresponding to the target character and used to describe the character visual data; Based on the second text information and the character style, character visual data of the target character is generated; wherein the character style is used to indicate a design theme of the target character.

2. The method according to claim 1, characterized in that: The method further comprises: In response to a first processing request of a user for the character visual data, obtaining an updated character style; Generate first updated character visual data of the target character according to the second text information and the updated character style.

3. The method according to claim 1, characterized in that The method further comprises: In response to a second processing request of the user for the character visual data, obtaining updated description information; Based on the updated description information, generating updated first text information corresponding to the target character and used to describe the character setting; Based on the updated first text information, generating updated second text information corresponding to the target character and used for describing character visual data; Second updated character visual data of the target character is generated according to the updated second text information and the character style.

4. The method according to any one of claims 1 to 3, characterized in that: The process of generating first text information corresponding to the target role and used to describe role settings based on the description information includes: Obtaining the character style; According to the description information and the character style, first text information corresponding to the target character and used to describe the character setting is generated.

5. The method according to any one of claims 1 to 4, characterized in that: The step of obtaining description information of the target role to be created includes: In response to a user's request to create visual data of a character, a preset control is displayed to the user; wherein the preset control includes a description information input control and a description information selection control, and the description information selection control includes at least one recommended description information of the character for the user to select; Based on the preset control displayed to the user, description information of the target role to be created is obtained.

6. The method according to any one of claims 1 to 5, characterized in that: The method further comprises: The character visual data is processed to obtain a visual avatar of the target character.

7. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: In response to a user's operation request for the first text information and / or the character visual data, the first text information, the second text information, the character style, and the character visual data are stored correspondingly; and / or, The first text information and the character visual data are added to the user's community list.

8. A device for generating character visual data, characterized in that: include: An acquisition unit, used to: acquire description information of a target role to be created; wherein the description information is used to indicate setting requirements of the target role; A first generating unit is used to: generate first text information corresponding to the target role and used to describe role settings based on the description information; A second generating unit is used to: generate second text information corresponding to the target character and used to describe the character visual data based on the first text information; The third generating unit is used to generate character visual data of the target character based on the second text information and the character style; wherein the character style is used to indicate a design theme of the target character.

9. An electronic device, characterized in that: include: a memory for storing a computer program product; A processor is used to execute the computer program product stored in the memory, and when the computer program product is executed, it implements the method described in any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method described in any one of claims 1 to 7 is implemented.

Images