Virtual character construction method, device and medium
By introducing character construction attribute groups and visual element extraction models, standardized virtual character concept art is generated, solving the problems of low efficiency and low accuracy caused by manual interpretation in virtual character construction, and realizing an efficient and accurate character construction process.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- YIDIAN LINGXI INFORMATION TECHNOLOGY (GUANGZHOU) CO LTD
- Filing Date
- 2026-02-09
- Publication Date
- 2026-06-09
AI Technical Summary
In existing technologies, the process of building virtual characters relies on manual interpretation, resulting in long development cycles, low efficiency, and low accuracy, requiring repeated communication and modification by multiple departments.
By introducing a pre-defined set of character attributes, the character description text is decomposed into visual elements. The visual elements are used to extract models and generate image models to produce standardized character concept art, reducing human interpretation bias and information omission.
This improved the alignment between character concept art requirements and development needs, shortened the development cycle, reduced modification and iteration costs, and increased development efficiency and accuracy.
Smart Images
Figure CN122176132A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of computer technology, and more specifically, to a method, device, and medium for constructing virtual characters. Background Technology
[0002] With the rapid development of the digital content industry, the efficiency and accuracy of virtual character creation have become core factors affecting product development cycles. Currently, the traditional process of virtual character creation generally adopts a linear work model from copywriting and concept art design to concept art optimization. This model is highly dependent on close collaboration among multiple positions such as planners and art designers.
[0003] Specifically, in the aforementioned linear workflow, after planners output character description text in natural language format, professional art designers must subjectively interpret the text based on their personal experience and then manually create character concept art. If the generated character concept art does not match the product positioning or actual needs, repeated communication, confirmation, and modifications are required between the planning and art design departments. This iterative process based on subjective understanding not only significantly prolongs the development cycle and reduces efficiency but also results in low accuracy of the final virtual characters. Summary of the Invention
[0004] One objective of this disclosure is to provide a new technical solution for virtual character construction.
[0005] According to a first aspect of this disclosure, a method for constructing a virtual character is provided, comprising: Obtain the first character description text and the preset character construction attribute group for the virtual character construction; wherein, the character construction attribute group includes multiple visual attributes for defining the appearance of the virtual character; The visual element extraction model is invoked to construct each visual attribute in the attribute group for the character, and the visual element corresponding to the visual attribute is extracted from the first character description text; The extracted visual elements are integrated into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character.
[0006] Optionally, the method further includes: The video generation model is invoked to generate a display video of the target virtual character based on the original character artwork.
[0007] Optionally, the step of fusing the extracted visual elements into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character includes: Guided by the character building attribute group, the reference character portrait is converted into corresponding reference character description text; wherein, the reference character description text includes visual elements of the reference character corresponding to each visual attribute in the character building attribute group; The visual elements extracted from the first character description text are merged into the corresponding visual elements of the reference character description text to generate a second character description text about the construction of the virtual character; The image generation model is invoked to generate the original character artwork of the target virtual character based on the reference character portrait and the second character description text. Optionally, the reference character portraits include multiple ones, and the image generation model generates the original character artwork of the target virtual character based on the reference character portraits and the second character description text, including: Obtain the image fusion weights corresponding to the multiple reference character portraits; The image generation model is invoked to generate the original character artwork of the target virtual character based on the second character description text, multiple reference character images and their corresponding image fusion weights. Optionally, the character construction attribute group includes physical structure. The image generation model generates the original character artwork of the target virtual character based on the reference character portrait and the second character description text; The image adjustment model is invoked to adjust the shape and structure of the reference character portrait to match the visual elements of the extracted shape and structure, thereby obtaining the adjusted reference character portrait; The image generation model is invoked to generate the original character artwork of the target virtual character based on the adjusted reference character portrait and the second character description text.
[0008] Optionally, the method further includes: Extract target visual elements from the extracted visual elements; wherein, the target visual elements are the key visual elements among the visual elements used to characterize the appearance style of the virtual character. Based on the extracted target visual elements, a character portrait that matches the target visual elements is searched in the set character image database and used as the reference character portrait.
[0009] Optionally, the method further includes: Extract target visual elements from the extracted visual elements; wherein, the target visual elements are the key visual elements among the visual elements used to characterize the appearance style of the virtual character. Based on the extracted target visual elements, a character portrait that matches the target visual elements is searched in the set character image database and used as a reference character portrait.
[0010] Optionally, the step of fusing the extracted visual elements into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character includes: The extracted visual elements are integrated into the corresponding visual attributes of the reference character image to generate the character concept art group of the target virtual character; wherein, the character concept art group includes at least one candidate character concept art corresponding to the virtual character; Candidate character concept art that meets the aforementioned visual attributes is selected from the character concept art group. Based on the candidate character concept art, the concept art of the target virtual character is generated. According to a second aspect of this disclosure, an electronic device is provided, including a memory and a processor, the memory being configured to store a computer program, and the processor being configured to execute the method according to the first aspect of this disclosure under the control of the computer program.
[0011] One beneficial effect of this disclosure is that, through the embodiments of this disclosure, by introducing a preset character construction attribute group to decompose the first character description text, the unstructured natural language character description is transformed into a corresponding set of visual elements, establishing a standardized mapping rule between character text and visual features, effectively eliminating subjective bias and information omissions in human interpretation, and simultaneously integrating the extracted visual elements into the corresponding visual attributes of the reference character portrait to generate the character original artwork, realizing targeted modification and stylized generation based on a controllable visual base, greatly improving the demand fit and generation controllability of the character original artwork, effectively shortening the development cycle, improving construction efficiency, and reducing generation deviation and subsequent modification and iteration costs.
[0012] The features and advantages of the embodiments of this specification will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description
[0013] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments of this specification and, together with their description, serve to explain the principles of these embodiments.
[0014] Figure 1 A schematic diagram is shown that can be used to implement hardware configurations according to embodiments of the present disclosure; Figure 2 A flowchart illustrating a method for constructing virtual characters according to some embodiments is shown; Figure 3 A schematic diagram illustrating the role-based attribute grouping according to some embodiments is shown; Figure 4 A flowchart illustrating a method for constructing virtual characters according to some embodiments is shown; Figure 5 A schematic diagram of a reference character image according to some embodiments is shown; Figure 6 A schematic diagram of a reference character image according to some embodiments is shown; Figure 7 A schematic diagram of a reference character image according to some embodiments is shown; Figure 8 A flowchart illustrating a method for constructing virtual characters according to some embodiments is shown; Figure 9 A flowchart illustrating a method for constructing virtual characters according to some embodiments is shown; Figure 10 A flowchart illustrating a method for constructing virtual characters according to some embodiments is shown; Figure 11 The illustrations show schematic diagrams of character concept art according to some embodiments; Figure 12 A flowchart illustrating a method for constructing virtual characters according to some embodiments is shown; Figure 13 A schematic diagram of the composition structure of a virtual character creation apparatus according to some embodiments is shown; Figure 14 A block diagram of an electronic device according to some embodiments is shown. Detailed Implementation
[0015] Various exemplary embodiments of this specification will now be described in detail with reference to the accompanying drawings.
[0016] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the embodiments of this specification or their application or use.
[0017] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.
[0018] It should be noted that all actions involving the acquisition of signals, information, or data in this embodiment are carried out in compliance with the relevant data protection laws and regulations of the country where the location is situated, and with authorization from the owner of the relevant equipment.
[0019] This disclosure provides a method for constructing virtual characters. For ease of understanding, the application scenarios of the virtual character construction method provided in this disclosure are illustrated below. Figure 1 As shown, Figure 1 This is a schematic diagram of the hardware structure of an electronic device for the virtual character construction method provided in this embodiment of the disclosure.
[0020] Electronic device 1000 is a device capable of running computer programs. These computer programs can be local applications installed on the electronic device, or web applications, lightweight applications, or mini-programs, etc., without limitation. The electronic device 1000 can be a mobile phone, tablet computer, PC, etc., without limitation.
[0021] like Figure 1 As shown, the electronic device 1000 may include a processor 1101, a memory 1102, an interface device 1103, a communication device 1104, an output device 1105, an input device 1106, etc. Figure 1 The hardware configuration shown is illustrative only and is not intended to limit this disclosure, its application, or its use.
[0022] The processor 1101 executes computer programs, which can be written using instruction sets of architectures such as x86, Arm, RISC, MIPS, and SSE. The memory 1102 includes, for example, ROM (Read-Only Memory), RAM (Random Access Memory), and non-volatile memory such as a hard disk. The interface device 1103 includes, for example, a USB interface, a network cable interface, and a headphone jack. The communication device 1104 is capable of wired or wireless communication. The communication device 1104 may include at least one short-range communication module, such as any module for short-range wireless communication based on short-range wireless communication protocols such as Hilink, WiFi (IEEE 802.11), Mesh, Bluetooth, ZigBee, Thread, Z-Wave, NFC, UWB, and LiFi. The communication device 1104 may also include a long-range communication module, such as any module for WLAN, GPRS, or 2G / 3G / 4G / 5G long-range communication. The output device 1105 may include, for example, an LCD screen or touch screen, and a speaker. Input device 1106 may include, for example, a touch screen, a keyboard, a microphone, various sensors, etc.
[0023] In this embodiment, the memory 1102 of the electronic device 1000 is used to store a computer program that controls the processor 1101 to perform a virtual character construction method according to any embodiment of the present disclosure.
[0024] The following is combined Figure 1 The electronic device shown illustrates the virtual character construction method provided in this disclosure embodiment. It should be noted that... Figure 1This is merely one application scenario of the virtual character construction method provided in this disclosure embodiment, and does not mean that the virtual character construction method can only be applied to... Figure 1 The application scenarios shown.
[0025] <First Embodiment> Figure 2 The diagram illustrates a flowchart of a virtual character creation method according to some embodiments. This method is implemented by an electronic device, such as... Figure 1 The electronic device 1000. The virtual character construction method may include the following steps S210-S230: Step S210: Obtain the first character description text and the preset character construction attribute group regarding the construction of the virtual character; wherein, the character construction attribute group includes multiple visual attributes used to define the appearance of the virtual character.
[0026] Virtual characters can be objects that move in a virtual scene, including but not limited to people and other non-human living beings (such as dogs, cats, etc.). The virtual scene mentioned above can be a virtual activity space that simulates the real world provided to users by entertainment products such as games with mobile interfaces.
[0027] The first character description text serves as the source of information for the construction of virtual characters, and is typically in the form of natural language text. This first character description text includes, but is not limited to: the virtual character's background story, identity, basic information, appearance details, personality traits, and abilities and gameplay positioning, such as the game character's class and skill attributes.
[0028] For example, planners can output first character description text based on product positioning. For instance, they can output first character description text for the anthropomorphic rabbit archer character in the game.
[0029] The character construction attribute group is a predefined structured parsing framework for standardizing the decomposition of the first character description text. It breaks down the appearance features of a virtual character into multiple distinct visual attributes, avoiding subjective interpretation biases in natural language text. The visual attributes involved in the character construction attribute group can be flexibly configured according to actual application scenarios to ensure that the character construction attribute group can adapt to the construction needs of different types of virtual characters.
[0030] In this embodiment, the electronic device can obtain the first character description text, clarify the core requirements and design direction of the virtual character, and establish a standardized parsing framework by pre-setting the character construction attribute group, so as to realize the correspondence between the requirement text and the parsing standard, and avoid the omission of visual elements or interpretation deviation in the subsequent visual element extraction process.
[0031] Step S220: Call the visual element extraction model, construct each visual attribute in the attribute group for the character, and extract the visual element corresponding to the visual attribute from the first character description text.
[0032] The visual element extraction model can be a large language model, which can accurately identify the semantic information in the first character description text and extract visual elements in a targeted manner according to the preset character attribute group, avoiding the subjective bias in the traditional manual extraction process. That is, the input of the visual element extraction model is the aforementioned first character description text, and the output of the visual element extraction model is the visual element corresponding to each visual attribute in the character attribute group extracted from the first character description text.
[0033] In this embodiment, the electronic device can input the first character description text into the visual element extraction model. With the help of the natural language understanding and structured parsing capabilities of the visual element extraction model, the unstructured requirements in the first character description text are transformed into visual elements that correspond one-to-one with each visual attribute in the character construction attribute group, providing accurate input for subsequent visual element fusion and character original artwork generation.
[0034] Step S230: The extracted visual elements are integrated into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character.
[0035] Among them, reference character portraits refer to character image materials used to provide visual references for target virtual characters. In essence, they are a kind of directional visual guidance information that can form clear constraints and directional guidance for the generation process of character concept art, ensuring that the generated character concept art of the target virtual character can specifically inherit or integrate its preset visual characteristics, such as style tone, form structure and texture.
[0036] Generally, the specific form of the reference character artwork is not limited, including but not limited to complete character illustrations, character illustrations, character concept art, and 3D rendered screenshots. As long as the image can clearly present the character's visual style, core form, or design tone, and can form a corresponding mapping with the visual attributes of the character's attribute group, it falls within the scope of the reference character portrait protected by this disclosure.
[0037] The specific form of the character concept art is not limited. The character concept art can be a three-view drawing such as a front view, side view, and top view, or it can be a full-body effect drawing of the character, a close-up drawing of the character's partial details, or a dynamic illustration of the character. As long as the image can clearly present the appearance features, body structure, visual style, or core design elements of the target virtual character, and can form a corresponding mapping with the visual attributes of the character's attribute group, it falls within the scope of the character concept art protected by this disclosure.
[0038] In this embodiment, the electronic device can first establish a precise mapping relationship between the extracted visual elements and the corresponding visual attributes in the reference character portrait. Then, through the execution logic of visual attribute matching, visual element replacement / supplementation, and global style unification calibration, the visual attributes of the reference character portrait are processed, and finally, the original character artwork of the target virtual character that meets the requirements of the first character description text and has visual coordination and integrity is generated.
[0039] Through the embodiments of this disclosure, by introducing a preset character construction attribute group to decompose the first character description text, the unstructured natural language character description is transformed into a corresponding set of visual elements, establishing a standardized mapping rule between character text and visual features, effectively eliminating subjective bias and information omissions in human interpretation, and simultaneously integrating the extracted visual elements into the corresponding visual attributes of the reference character portrait to generate character concept art, realizing targeted modification and stylized generation based on a controllable visual base, greatly improving the demand fit and generation controllability of character concept art, effectively shortening the development cycle, improving construction efficiency, and reducing generation deviation and subsequent modification and iteration costs.
[0040] <Second Embodiment> Unlike the first embodiment described above, in this embodiment, by using a character construction attribute group that includes multi-dimensional visual attributes such as historical background, color matching, and decorative style, the standardized decomposition and targeted extraction of the visual features of virtual characters can be achieved, thereby improving the accuracy of character concept art generation.
[0041] The character construction attribute group can include Figure 3 The visual attributes shown include at least some of the historical context, color scheme, decorative style, form and structure, social function, and functional props.
[0042] The aforementioned historical context attributes are used to define the era, worldview, and scene tone of the virtual character, providing a basic positioning for the overall visual style of the virtual character.
[0043] The aforementioned color matching attributes are used to define the main color scheme, color matching relationship, and color tone of the virtual character's appearance, determine the core visual color characteristics of the virtual character, and ensure that the color expression is consistent with the overall style.
[0044] The aforementioned decorative style attributes are used to characterize the design expression of virtual character clothing, patterns, and accessories, reflect the appearance and decorative features and style tendencies of virtual characters, and enhance the visual recognizability of characters.
[0045] The aforementioned physical structure attributes are used to describe the basic physical characteristics of a virtual character, such as body proportions, body contours, and limb shapes. They are the core structural constraints of the virtual character's appearance, ensuring the rationality of movements and visual coordination.
[0046] The aforementioned social functional attributes are used to characterize the identity, occupation, and social positioning of virtual characters within their worldview, and to associate the functional design features of the virtual character's appearance with the character's positioning, so that the visual expression is highly consistent with the character's positioning.
[0047] The aforementioned functional prop attributes are used to define the practical props and equipment equipped and used by virtual characters. These are physical visual elements that match the functional positioning of virtual characters, enhancing the correlation between character functions and visuals.
[0048] <Third Embodiment> Figure 4 A flowchart illustrating a virtual character construction method according to some embodiments is shown. Unlike the first embodiment described above, in this embodiment, by extracting key visual elements characterizing the character's appearance style and automatically matching reference character portraits based on these key visual elements, the efficiency of reference portrait selection and the accuracy of style matching can be improved. For example... Figure 4 As shown, the virtual character construction method of this embodiment may include the following steps S410 to S450: Step S410: Obtain the first character description text and the preset character construction attribute group regarding the virtual character construction.
[0049] Step S420: Call the visual element extraction model, construct the visual attribute in the attribute group for each visual attribute of the character, and extract the visual element corresponding to the visual attribute from the first character description text.
[0050] Step S430: Extract the target visual element from the extracted visual elements.
[0051] Among them, the target visual element is the key visual element used to represent the appearance style of the virtual character.
[0052] In this embodiment, the electronic device can further filter out key visual elements that uniquely represent the appearance style of the virtual character from extracted visual attributes such as historical background, color matching, decorative style, body structure, social function, and functional props.
[0053] Step S440: Based on the extracted target visual elements, search for a character image that matches the target visual elements in the set character image database, and use it as a reference character image.
[0054] In this embodiment, the electronic device can use extracted target visual elements as query conditions to search for character images in a set character image database. During the search process, a visual feature similarity comparison algorithm can be used to match the appearance style features of each character image in the set character image database with the target visual elements one by one, and select one or more character images with a similarity reaching a preset threshold as reference character images. Typically, this preset threshold can be set according to actual needs, such as 80%. For example, taking the anthropomorphic rabbit archer character in the game as an example, the reference character image could be... Figure 5 The character portrait shown is 51. Figure 6 The character portrait shown is 61 and Figure 7 At least one of the character portraits shown in character portrait 71.
[0055] Step S450: The extracted visual elements are integrated into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character.
[0056] <Fourth Embodiment> Figure 8 A flowchart illustrating a virtual character construction method according to some embodiments is shown. Unlike the first embodiment described above, this embodiment can convert a reference character image into standardized reference character description text based on a character construction attribute group, and then fuse the extracted visual elements to generate a second character description text. By generating character concept art through dual constraints of image and text, precise feature alignment can be achieved, style and element deviations can be reduced, and the accuracy and consistency of character concept art generation can be improved. Figure 8 As shown, the virtual character construction method of this embodiment may include the following steps S810 to S850: Step S810: Obtain the first character description text and the preset character construction attribute group regarding the virtual character construction.
[0057] Step S820: Call the visual element extraction model, construct the visual attribute in the attribute group for each visual attribute of the character, and extract the visual element corresponding to the visual attribute from the first character description text.
[0058] Step S830: Guided by the character construction attribute group, the reference character portrait is converted into the corresponding reference character description text.
[0059] The reference character description text may include visual elements corresponding to each visual attribute in the character building attribute group for the reference character.
[0060] In this embodiment, the electronic device can perform image semantic analysis on the reference character portrait guided by a preset character construction attribute group, and map each type of visual feature in the reference character portrait to each visual attribute under the character construction attribute group to form a structured reference character description text that records the visual elements corresponding to each visual attribute.
[0061] Step S840 involves merging the visual elements extracted from the first character description text into the corresponding visual elements of the reference character description text to generate a second character description text about the construction of the virtual character.
[0062] In one example, the reference character profile may include a second character description text. The electronic device may, according to the principle of matching the same visual attribute, merge, replace or modify the visual elements extracted from the first character description text and the visual elements under the same visual attribute in the reference character description text corresponding to the reference character profile, and integrate them to form a second character description text that takes into account both the user's original needs and the visual characteristics of the reference character.
[0063] In one example, there may be multiple reference character profiles. The electronic device can first obtain the visual element fusion weights corresponding to the multiple reference character profiles. These visual element fusion weights are used to characterize the fusion ratio between the visual elements in the corresponding reference character description text and the required elements extracted from the first character description text. Then, the visual elements extracted from the first character description text are combined with the fusion weights of each reference character profile and weighted fusion with the corresponding visual elements with the same visual attributes in the multiple reference character description texts to generate a second character description text. This second character description text can inherit the core visual features of each reference character according to the weight ratio while retaining the user's original character requirements.
[0064] Step S850: Invoke the image generation model to generate the original character artwork of the target virtual character based on the reference character portrait and the second character description text.
[0065] In one example, a reference character portrait may be included. In this case, the electronic device can call an image generation model, using the unique reference character portrait as the visual style and shape base, and the second character description text as the precise requirement constraint. Through dual guidance of image and structured text, the final character concept art of the target virtual character is obtained, which inherits the visual characteristics of the reference character and matches the original character construction requirements.
[0066] In one example, there may be multiple reference character portraits. The image generation model generates the original character artwork of the target virtual character based on the reference character portraits and the second character description text. This can be achieved in the following way: obtain the image fusion weights corresponding to the multiple reference character portraits; call the image generation model to generate the original character artwork of the target virtual character based on the second character description text, the multiple reference character images and their corresponding image fusion weights.
[0067] In this example, the electronic device can first obtain the image fusion weights corresponding to each reference character portrait. These image fusion weights are used to represent the proportion of the visual features of the corresponding reference character portrait in the final character original artwork. Then, the second character description text, multiple reference character portraits and their corresponding image fusion weights are input into the image generation model. The image generation model fuses the visual features of each reference character portrait according to the image fusion weights and follows the constraints of the second character description text to generate a character original artwork that takes into account multiple reference styles and target requirements.
[0068] <Fifth Embodiment> Figure 9 A flowchart illustrating a virtual character construction method according to some embodiments is shown. Unlike the first embodiment described above, in this embodiment, the shape and structure of a reference character image are first adapted and adjusted before image generation. This improves the matching degree of shape features, reduces generation errors, and enhances the accuracy and fit of the character's original artwork construction. Figure 9 As shown, the virtual character construction method of this embodiment may include the following steps S910 to S940: Step S910: Obtain the first character description text and the preset character construction attribute group regarding the virtual character construction.
[0069] Step S920: Call the visual element extraction model, construct the visual attribute in the attribute group for each visual attribute of the character, and extract the visual element corresponding to the visual attribute from the first character description text.
[0070] The character building attribute group can include physical structure.
[0071] Step S930: Call the image adjustment model to adjust the shape and structure of the reference character portrait to match the visual elements of the extracted shape and structure, and obtain the adjusted reference character portrait.
[0072] In this embodiment, the electronic device can call an image adjustment model, for example, to adjust the head-to-body ratio in the body structure of the reference character portrait to match the extracted visual elements of the head-to-body ratio, thereby obtaining the adjusted reference character portrait.
[0073] Step S940: The image generation model is invoked to generate the original character artwork of the target virtual character based on the adjusted reference character portrait and the second character description text.
[0074] <Sixth Embodiment> Figure 10 A flowchart illustrating a virtual character construction method according to some embodiments is shown. Unlike the first embodiment described above, in this embodiment, a group of character concept art containing multiple candidate concept art pieces is generated and subjected to compliance screening to select high-quality candidates that match visual attributes, effectively improving the quality of the final character concept art generation. For example... Figure 10 As shown, the virtual character construction method of this embodiment may include the following steps S1010 to S1050: Step S1010: Obtain the first character description text and the preset character construction attribute group regarding the virtual character construction.
[0075] Step S1020: Call the visual element extraction model, construct the visual attribute in the attribute group for each visual attribute of the character, and extract the visual element corresponding to the visual attribute from the first character description text.
[0076] Step S1030: The extracted visual elements are integrated into the corresponding visual attributes of the reference character image to generate the character concept art group of the target virtual character.
[0077] The character concept art group may include at least one candidate character concept art corresponding to a virtual character.
[0078] Step S1040: Screen out candidate character concept art that meets the visual attributes from the character concept art group.
[0079] In this embodiment, the electronic device can select candidate character concept art that meets the requirements of each extracted visual attribute and corresponding visual element from the generated character concept art group (including multiple candidate concept art), laying the foundation for the subsequent determination of the character concept art of the target virtual character.
[0080] Step S1050: Generate the original character artwork of the target virtual character based on the candidate character artwork.
[0081] In this embodiment, the electronic device can directly use candidate character concept art as the character concept art for the target virtual character. The electronic device can also fine-tune and optimize the candidate character concept art, combining visual elements extracted in the early stages to correct visual elements such as shape structure and color matching, and then determine the fine-tuned candidate character concept art as the character concept art for the target virtual character; or, based on user feedback, select one candidate character concept art from multiple candidate character concept art as the character concept art for the target virtual character. Taking the anthropomorphic rabbit archer character in the game as an example, the generated character concept art could be... Figure 11The character's original artwork shown is 110.
[0082] <Seventh Embodiment> Character concept art and its corresponding display videos are the core display carriers of virtual characters. In the current technology, display videos need to be produced manually by professionals after the character concept art is generated. This process is not only cumbersome and inefficient, but also prone to problems such as inconsistencies in visual style and elements between the video and the character concept art, making it difficult to adapt to the needs of rapid display and agile iteration of virtual characters.
[0083] Figure 12 A flowchart illustrating a virtual character construction method according to some embodiments is shown. Unlike the first embodiment described above, in this embodiment, after generating the character's original artwork, a video generation model is directly called to produce a display video of the target virtual character. This eliminates the need for manual motion design and editing, improving video generation efficiency and reducing manpower and collaboration costs. Simultaneously, it ensures a high degree of consistency in visual style and core elements between the video and the original artwork, enhancing the coherence of the displayed content and adapting to the needs of rapid character display and iteration. Figure 12 As shown, the virtual character construction method of this embodiment may include the following steps S1210 to S1240: Step S1210: Obtain the first character description text and the preset character construction attribute group regarding the virtual character construction.
[0084] Step S1220: Call the visual element extraction model, construct the visual attribute in the attribute group for each visual attribute of the character, and extract the visual element corresponding to the visual attribute from the first character description text.
[0085] Step S1230: The extracted visual elements are integrated into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character.
[0086] Step S1240: Call the video generation model to generate a display video of the target virtual character based on the original character artwork.
[0087] The video generation model can be an artificial intelligence generation model with image-driven video generation capabilities. The input of the video generation model can include at least the original character artwork, and can further include preset motion parameters, camera parameters, duration configuration and action description information, etc. The output can be a dynamic display video of the target virtual character that is consistent with the visual features of the original character artwork.
[0088] Specifically, electronic devices can generate storyboard description information for virtual character display based on the original character artwork and preset action description information, and then render a sequence of storyboard images corresponding to the storyboard description information. The video generation model is then called to render a coherent and stylistically consistent dynamic display video of the target virtual character, using the storyboard image sequence as keyframes or visual constraints.
[0089] <Eighth Embodiment> This embodiment provides a virtual character creation device. Figure 13 A schematic diagram of the composition structure of a virtual character creation apparatus according to an embodiment of the present disclosure is shown. Figure 13 As shown, the virtual character creation device 1300 may include an acquisition module 1310, an extraction module 1320, and a generation module 1330.
[0090] The acquisition module 1310 is used to acquire a first character description text about the construction of a virtual character and a preset character construction attribute group; wherein, the character construction attribute group includes multiple visual attributes for defining the appearance of the virtual character; The extraction module 1320 is used to call the visual element extraction model, construct an attribute group for each visual attribute of the character, and extract the visual element corresponding to the visual attribute from the first character description text. The generation module 1330 is used to integrate the extracted visual elements into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character.
[0091] In some embodiments, the generation module 1330 is further configured to invoke a video generation model to generate a display video of the target virtual character based on the character's original artwork.
[0092] In some embodiments, the generation module 1330 is specifically used to convert the reference character portrait into corresponding reference character description text, guided by the character construction attribute group; wherein, the reference character description text includes visual elements of the reference character corresponding to each visual attribute in the character construction attribute group; each visual element extracted from the first character description text is merged into the corresponding visual element of the reference character description text to generate a second character description text about the construction of the virtual character; and an image generation model is invoked to generate the original character artwork of the target virtual character based on the reference character portrait and the second character description text.
[0093] In some embodiments, the generation module 1330 is specifically used to obtain the image fusion weights corresponding to the multiple reference character images respectively; and call the image generation model to generate the original character artwork of the target virtual character based on the second character description text, the multiple reference character images and their corresponding image fusion weights.
[0094] In some embodiments, the character construction attribute group includes a body structure. The generation module 1330 is specifically used to call an image adjustment model to adjust the body structure of the reference character portrait to match the visual elements of the extracted body structure, thereby obtaining the adjusted reference character portrait; and to call the image generation model to generate the original character artwork of the target virtual character based on the adjusted reference character portrait and the second character description text.
[0095] In some embodiments, the device 1300 further includes a lookup module (not shown in the figure).
[0096] The extraction module 1320 is further configured to extract target visual elements from the extracted visual elements; wherein, the target visual element is a key visual element among the visual elements used to characterize the appearance style of the virtual character; The search module is used to search for a character portrait that matches the extracted target visual element in a set character image database, and use it as the reference character portrait.
[0097] In some embodiments, the character building attribute set includes at least some visual attributes from the following: historical setting, color scheme, decorative style, physical structure, social function, and functional props.
[0098] In some embodiments, the generation module 1330 is specifically used to fuse the extracted visual elements into the corresponding visual attributes of the reference character image to generate the character concept art group of the target virtual character; wherein, the character concept art group includes at least one candidate character concept art corresponding to the virtual character; screening out candidate character concept art that satisfies the visual attributes from the character concept art group; and generating the character concept art of the target virtual character based on the candidate character concept art. <Electronic Device Examples> This embodiment provides an electronic device for implementing a virtual character construction method according to any embodiment of this disclosure. Figure 14 The basic hardware components of this electronic device are shown. For example... Figure 14 As shown, the electronic device 1400 includes a processor 1410 and a memory 1420 for storing a computer program for controlling the processor 1410 to operate in order to control the electronic device 1400 to perform a method according to any embodiment of the present disclosure.
[0099] This disclosure also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements a virtual character construction method according to any embodiment of this disclosure.
[0100] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.
[0101] This disclosure may be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement any of the methods in the foregoing embodiments of this disclosure.
[0102] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media may include, for example, electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), static random access memory (SRAM), compact disc-read-only memory (CD-ROM), digital versatile disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any combination thereof. The computer-readable storage medium used herein is not to be interpreted as a transient signal itself, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.
[0103] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include one or more of copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to computer-readable storage media in the respective computing / processing device.
[0104] The computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source or object programs written in any combination of one or more programming languages, including object-oriented programming languages (such as Smalltalk, C++, etc.) and conventional procedural programming languages (such as the "C" language or similar programming languages). The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network (e.g., a local area network or a wide area network), or it may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays, or programmable logic arrays, can execute computer-readable program instructions to implement various aspects of the embodiments of this disclosure by utilizing state information from the computer-readable program instructions.
[0105] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus, and computer program products according to embodiments of this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.
[0106] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.
[0107] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions that execute on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.
[0108] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions. It should be noted that implementation in hardware, implementation in software, and implementation using a combination of software and hardware are all equivalent.
[0109] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein. The scope of this disclosure is defined by the appended claims.
Claims
1. A method for constructing virtual characters, wherein, include: Obtain the first character description text and the preset character construction attribute group for the virtual character construction; wherein, the character construction attribute group includes multiple visual attributes for defining the appearance of the virtual character; The visual element extraction model is invoked to construct each visual attribute in the attribute group for the character, and the visual element corresponding to the visual attribute is extracted from the first character description text; The extracted visual elements are integrated into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character.
2. The method according to claim 1, wherein, The method further includes: The video generation model is invoked to generate a display video of the target virtual character based on the original character artwork.
3. The method according to claim 1, wherein, The step of fusing the extracted visual elements into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character includes: Guided by the character building attribute group, the reference character portrait is converted into corresponding reference character description text; wherein, the reference character description text includes visual elements of the reference character corresponding to each visual attribute in the character building attribute group; The visual elements extracted from the first character description text are merged into the corresponding visual elements of the reference character description text to generate a second character description text about the construction of the virtual character; The image generation model is invoked to generate the original character artwork of the target virtual character based on the reference character portrait and the second character description text.
4. The method according to claim 3, wherein, The reference character portraits include multiple ones. The image generation model generates the original character artwork of the target virtual character based on the reference character portraits and the second character description text, including: Obtain the image fusion weights corresponding to the multiple reference character portraits; The image generation model is invoked to generate the original character artwork of the target virtual character based on the second character description text, multiple reference character images and their corresponding image fusion weights.
5. The method according to claim 3, wherein, The character construction attribute group includes physical structure. The image generation model generates the original character artwork of the target virtual character based on the reference character portrait and the second character description text; The image adjustment model is invoked to adjust the shape and structure of the reference character portrait to match the visual elements of the extracted shape and structure, thereby obtaining the adjusted reference character portrait; The image generation model is invoked to generate the original character artwork of the target virtual character based on the adjusted reference character portrait and the second character description text.
6. The method according to claim 1, wherein, The method further includes: Extract target visual elements from the extracted visual elements; wherein, the target visual elements are the key visual elements among the visual elements used to characterize the appearance style of the virtual character. Based on the extracted target visual elements, a character portrait that matches the target visual elements is searched in the set character image database and used as the reference character portrait.
7. The method according to any one of claims 1 to 6, wherein, The character construction attribute group includes at least some of the visual attributes from the following: historical background, color scheme, decorative style, physical structure, social function, and functional props.
8. The method according to any one of claims 1 to 6, wherein, The step of fusing the extracted visual elements into the corresponding visual attributes of the reference character portrait to generate the original character artwork of the target virtual character includes: The extracted visual elements are integrated into the corresponding visual attributes of the reference character image to generate the character concept art group of the target virtual character; wherein, the character concept art group includes at least one candidate character concept art corresponding to the virtual character; Candidate character concept art that meets the aforementioned visual attributes is selected from the character concept art group. Based on the candidate character concept art, the concept art of the target virtual character is generated.
9. An electronic device, wherein, It includes a memory and a processor, the memory being used to store a computer program, and the processor being used, under the control of the computer program, to execute the method according to any one of claims 1 to 8.
10. A computer-readable storage medium, wherein, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the method according to any one of claims 1 to 8.