Art asset generation method, device, and medium

Abstract

This disclosure relates to a method, device, and medium for generating art assets. The method includes: obtaining user requirements for generating art assets in a game project, and an art specification document corresponding to the game project; wherein the art specification document defines the visual attributes of art elements in the game project; retrieving target art elements associated with the user requirements from the art specification document, and constructing art asset generation conditions adapted to the user requirements based on the visual attributes of the target art elements; constructing a task set for generating art assets based on the user requirements; wherein the task set includes at least one generation task; and executing each generation task in the task set, guided by the art asset generation conditions, to generate target art assets for the game project; wherein the target art assets are art assets that meet the user requirements and the art asset generation conditions.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more specifically, to a method, device, and medium for generating art assets. Background Technology

[0002] The rapid development of AI technology has driven the game industry to introduce AI tools to assist in the generation of art assets, in order to meet diverse production needs such as characters, scenes, and promotional videos. Currently, most mainstream tools are web-based or application-based, allowing users to generate art assets by inputting text prompts or uploading reference images. While the generated art assets are visually appealing, they often fail to meet the standards for game project implementation, requiring extensive manual revisions and thus failing to meet the core need for large-scale, standardized production of game art assets. Summary of the Invention

[0003] One objective of this disclosure is to provide a new technical solution for generating art assets.

[0004] According to a first aspect of this disclosure, a method for generating art assets is provided, comprising: Obtain user requirements for the generation of art assets for the game project, as well as the corresponding art specification document for the game project; wherein, the art specification document is used to define the visual attributes of art elements in the game project; Based on the user's requirements, target art elements associated with the user's requirements are retrieved from the art specification document, and art asset generation conditions adapted to the user's requirements are constructed based on the visual attributes of the target art elements. A set of tasks for generating art assets is constructed based on the user's requirements; wherein the set of tasks includes at least one generation task; Guided by the aforementioned art asset generation conditions, each generation task in the task set is executed to generate the target art asset for the game project; wherein, the target art asset is an art asset that meets the user's needs and the aforementioned art asset generation conditions.

[0005] Optionally, the method further includes: Determine the execution workflow corresponding to the task set; wherein the execution workflow indicates the task dependencies between each generation task in the task set, and the input data of each generation task; wherein the input data of the generation task includes at least the task description information of the corresponding generation task, and at least one of the art asset generation conditions and intermediate execution results, wherein the intermediate execution results are the execution results output by the upstream generation task that meet the art asset generation conditions; The execution of each generation task in the art asset generation task set includes: Following the instructions of the execution workflow, each generation task in the art asset generation task set is executed sequentially.

[0006] Optionally, the task set includes a first generation task for drawing art assets, and a second generation task for verifying whether the first execution result of the first generation task meets the conditions for generating the art assets. The process of executing each generation task in the task set, guided by the aforementioned art asset generation conditions, includes: The first generation task is executed using the aforementioned art asset generation conditions as input, and a first execution result is obtained; Using the art asset generation conditions and the first execution result as input, the second generation task is executed to obtain the second execution result; wherein, if the first execution result fails the verification, the second execution result also includes a correction item based on the art asset generation conditions; If the second execution result indicates that the first execution result failed the verification, the first generation task is re-executed with the art asset generation conditions, the correction item, and the first execution result as input to obtain the corrected first execution result. Optionally, the step of executing each generation task in the task set, guided by the art asset generation conditions, includes: For each generation task, a matching functional module is searched in the functional module registry, and the matching functional module is called to execute the corresponding generation task guided by the art asset generation conditions.

[0007] Optionally, the functional module registry includes the deployment locations of the registered functional modules, wherein different deployment locations correspond to different communication protocols. The invocation of the matching functional module, guided by the art asset generation conditions, to execute the corresponding generation task includes: By using the communication protocol corresponding to the matching functional module, the matching functional module is invoked to execute the corresponding generation task guided by the art asset generation conditions.

[0008] Optionally, the task set for generating art assets based on the user's needs includes: The user requirements are semantically expanded to generate multiple storyboard scripts with temporal relationships; The task set is constructed based on the multiple storyboard scripts.

[0009] Optionally, the task set includes: a first generation task and a second generation task, wherein the first generation task is used to draw character images in the storyboard, and the second generation task is used to verify whether the first execution result of the first generation task meets the art asset generation conditions; if the first execution result fails the verification, the second execution result of the second generation task also includes a correction item compared with the art asset generation conditions, and the correction item triggers the re-execution of the first generation task until a target character image that meets the art asset generation conditions is obtained; The task set also includes a number of third generation tasks that correspond one-to-one with the multiple storyboard scripts. The third generation tasks are used to generate corresponding storyboard videos based on the target character image and the corresponding storyboard script. The task set also includes a fourth generation task, which is used to generate the target art asset based on multiple storyboard videos.

[0010] Optionally, the third generation task includes a first subtask and a second subtask; the first subtask is used to generate a storyboard image based on the target character image and the corresponding storyboard script; the second subtask is used to generate a storyboard video based on the storyboard image and dynamic parameters.

[0011] 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 described according to the first aspect of this disclosure under the control of the computer program.

[0012] According to a third aspect of this disclosure, a computer-readable storage medium is provided, on which a computer program is stored, the computer program implementing the method according to a first aspect of this disclosure when executed by a processor.

[0013] One beneficial effect of this disclosure is that, based on user needs and game project-specific art specification documents, this disclosure can accurately match the complex art specifications of game projects by automatically searching for associated specification constraints, constructing compliant generation conditions, and breaking down generation tasks. This effectively avoids the pain point of existing AI tools generating non-compliant art assets that require a large amount of manual secondary correction, significantly improving the compliance and production efficiency of game art assets, reducing the operational threshold for non-technical personnel, and achieving a synergistic balance between the needs of large-scale production and the uniformity of specifications.

[0014] 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

[0015] 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.

[0016] Figure 1 A schematic diagram of the hardware configuration of an art asset generation system that can be used to implement the art asset generation method according to embodiments of the present disclosure is shown; Figure 2 A schematic diagram of the hardware configuration of an electronic device that can be used to implement the art asset generation method according to embodiments of the present disclosure is shown; Figure 3 A flowchart illustrating a method for generating art assets according to some embodiments is shown; Figure 4 A flowchart illustrating a method for generating art assets according to other embodiments is shown; Figure 5 A flowchart illustrating a method for generating art assets according to other embodiments is shown; Figure 6 A flowchart illustrating a method for generating art assets according to other embodiments is shown; Figure 7 A flowchart illustrating a method for generating art assets according to other embodiments is shown; Figure 8 A schematic diagram of the composition structure of an art asset generation apparatus according to some embodiments is shown; Figure 9 A schematic diagram of the composition structure of an electronic device according to some embodiments is shown. Detailed Implementation

[0017] Various exemplary embodiments of this specification will now be described in detail with reference to the accompanying drawings.

[0018] 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.

[0019] 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.

[0020] 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.

[0021] This disclosure provides a method for generating art assets. For ease of understanding, the application scenarios of the art asset generation method provided in this disclosure are illustrated below. Figure 1 As shown, Figure 1 A schematic diagram of the hardware configuration of an art asset generation system that can be used to implement the art asset generation method according to embodiments of the present disclosure is shown.

[0022] like Figure 1 As shown, the art asset generation system 1 may include a user terminal 11, an intelligent agent server 12, a message control platform (MCP) service bus 13, and an underlying functional service cluster 14.

[0023] The user terminal 11 serves as the entry point for user interaction with the system, providing a web-based or application-based interface. It allows users to submit requests via natural language, sketches, or other means, and to receive the final generated target art assets. User requests can be transmitted to the intelligent agent server 12 via WebSocket or Hypertext Transfer Protocol (HTTP) to ensure the real-time performance and stability of the interaction.

[0024] Among them, the intelligent agent server 12 is the core control and scheduling hub of the system. It is mainly responsible for parsing user requirements, generating task sets, scheduling underlying services, and handling logical decisions and feedback corrections during task execution. It is the brain of the entire generation process. The intelligent agent server 12 may include a context manager 121, an orchestration engine 122, and a task queue executor 123.

[0025] The context manager 121 is responsible for maintaining the state information of the entire generation process. Through a state database, information persistence and traceability are achieved, providing contextual support for task scheduling and logical decisions. Specifically, the context manager 121 is responsible for maintaining a global state object (AgentState), which is a persistent dictionary structure containing: messages (dialogue history), requirements (original user requirements), current_asset_ptr (pointer to the currently processed asset), style_guide (style constraints extracted from the document), and task_queue (list of tasks to be executed). Implementation details: A SQLite database is used for state checkpointing to ensure that in the event of a system interruption during long-running generation processes, recovery from the breakpoint is possible.

[0026] The orchestration engine 122, built upon finite state graph theory, is responsible for task scheduling and logical processing. It features a visual feedback loop. The reviewer node utilizes a multimodal model with visual perception capabilities to analyze generated images / videos, comparing them against the original user requirements and the game project's specific art guidelines. If the check fails, the system automatically triggers redrawing or editing commands, forming an automated closed loop of generation-review-correction without manual intervention. For example, the reviewer node can construct a prompt: "As the art director, please check if this image matches the description: [Prompt]. Specifically, check for any limb errors and whether the style is [Style]. If it matches, reply PASS; otherwise, reply REJECT with specific modification suggestions." If a REJECT is received, the system automatically converts the "modification suggestions" into a new prompt and reschedules the Artist node for repair.

[0027] Task queue executor 123 is responsible for parsing task sets and generating execution workflows. It achieves orderly task scheduling through task queues, ensuring the automation and standardization of the generation process and avoiding task execution chaos and connection deviations. Each task in the queue contains specific tool call instructions, parameters, and context placeholders (such as {{last_output}}). Task queue executor 123 parses and executes the queue sequentially, enabling unattended production of long workflows.

[0028] The MCP service bus 13, serving as a hybrid transport layer, acts as the communication hub connecting the intelligent agent server 12 and the underlying functional service cluster 14. It supports both local (low latency) and remote (long polling) communication protocols, automatically adapting to the deployment location and performance requirements of different services to ensure efficient and stable cross-module data transmission. The local channel can be Stdio transmission, referring to communication with child processes via local standard input / output streams, used for low-latency, high-frequency calls to local tools. The remote channel can be SSE (Server-Sent Events) transmission, referring to HTTP long-connection communication via Server-Sent Events, used to call remote Application Programming Interface (API) services.

[0029] The underlying functional service cluster 14 can be a provider of specific generation capabilities, such as, but not limited to, drawing services, video services, visual services, editing services, and a retrieval-augmented generation (RAG) knowledge base. Among these, project specification enhancement (constraint-based RAG) refers to using a vector database to store unstructured art specification documents. When generating a Prompt, the system automatically retrieves relevant constraints and injects context, forcing the generated model to conform to the specific project's private specifications.

[0030] Figure 2 A schematic diagram of the hardware configuration of an electronic device 1000 that can be used to implement the art asset generation method according to embodiments of the present disclosure is shown. The electronic device 1000 may be configured with an art asset generation system, which may be... Figure 1 The art asset generation system shown is 1.

[0031] like Figure 2 As shown, electronic device 1000 is a device capable of running computer programs. These programs can be local applications installed on the electronic device, or web applications, lightweight applications, or mini-programs, etc., without limitation. Electronic device 1000 can be a mobile phone, tablet computer, PC, etc., without limitation.

[0032] 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.

[0033] 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.

[0034] In this embodiment, the memory 1102 of the electronic device 1000 is used to store a computer program that controls the processor 1101 to operate in order to execute the art asset generation method according to any embodiment of this disclosure.

[0035] The following is combined with Figure 2 The electronic device shown illustrates the art asset generation method provided in this disclosure embodiment. It should be noted that... Figure 2 This is merely one application scenario of the art asset generation method provided in this disclosure embodiment, and does not mean that the art asset generation method can only be applied to... Figure 2 The application scenarios shown.

[0036] <First Embodiment> Figure 3 The diagram illustrates a flowchart of a method for generating art assets according to some embodiments. This method is implemented by an electronic device, such as... Figure 2 Electronic device 1000. The method for generating art assets may include the following steps S310-S340: Step S310: Obtain user requirements for the generation of art assets for the game project, as well as the corresponding art specification document for the game project; wherein, the art specification document is used to define the visual attributes of art elements in the game project.

[0037] Among them, the art assets of a game project include, but are not limited to, character art assets, scene art assets, user interface art assets, special effects art assets, promotional art assets, and narrative art assets.

[0038] The aforementioned character art assets may include character images, which can be understood as character concept art, and the character images may be three views of the character, such as a front view, a side view, and a back view.

[0039] The aforementioned scene art assets can include basic scene resources and scene detail resources. Basic scene resources can include terrain, buildings, and natural environment elements under the game's world view; scene detail resources can include scene props, interactive trigger elements, and scene atmosphere textures.

[0040] The aforementioned user interface art assets may include functional interface resources and interactive feedback resources. The aforementioned special effects art assets may include skill and combat effects, as well as environmental and interactive effects.

[0041] The aforementioned promotional art assets may include game promotional posters and game promotional videos. The aforementioned narrative art assets may include game cutscenes and game story short films.

[0042] In this embodiment, on the one hand, the electronic device can receive user input regarding the generation of art assets for the game project through a preset human-computer interaction interface, such as the web interface or application interface of the user terminal. The user input may include, but is not limited to, natural language text descriptions and / or sketch uploads.

[0043] Each game project has its own dedicated art specification document. This document can be a guiding document developed during the game project's development process to unify the visual attributes of all art assets within the project, ensuring compliance and consistency. These visual attributes can include all characteristic parameters of the art elements in their visual presentation, including but not limited to style type, color scheme standards, texture representation, proportions, detail requirements, and logical consistency requirements.

[0044] In this embodiment, the electronic device can receive the game project's exclusive art specification document uploaded by the user through a preset document upload interface, or it can directly retrieve the pre-archived and stored art specification document from the game project's exclusive private knowledge base that is pre-associated with this system. The document format includes, but is not limited to, PDF and Word, to ensure compatibility with common specification document storage formats in game projects.

[0045] Typically, the acquired art specification documents can be converted into vector data through an embedding model and stored in a system-related vector database, such as the RAG knowledge base, forming a dedicated art specification vector index library for the game project.

[0046] Specifically, users can submit requests via their terminals, such as "Help me make a short post-apocalyptic video featuring a mecha girl as the main character, set in a sandstorm-ravaged wasteland." These requests can be transmitted to the context manager on the agent server via WebSocket / HTTP protocol. Simultaneously, the system retrieves the game project's specific art specification document from the RAG knowledge base. This document has been converted into vector data storage using an embedding model. The context manager initializes the requirement state database and checkpoints, recording the core features of the requirement and the specification version information.

[0047] Step S320: Based on the user's requirements, retrieve the target art elements associated with the user's requirements from the art specification document, and construct art asset generation conditions that adapt to the user's requirements based on the visual attributes of the target art elements.

[0048] In this embodiment, the electronic device can vectorize user requirements to generate corresponding requirement vectors; and using these requirement vectors as search criteria, it can accurately retrieve the specification fragments corresponding to the target art elements associated with the user requirements from a vector database that stores a vector index of pre-stored art specification documents. These specification fragments corresponding to the target art elements associated with the user requirements are typically the core constraints related to the visual attributes of the target art elements within the art specification documents.

[0049] In this embodiment, the electronic device can automatically construct standardized art asset generation conditions that can be recognized and executed by the generation model based on the visual attributes of the target art elements retrieved above and in combination with user needs. The art asset generation conditions include, but are not limited to, positively guided Prompt constraint statements, visual quantification parameters, and a list of prohibited rules for negative constraints, ensuring that the generation model strictly follows the game project's exclusive art specifications while accurately meeting the user's actual creative needs when performing subsequent generation tasks.

[0050] Specifically, the orchestration engine's Manager node vectorizes user requirements, generating requirement vectors. Through a remote channel on the MCP service bus, it retrieves target art element specification fragments matching the requirement vectors from the RAG knowledge base, such as requiring mechs to have a rusty yellow metallic texture. Combining user requirements with the retrieved visual attributes, it constructs conditions for generating art assets, including positive prompt constraints, visual quantification parameters such as style vectors and color ranges, and a reverse list of disallowed rules.

[0051] Through this embodiment, electronic devices can accurately match user needs with the art specifications of game projects. By extracting associated visual constraints, they can construct art asset generation conditions that are both demand-adaptable and compliant, providing clear rule guidance for the execution of subsequent generation tasks.

[0052] Step S330: Construct a task set for generating art assets based on the user's requirements; wherein the task set includes at least one generation task.

[0053] In this embodiment, the electronic device can, based on user needs and combined with the established conditions for generating art assets, structurally decompose and logically plan the art asset generation process to form a corresponding generation task set. The generation task set can contain at least one generation task that can be independently scheduled and executed by the system. Each generation task can correspond to a single type of art asset production stage, including but not limited to character image generation, scene storyboard drawing, and video clip compositing. Furthermore, the generation task set can be further configured and its specific execution content, inter-task dependencies, and execution order can be clearly defined, providing a standardized execution basis for the subsequent orderly and automated execution of the art asset generation process.

[0054] Specifically, the orchestration engine can break down independently executable generation tasks based on user needs and art asset generation conditions, such as character drawing, storyboard generation, and video compositing. The task queue executor then generates an execution workflow containing task dependencies. During the construction of the task set, the context manager can update the task status in real time.

[0055] Step S340: Guided by the art asset generation conditions, execute each generation task in the task set to generate the target art asset of the game project; wherein, the target art asset is an art asset that meets the user requirements and the art asset generation conditions.

[0056] In this embodiment, the electronic device can be guided by the conditions for generating art assets and combined with user needs to call the generation model to execute each generation task in the task set, so as to generate art assets of the game project that meet the user needs and the conditions for generating art assets.

[0057] After all generation tasks are completed, the electronic device integrates the results of each task to ultimately generate the target art assets for the game project. In other words, the target art assets undergo multi-dimensional verification to accurately match user needs and fully comply with the game project's art specifications corresponding to the art asset generation conditions, thus achieving compliance and precision in the art assets.

[0058] Specifically, following the instructions of the execution workflow, the system matches the underlying services corresponding to each task through the functional module registry and completes the invocation using an adapted communication protocol. For example, for tasks with low latency requirements, such as character drawing, the drawing service can be invoked through the local channel of the MCP service bus. Each generation task adheres to the art asset generation conditions throughout the entire process and finally returns to the user terminal via the WebSocket / HTTP protocol, completing the delivery of the target art asset.

[0059] Through the embodiments disclosed herein, based on user needs and game project-specific art specification documents, by automatically retrieving associated specification constraints, constructing compliant generation conditions, and breaking down generation tasks, it is possible to accurately match the complex art specifications of game projects. This effectively avoids the pain point of existing AI tools generating non-compliant art assets that require a large amount of manual secondary correction, significantly improving the compliance and production efficiency of game art assets, reducing the operational threshold for non-technical personnel, and achieving a synergistic balance between the needs of large-scale production and the uniformity of specifications.

[0060] <Second Embodiment> Figure 4 The diagram illustrates a flowchart of an art asset generation method according to some embodiments. Unlike the first embodiment described above, this embodiment utilizes a workflow that matches task sets to clarify the dependencies and standardized inputs of each generation task, enabling orderly scheduling and automated execution of the generation process. Simultaneously, by relying on the reuse of upstream compliant intermediate results, the standardization of inputs at each stage is ensured, avoiding task execution chaos and connection deviations, improving overall generation efficiency and process stability, and further ensuring that the final art asset generation meets regulatory requirements. Figure 4 As shown, the art asset generation method of this embodiment may include the following steps S410 to S440: Step S410: Obtain user requirements for the generation of art assets for the game project, as well as the corresponding art specification documents for the game project.

[0061] Step S420: Based on user needs, retrieve target art elements related to user needs from the art specification document, and construct art asset generation conditions that adapt to user needs based on the visual attributes of the target art elements.

[0062] Step S430: Construct a task set for generating art assets based on user needs, and determine the execution workflow for the corresponding task set.

[0063] The task set can include multiple generated tasks, and the execution workflow can indicate the task dependencies between the generated tasks in the task set, as well as the input data for each generated task.

[0064] The input data for the generation task may include at least the task description information of the corresponding generation task, as well as at least one of the art asset generation conditions and intermediate execution results. The intermediate execution results may be the execution results output by the upstream generation task that meet the art asset generation conditions.

[0065] Step S440: Guided by the conditions for generating art assets, and following the instructions of the execution workflow, execute each generation task in the task set in sequence to generate the target art assets for the game project.

[0066] Within this system, subtasks can automatically flow between each other using placeholder references. Specifically, placeholders refer to a dynamic variable referencing mechanism (such as {{current_image}}) used in the task queue. This allows subsequent tasks to automatically obtain the output of preceding tasks as input parameters, thus enabling automated data flow across tools and modalities. For example, when task 1 is completed, the task queue executor automatically replaces the placeholder, allowing the system to plan and generate 10 different video scenes at once, and execute them automatically, either sequentially or in parallel, without manual intervention at each step.

[0067] Specifically, the task queue executor is responsible for parsing the task set and generating the execution workflow, clarifying the dependencies and input data of each task. The system schedules each functional module sequentially through the MCP service bus according to the workflow's timing requirements, ensuring an orderly and consistent generation process.

[0068] In this embodiment, the electronic device can execute each generation task in the art asset generation task set sequentially according to the instructions of the execution workflow and the task dependencies and preset execution sequence between each generation task. Each generation task is based on the input data indicated by the execution workflow, ensuring that the generation process is orderly and the constraints are unified, thereby completing the automated generation of the target art asset.

[0069] <Third Embodiment> Figure 5 A flowchart illustrating a method for generating art assets according to some embodiments is shown. Unlike the first embodiment described above, this embodiment separates generation and verification into independent tasks, forming an automated closed-loop verification mechanism. This mechanism can automatically identify non-compliant results and generate targeted corrections. Based on these corrections, the generation task is re-executed, achieving closed-loop compliance assurance. This significantly reduces the workload of manual review and modification, improving the stability of the generation process and the compliance quality of the final assets. Figure 5 As shown, the conditions for generating art assets in this embodiment may include the following steps S510 to S560: Step S510: Obtain user requirements for the generation of art assets for the game project, as well as the corresponding art specification documents for the game project.

[0070] Step S520: Based on user needs, retrieve target art elements related to user needs from the art specification document, and construct art asset generation conditions that adapt to user needs based on the visual attributes of the target art elements.

[0071] Step S530: Construct a first generation task for drawing art assets based on user needs, and a second generation task for verifying whether the first execution result of the first generation task meets the conditions for generating art assets.

[0072] Step S540: Execute the first generation task with the art asset generation conditions as input, and obtain the first execution result.

[0073] Specifically, the Artist node of the orchestration engine can call the drawing service through the local channel of the MCP service bus to generate the initial character image, which is the first execution result.

[0074] Step S550: Using the art asset generation conditions and the first execution result as input, execute the second generation task to obtain the second execution result.

[0075] In cases where the first execution result fails validation, the second execution result may also include correction items based on the art asset generation conditions. These correction items typically address specific modification requirements identified in the first execution result regarding non-compliant visual elements, parameter deviations, and detail defects, based on the art asset generation conditions. They provide a clear direction for corrections in subsequent iterations of the first generation task.

[0076] Specifically, the orchestration engine's review node can call the visual service to validate the art asset generation conditions, user requirements, and initial character images as input. If the validation fails, a second execution result containing corrections such as "adding a severe rust effect to the mech" is generated and fed back to the management scheduling node through the orchestration engine.

[0077] Step S560: If the second execution result indicates that the first execution result failed the verification, the first generation task is re-executed with the art asset generation conditions, correction items and the first execution result as input to obtain the corrected first execution result.

[0078] In this embodiment, when the electronic device performs the correction task, it usually adjusts the generation logic according to the correction item to obtain the first execution result after correction that meets the conditions for generating art assets, thereby realizing the automated closed-loop correction of non-compliant output.

[0079] Specifically, the management and scheduling node can update the conditions for generating art assets, triggering the artist node to call the drawing service again until a target character image that meets the specifications is generated.

[0080] <Fourth Embodiment> Figure 6 A flowchart illustrating a method for generating art assets according to some embodiments is shown. Unlike the first, second, and third embodiments described above, this embodiment utilizes a functional module registry and protocol adaptation calling mechanism to achieve unified scheduling of multi-protocol, multi-deployment-location functional modules, enhancing system scalability and compatibility, and ensuring efficient and stable task execution. For example... Figure 6 As shown, the art asset generation method of this embodiment may include the following steps S610 to S640: Step S610: Obtain user requirements for the generation of art assets for the game project, as well as the corresponding art specification documents for the game project.

[0081] Step S620: Based on user needs, retrieve target art elements related to user needs from the art specification document, and construct art asset generation conditions that adapt to user needs based on the visual attributes of the target art elements.

[0082] Step S630: Construct a task set for generating art assets based on user needs; wherein the task set includes at least one generation task.

[0083] Step S640: For each generation task, search for the matching functional module in the functional module registry, and call the matching functional module to execute the corresponding generation task guided by the art asset generation conditions.

[0084] The function module registry can include the deployment location of the registered function modules, where different deployment locations correspond to different communication protocols.

[0085] In this embodiment, the invocation of the matching functional module guided by the art asset generation conditions to execute the corresponding generation task can be implemented in the following way: through the communication protocol corresponding to the matching functional module, the matching functional module is invoked guided by the art asset generation conditions to execute the corresponding generation task.

[0086] Specifically, to achieve modular and standardized execution of generation tasks, when guiding the execution of various tasks based on art asset generation conditions, it is necessary to first search and match the corresponding functional implementation modules in a pre-defined functional module registry according to the functional requirements of each generation task. The functional module registry pre-registers the deployment locations of each functional module and configures corresponding communication protocols for different deployment locations. After determining the matching module, the electronic device will use a communication protocol adapted to the module's deployment location to complete the module's communication connection and invocation. Constrained by the art asset generation conditions, it will drive the matching module to execute the corresponding generation task, ensuring compatibility and execution reliability of module invocation under different deployment modes.

[0087] For example, for each generation task, the system retrieves a matching service from the functional module registry (such as a drawing service for character drawing and a video service for video compositing), and automatically selects a communication protocol based on the service's deployment location, such as calling a locally deployed service through a local channel and a remote service through a remote channel.

[0088] <Fifth Embodiment> Figure 7 The diagram illustrates a flowchart of an art asset generation method according to some embodiments. Unlike the first, second, and third embodiments described above, this embodiment automatically generates a chronological storyboard script by semantically expanding user requirements and constructs a complete task set including a closed-loop character drawing verification process, step-by-step generation of storyboard videos, and final asset integration. This achieves fully automated generation of art assets from requirements to target assets, ensuring both the compliance of character materials and standardization of the video art asset production process, significantly reducing manual intervention and production costs. Figure 7 As shown, the art asset generation method of this embodiment may include the following steps S710 to S740: Step S710: Obtain user requirements for the generation of art assets for the game project, as well as the corresponding art specification documents for the game project.

[0089] Step S720: Based on user needs, retrieve target art elements related to user needs from the art specification document, and construct art asset generation conditions that adapt to user needs based on the visual attributes of the target art elements.

[0090] Step S730: Semantically expand the user requirements, generate multiple storyboard scripts with temporal relationships, and construct a task set based on the multiple storyboard scripts.

[0091] The task set may include: a first generation task and a second generation task. The first generation task can be used to draw character images in the storyboard script, and the second generation task can be used to verify whether the first execution result of the first generation task meets the conditions for generating art assets.

[0092] In this embodiment, if the first execution result fails the verification, the second execution result of the second generation task may also include a correction item based on the art asset generation conditions. The correction item can trigger the re-execution of the first generation task until a target character image that meets the art asset generation conditions is obtained. After passing the review node, the feature vector or random seed value of the character image can be locked and stored in the context manager. When generating storyboards in subsequent stages, the newly generated frames will be required to contain the visual features of the character image.

[0093] The task set may also include multiple third generation tasks corresponding one-to-one with multiple storyboard tasks. The third generation tasks can be used to generate corresponding storyboard videos based on the target character image and the corresponding storyboard script. The third generation tasks may include a first subtask and a second subtask. The first subtask can be used to generate storyboard images based on the target character image and the corresponding storyboard script; the second subtask can be used to generate storyboard videos based on the storyboard images and dynamic parameters.

[0094] The task set may also include a fourth generation task, which can be used to generate target art assets based on multiple storyboard videos.

[0095] Step S740: Guided by the conditions for generating art assets, execute each generation task in the task set to generate the target art assets for the game project; wherein, the target art assets are art assets that meet the user's needs and the conditions for generating art assets.

[0096] Specifically, the user's requirements are as follows: The user wants to create a promotional video for a "post-apocalyptic sci-fi" game, requiring the protagonist to be a "girl wearing tattered mecha," with the background being an "abandoned city in a sandstorm." The management and scheduling nodes of the orchestration engine can call a large language model to semantically expand the user's requirements, generating three storyboards with a temporal relationship. For example, the first storyboard might be: the girl walking through ruins; the second storyboard might be: the girl discovering relics from an old era; and the third storyboard might be: encountering a mechanical enemy. The task set built based on the storyboards can include: First generation task: the artist node calls the drawing service to generate character images. Second generation task: the review node calls the visual service to verify the character's compliance and generate corrections. Third generation task: the video processing node (Videographer) calls the drawing service to generate storyboard images, and then calls the video service to generate video clips (including image generation and video compositing subtasks). Fourth generation task: the editing service is called to stitch the storyboard videos together to generate the final promotional video. All tasks are executed collaboratively through the MCP service bus, achieving fully automated generation from requirements to video assets.

[0097] <Sixth Embodiment> This embodiment provides an art asset generation device. Figure 8 A schematic diagram of the composition structure of an art asset generation apparatus according to an embodiment of the present disclosure is shown. Figure 8 As shown, the art asset generation device 800 may include an acquisition module 810, a retrieval module 820, a first generation module 830, and a fourth generation module 840.

[0098] The acquisition module 810 is used to acquire user requirements for the generation of art assets for a game project, as well as the corresponding art specification document for the game project; wherein, the art specification document is used to define the visual attributes of art elements in the game project; The retrieval module 820 is used to retrieve target art elements associated with the user's needs in the art specification document according to the user's needs, and to construct art asset generation conditions that are adapted to the user's needs based on the visual attributes of the target art elements. The first generation module 830 is used to construct a task set for generating art assets based on the user's requirements; wherein the task set includes at least one generation task. The fourth generation module 840 is used to execute each generation task in the task set guided by the art asset generation conditions, so as to generate the target art asset of the game project; wherein the target art asset is an art asset that meets the user requirements and the art asset generation conditions.

[0099] In some embodiments, the device 800 further includes a determining module (not shown).

[0100] A determination module is used to determine the execution workflow corresponding to the task set; wherein the execution workflow indicates the task dependencies between each generation task in the task set, and the input data of each generation task; wherein the input data of the generation task includes at least the task description information of the corresponding generation task, and at least one of the art asset generation conditions and intermediate execution results, wherein the intermediate execution results are the execution results output by the upstream generation task that meet the art asset generation conditions; The second generation module 840 is specifically used to sequentially execute each generation task in the art asset generation task set according to the instructions of the execution workflow.

[0101] In some embodiments, the task set includes a first generation task for drawing art assets, and a second generation task for verifying whether a first execution result of the first generation task meets the conditions for generating the art assets. The second generation module 840 is specifically used to execute the first generation task with the art asset generation conditions as input to obtain a first execution result; and to execute the second generation task with the art asset generation conditions and the first execution result as input to obtain a second execution result; wherein, if the first execution result fails the verification, the second execution result also includes a correction item based on the art asset generation conditions; and if the second execution result indicates that the first execution result fails the verification, the first generation task is re-executed with the art asset generation conditions, the correction item, and the first execution result as input to obtain a corrected first execution result. In some embodiments, the second generation module 840 is specifically used to search for a matching functional module in the functional module registry for each generation task, and to call the matching functional module to execute the corresponding generation task guided by the art asset generation conditions.

[0102] In some embodiments, the functional module registry includes the deployment locations of the registered functional modules, wherein different deployment locations correspond to different communication protocols. The second generation module 840 is specifically used to call the matching functional module through the communication protocol corresponding to the matching functional module, guided by the art asset generation conditions, to execute the corresponding generation task.

[0103] In some embodiments, the first generation module 830 is specifically used to semantically expand the user requirements and generate multiple storyboard scripts with temporal relationships; and to construct the task set based on the multiple storyboard scripts. In some embodiments, the task set includes: a first generation task for drawing character images in the storyboard, and a second generation task for verifying whether the character images meet the art asset generation conditions. The first generation task is used to draw character images in the storyboard, and the second generation task is used to verify whether the character images meet the art asset generation conditions. If the character image fails the verification, the second execution result of the second generation task further includes a correction item based on the art asset generation conditions. The correction item triggers the re-execution of the first generation task until a target character image that meets the art asset generation conditions is obtained. The task set also includes a number of third generation tasks that correspond one-to-one with the multiple storyboard task scripts. The third generation tasks are used to generate corresponding storyboard videos based on the target character image and the corresponding storyboard script. The task set also includes a fourth generation task, which is used to generate the target art asset based on multiple storyboard videos.

[0104] In some embodiments, the third generation task includes a first subtask and a second subtask; the first subtask is used to generate a storyboard image based on the target character image and the corresponding storyboard script; the second subtask is used to generate a storyboard video based on the storyboard image and dynamic parameters.

[0105] <Electronic Device Examples> This embodiment provides an electronic device for implementing an art asset generation method according to any embodiment of this disclosure. Figure 9 A schematic diagram of the electronic device's structure is shown. (For example...) Figure 9 As shown, the electronic device 900 includes a processor 910 and a memory 920. The memory 920 stores a computer program that controls the processor 910 to operate in order to control the electronic device 900 to execute an art asset generation method according to any embodiment of the present disclosure.

[0106] This disclosure also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements an art asset generation method according to any embodiment of this disclosure.

[0107] 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.

[0108] 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.

[0109] 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.

[0110] 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.

[0111] 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.

[0112] 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.

[0113] 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.

[0114] 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.

[0115] 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.

[0116] 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 generating art assets, wherein, include: Obtain user requirements for the generation of art assets for the game project, as well as the corresponding art specification document for the game project; wherein, the art specification document is used to define the visual attributes of art elements in the game project; Based on the user's requirements, target art elements associated with the user's requirements are retrieved from the art specification document, and art asset generation conditions adapted to the user's requirements are constructed based on the visual attributes of the target art elements. A set of tasks for generating art assets is constructed based on the user's requirements; wherein the set of tasks includes at least one generation task; Guided by the aforementioned art asset generation conditions, each generation task in the task set is executed to generate the target art asset for the game project; wherein, the target art asset is an art asset that meets the user's needs and the aforementioned art asset generation conditions.

2. The method according to claim 1, wherein, The method further includes: Determine the execution workflow corresponding to the task set; wherein the execution workflow indicates the task dependencies between each generation task in the task set, and the input data of each generation task; wherein the input data of the generation task includes at least the task description information of the corresponding generation task, and at least one of the art asset generation conditions and intermediate execution results, wherein the intermediate execution results are the execution results output by the upstream generation task that meet the art asset generation conditions; The execution of each generated task in the task set includes: According to the instructions of the execution workflow, each generation task in the task set is executed sequentially.

3. The method according to claim 1, wherein, The task set includes a first generation task for drawing art assets, and a second generation task for verifying whether the first execution result of the first generation task meets the conditions for generating the art assets. The process of executing each generation task in the task set, guided by the aforementioned art asset generation conditions, includes: The first generation task is executed using the aforementioned art asset generation conditions as input, and a first execution result is obtained; Using the art asset generation conditions and the first execution result as input, the second generation task is executed to obtain the second execution result; wherein, if the first execution result fails the verification, the second execution result also includes a correction item based on the art asset generation conditions; If the second execution result indicates that the first execution result failed the verification, the first generation task is re-executed with the art asset generation conditions, the correction item, and the first execution result as input to obtain the corrected first execution result.

4. The method according to any one of claims 1 to 3, wherein, The process of executing each generation task in the task set, guided by the aforementioned art asset generation conditions, includes: For each generation task, a matching functional module is searched in the functional module registry, and the matching functional module is called to execute the corresponding generation task guided by the art asset generation conditions.

5. The method according to claim 4, wherein, The functional module registry includes the deployment locations of the registered functional modules, where different deployment locations correspond to different communication protocols. The invocation of the matching functional module, guided by the art asset generation conditions, executes the corresponding generation task, including: By using the communication protocol corresponding to the matching functional module, the matching functional module is invoked to execute the corresponding generation task guided by the art asset generation conditions.

6. The method according to claim 1, wherein, The task set for generating art assets based on the user's needs includes: The user requirements are semantically expanded to generate multiple storyboard scripts with temporal relationships; The task set is constructed based on the multiple storyboard scripts.

7. The method according to claim 6, wherein, The task set includes: a first generation task and a second generation task. The first generation task is used to draw character images in the storyboard. The second generation task is used to verify whether the first execution result of the first generation task meets the art asset generation conditions. If the first execution result fails the verification, the second execution result of the second generation task also includes a correction item compared with the art asset generation conditions. The correction item triggers the re-execution of the first generation task until a target character image that meets the art asset generation conditions is obtained. The task set also includes a number of third generation tasks that correspond one-to-one with the multiple storyboard scripts. The third generation tasks are used to generate corresponding storyboard videos based on the target character image and the corresponding storyboard script. The task set also includes a fourth generation task, which is used to generate the target art asset based on multiple storyboard videos.

8. The method according to claim 7, wherein, The third generation task includes a first subtask and a second subtask; the first subtask is used to generate a storyboard image based on the target character image and the corresponding storyboard script; the second subtask is used to generate a storyboard video based on the storyboard image and dynamic parameters.

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.

Images