Self-generating AI game engine system and operation method

Abstract

The application discloses a self-generating AI game engine system and operation method, which is applied to game production, multimedia or learning material generation mode, develops a new self-generating production mode, completely changes the present production mode and execution mode, and in the game field, the game itself has a self-learning, correction and manufacturing mode, so that the game content is full of change rate.

Description

Technical Field

[0001] This invention belongs to the field of game technology, and in particular relates to a self-generating AI game engine system and its operation method. Background Technology

[0002] The shortcomings of existing game engine generation technology can be roughly summarized into the following seven points: 1. 3D assets: These require art designers, 3D modelers, material painters, character skeleton designers, animators, special effects artists, programmers, etc., to complete. Moreover, once completed, the data volume is huge, and modifications require a lot of time to rebuild, which is time-consuming and extremely costly.

[0003] 2.2D assets: Although there are no 3D modelers, the other processes are the same as for 3D assets.

[0004] 3. The design of scenario modes also requires specialized personnel, and it is often revised. This results in the manual modification of both 3D and 2D materials, which takes a lot of time.

[0005] 4. To see initial results, it usually requires a lot of manpower and a long production time, which results in a large production cost.

[0006] 5. For players or users: It's the same scenario every time, especially the game and multimedia parts. The content of completing the game is exactly the same every time, with no changes.

[0007] 6. If the game is an interactive digital teaching material: the content of the learning materials remains unchanged. It will not automatically reorganize the learning content to a new level as the learner's ability improves. Instead, the learner has to adjust the difficulty himself, but then finds that the difficulty suddenly increases too much, which causes a learning gap.

[0008] 7. Using current game engines requires a lot of learning time. The programming language alone is very difficult. 3D and 2D assets need to be processed after being imported into the engine. In addition, animation and packaging require a lot of learning data, resulting in a very long learning curve.

[0009] How to solve the above problems is a topic that game engine developers are actively working on. Summary of the Invention

[0010] The main purpose of this invention is to provide a self-generating AI game engine system and its operation method. Designers only need to propose a game outline, and then the AI ​​will randomly generate a pattern outline and structure. The designers can then select, add, modify, or increase the weight of elements step by step according to the AI ​​prompts. The AI ​​will then begin to organize the entire game architecture, levels, maps, material features, etc. The AI ​​will also start to assign AI_IDs to create scenes and NPCs. Finally, the designers can play the game, and after confirmation, the large model will be refined, scaled down, and packaged.

[0011] The secondary objective of this invention is to provide a self-generating AI game engine system and its operation method. When the AI ​​engine is installed in the game, the AI ​​changes the plot based on the weight factors authorized by the player due to the player's behavior, which will become diverse. However, since the large model has become closed, only the built-in dedicated large model will be used to change the plot again, making the game content full of random changes.

[0012] To achieve the above objectives, the specific technical solution of the self-generating AI game engine system and operation method of the present invention is as follows: The self-generating AI game engine system consists of: a designer terminal, a user terminal, an AI game generation system, and a logic parsing system. The designer terminal is an electronic device, such as a tablet, computer, or mobile phone, used to connect to the AI ​​game generation system. The designer terminal is used to input modeling keywords and provide the aforementioned keywords to the AI ​​game generation system. The designer terminal also receives and displays the preliminary model generated collaboratively by the AI ​​game generation system and the logic parsing system. The client is also an electronic device, such as a tablet, computer, or mobile phone. The user operates the game through the client and is connected to a game control center. The game control center transmits the key operation process of the client to the AI ​​game generation system. The AI ​​game generation system consists of a large text model, a large image model, a large physics and chemistry model, a large state reasoning model, and a custom large model. Based on the keyword requirements of the designer, the large text model automatically generates a game script. The designer then selects to delete or retain certain plots and elements and regenerates the game script. Based on the keyword requirements of the designer, the large image model generates relevant characters, objects and scenes, and any deficiencies are determined by the large image model itself. Based on the keyword requirements of the designer, this physical and chemical model assigns attributes related to characters, objects, and scenes. Based on the keyword requirements of the designer, this state reasoning model adjusts the attributes of characters, objects, and scenes through the state reasoning module. Through the AI ​​automatic learning principle, the state reasoning model generates dynamic state balance and then generates corresponding balance coefficients for characters, objects, and scenes. The designer can initially set these balance coefficients and set unique attributes for specific characters, objects, or scenes. Furthermore, without the designer's intervention, the state reasoning model automatically balances the game script, characters, objects, and images through the text model, image model, and physical / chemical model. Furthermore, after processing the aforementioned large text model, large image model, large physicochemical model, and large state reasoning model, the AI ​​game generation system generates corresponding object numbers, scene numbers, game script numbers, and character numbers for each object, scene, game script, and character. In addition, the AI ​​game generation system further generates 3D coordinates, depth data, color materials, physical and chemical simulations, small AI behavior models, state and sampling sticker sequences, and then integrates and packages the aforementioned 7 data into a PW point. The logic parsing system generates game models based on the various PW points, game scripts, characters, objects, scenes, related attributes, and balance coefficients generated by the AI ​​game generation system. After receiving the key operation process from the user, it sends the data back to the AI ​​game generation system according to the key weight ratio to correct the game script, characters, objects, scenes, related attributes, and balance coefficients.

[0013] This invention provides a method for game designers to create games using the aforementioned system, comprising: Step 1: Using the AI ​​game generation system's prompts, the designer inputs the game outline in sequence; Step 2: The AI ​​game generation system calculates the random pattern, outline, and structure. Based on the prompts from the AI ​​game generation system, the designer creates characters, objects, scenes, and related attributes. The AI ​​game generation system also generates multiple PW points used in the game. The designer then adjusts the detailed parameters and weight ratios of the AI ​​game generation system. Step 3: The logic parsing system generates models of characters, objects, scenes, and related attributes according to the AI ​​game generation system settings. These models are then sent to the designer's end for initial model testing. After subsequent editing and adjustment of the attributes, the models of characters, objects, scenes, and related attributes are created. Finally, the game is coded and packaged to complete the game production.

[0014] The present invention provides a method for game designers to create games through the above-mentioned system, wherein the game is played by operating a user terminal, the game control center detects the key operation process of the user terminal and transmits it to the AI ​​game generation system, the state reasoning big model of the AI ​​game generation system corrects the game script, characters, objects, scenes and related attributes, and then the logic parsing system provides the user terminal with the game based on the corrected generation model.

[0015] The present invention has the following advantages: 1. PW points are generated through an AI game generation system. No 2D or 3D modeling is required. All of them are sampled by AI. The PW points are used in conjunction with object numbers, scene numbers, game script numbers, and character numbers to complete the process automatically.

[0016] 2. Designers only need to write the game outline, key points, prompts and their weights. The overall game content is generated by AI, or the AI ​​provides suggested options.

[0017] 3. The AI ​​game generation system and the logic analysis system work together, saving a lot of manpower. The preprocessing mode allows users to see the results of the initial model first, and then refine the model later.

[0018] 4. By utilizing the PW point and logic analysis system, the game's plot development is based on changes in the design weights on the designer's end. When the user's end reaches a key operation, the AI ​​game generation system automatically makes changes, so the game content is not easy to repeat and has variability.

[0019] 5. The AI ​​game generation system in this case reorganizes the difficulty level based on the previous execution and results to avoid making the game process too difficult.

[0020] 6. Since the game is entirely executed by AI, the materials and programs are almost all handled by AI. Developers only need to provide ideas, make choices, and adjust weights, and the rest is handled by AI. Developers only need to wait for the results and make appropriate adjustments, eliminating the need for a lot of learning time and manpower required in the past. Attached Figure Description

[0021] Figure 1 This is an overall system diagram of the present invention; Figure 2 The flowchart for creating the game script of the large text model of this invention; Figure 3 This is a flowchart illustrating the establishment of the AI ​​game generation system of the present invention; Figure 4 This is a flowchart of the state reasoning large model of the present invention; Figure 5 An explanatory diagram illustrating the model established for the PW points of this invention; Figure 6A flowchart illustrating the process of establishing PW points in the AI ​​game generation system of this invention; Figure 7 This is a diagram of the PW point processing mode of the present invention; Figure 8 This is a flowchart of the AI ​​game generation module of the present invention when it does not intervene in the game; Figure 9 This is a schematic diagram illustrating how users of this invention generate sidelines in the game flow during key operational processes. Figure 10 This invention provides a flowchart of a new game model generated by an AI game generation module in key operational processes for users. Explanation of markings in the diagram: 10: Designer's End 20: User End 21: Game Control Center 30: AI Game Generation System 31: Large Text Model 32: Large Image Model 33: Large-scale physical and chemical model 34: Large-scale model of state reasoning 35: Custom Large Model 40: Logical Analysis System A: Item list area B: Text or image description area C: Scene Editing Area D: Object Property List E: Object definition list and records F: Object material Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are intended to be within the scope of the invention and form different embodiments. For example, in the claims, any of the claimed embodiments can be used in any combination.

[0024] The following is a reference to the appendix. Figure 1 To be continued Figure 10This invention describes a self-generating AI game engine system and its operation method.

[0025] like Figure 1 As shown, the self-generating AI game engine system and operation method of the present invention comprises: a designer terminal 10, a user terminal 20, an AI game generation system 30, and a logic parsing system 40.

[0026] like Figure 1 As shown, the designer terminal 10 is an electronic device, which can be a tablet, a computer, or a mobile phone, used to connect to the AI ​​game generation system 30. The designer terminal 10 is used to input keywords for modeling and provides the aforementioned keywords to the AI ​​game generation system 30. The designer terminal 10 also receives and displays the preliminary model generated collaboratively by the AI ​​game generation system and the logic analysis system 40.

[0027] like Figure 1 As shown, the user terminal 20 is also an electronic device, which can be a tablet, a computer, or a mobile phone. The user terminal 20 is used to perform game operations. The user terminal 20 is connected to a game control center 21, through which the game control center 21 transmits the key operation process of the user terminal 20 to the AI ​​game generation system 30.

[0028] like Figure 1 As shown, the AI ​​game generation system 30 is composed of a large text model 31, a large image model 32, a large physics and chemistry model 33, a large state reasoning model 34, and a custom large model 35.

[0029] Please see Figure 1 and Figure 2 The text model 31 automatically generates a game script based on the keyword requirements of the designer's terminal 10. The designer can then choose to delete or retain certain plots and elements to regenerate the game script.

[0030] Please see Figure 1 , Figure 3 Given that the conventional method of creating game models requires designers to learn how to create 2D and 3D objects, asset databases, and program coding IDEs, which is quite complex and requires a lot of manpower, resulting in high costs, the large image model 32 generates relevant characters, objects, and scenes based on the keyword requirements of the designer's end 10, and the large image model 32 automatically judges any deficiencies.

[0031] like Figure 1 As shown, the physical and chemical model 33 assigns character, object and scene-related attributes based on the keyword requirements of the designer end 10.

[0032] Please see Figure 5 As shown, the method of establishing the game model using the large image model 32 and the large physical and chemical model 33 is further explained, such as... Figure 5 In the object list area A, only the designer specifies the main objects; other secondary objects are defined by the AI ​​according to the context set by the designer. Figure 5 Section B, which contains text or images, allows designers to describe the objects or storylines they want to create using text or reference images. The AI ​​will automatically identify any shortcomings and provide suggestions or modifications. There is no program section, so no programming is required. Figure 5 Scene editing area C: Designers can customize the position of objects, or the AI ​​can determine it automatically. Object material F is the edge area of ​​the object, and the star indicates the scene number. In this cubic area, the prototype can be displayed or the object style does not need to be displayed. Figure 5 Object Property List D: Lists the properties contained in the currently edited object, divided into custom and AI-added sections. Figure 5 Object definition list and records E: Lists the object records that the designer has edited. For example... Figure 4 The object material F in the settings can be selected from materials such as NPCs, NPCBOSSes, houses, buildings, roads, trees, stones, haystacks, flames, special effects, starry sky, etc. All objects defined by the user will be displayed in this way. AI-generated materials will not appear in the window unless the system switches to preview display mode. Moreover, the AI ​​game generation system will assign corresponding numbers (AI_ID) to the generated characters, objects, scenes, materials, etc., for subsequent editing.

[0033] like Figure 6 As shown, further explanation regarding object material F is as follows: Figure 5 If an object material F contains only one star, it indicates lower pixel count and faster loading speed. Conversely, if an object material F contains multiple stars, it indicates higher detail and more data to load.

[0034] like Figure 1 , Figure 4 As shown, the state reasoning model 34, based on the keyword requirements of the designer's end 10, adjusts the attributes of characters, objects, and scenes through the state reasoning module 34. Using AI automatic learning principles, the state reasoning model 34 generates dynamic state balance, and then generates corresponding balance coefficients for characters, objects, and scenes. The designer's end 10 can initially set these balance coefficients and set unique attributes for specific characters, objects, or scenes, and further... Figure 7As shown, without interference from the designer end 10, the state reasoning big model 34 automatically balances the game script, characters, objects, and images through the text big model 31, image big model 32, and physical and chemical big model 33.

[0035] like Figure 7 As shown, the AI ​​game generation system 30, after processing by the aforementioned large text model 31, large image model 32, large physicochemical model 33, and large state reasoning model 34, generates corresponding object numbers, scene numbers, game script numbers, and character numbers for each object, scene, game script, and character. Furthermore, the AI ​​game generation system 30 further generates 3D coordinates, depth data, color and material properties, physicochemical simulation, AI behavior mini-model, state, and sampling patch sequences. Then, it integrates and packages the aforementioned seven data points into a PW point (information point, hereinafter referred to as PW point). The aforementioned PW point can then be used in game scripts, objects, scenes, and characters, and as... Figure 6 The weight is increased or decreased based on the number of stars.

[0036] This custom large model 35 series can directly change or modify the AI ​​game generation system 30, allowing for overall adjustments and modifications to the game.

[0037] like Figure 1 , Figure 7 As shown, the logic parsing system 40 generates a game model based on the various PW points, game scripts, characters, objects, scenes, related attributes, balance coefficients, etc. generated by the AI ​​game generation system 30. After receiving the key operation process from the user terminal 20, it sends the key weight ratio back to the AI ​​game generation system 30 to correct the game scripts, characters, objects, scenes, related attributes, and balance coefficients.

[0038] The above is an introduction to the system of the present invention. Next, the method and process of creating a game according to the present invention will be introduced: Step 1: Using the designer terminal 10, input the game outline in sequence according to the prompts of the AI ​​game generation system 30.

[0039] Step 2: The AI ​​game generation system 30 calculates a random pattern, outline, and structure. Based on the prompts from the AI ​​game generation system 30, the designer creates characters, objects, scenes, and related attributes. The AI ​​game generation system 30 also generates multiple PW points used in the game. The designer then adjusts the detailed parameters and weighting ratios of the AI ​​game generation system 30. If the designer has some understanding of game design, they can directly manipulate the custom large model 35 to adjust attributes such as the game script, characters, objects, and scenes.

[0040] Step 3: The logic parsing system 40 generates models of characters, objects, scenes and related attributes according to the settings of the AI ​​game generation system 30 and sends them to the designer end 10 for initial model testing. After subsequent editing and adjustment of attributes, the models of characters, objects, scenes and related attributes are generated. Finally, the game production is completed by coding and packaging.

[0041] Because current game engines are loaded within the game itself, providing gameplay variety and preventing the game content from becoming too monotonous, therefore, Figures 8 to 10 As shown, after the user terminal 20 (player) starts the game, the game control center 21 detects the key operation process of the user terminal 20, such as... Figure 8 The status indicates that permissions have not been granted, and the game will follow a single path.

[0042] like Figure 9 , Figure 10 As shown, when the AI ​​game generation system 30 is granted access, the game control center 21 detects the key operation process of the user terminal 20 and transmits the information to the AI ​​game generation system 30. The state reasoning big model 34 of the AI ​​game generation system 30 modifies the game script, characters, objects, scenes and related attributes. Then, the logic parsing system 40 generates different game route branches based on the modified generation model and provides the user terminal with the game to play.

[0043] Therefore, the features and effects of this invention are briefly described as follows: 1. PW points are generated through the AI ​​game generation system 30. No 2D or 3D modeling is required. The entire process is completed by AI sampling and PW points, along with object numbers, scene numbers, game script numbers, and character numbers, in an automated manner.

[0044] 2. Designers only need to write the game outline, key points, prompts and their weights. The overall game content is generated by AI, or the AI ​​provides suggested options.

[0045] 3. The AI ​​game generation system 30 and the logic analysis system 40 work together, saving a lot of manpower. The preprocessing mode allows users to see the results of the initial model first, and then refine the model later.

[0046] 4. By utilizing the PW point and logic analysis system 40, the game plot can be changed according to the design weight changes of the designer terminal 10. When the user terminal 20 reaches the key operation process, the AI ​​game generation system 30 will automatically change it. Therefore, the game content is not easy to repeat and has variability.

[0047] 5. The AI ​​game generation system 30 in this case reorganizes the difficulty level based on the previous execution and results to avoid making the game process too difficult.

[0048] 6. Since the game is entirely executed by AI, the materials and programs are almost all handled by AI. Developers only need to provide ideas, make choices, and adjust weights, and the rest is handled by AI. Developers only need to wait for the results and make appropriate adjustments, eliminating the need for a lot of learning time and manpower required in the past.

[0049] In conclusion, this invention utilizes [the technology / method]; and its structure has not been seen in any books or publications or publicly used, thus meeting the requirements for an invention patent application. We earnestly request your esteemed authority to grant the patent as soon as possible, and we are deeply grateful.

[0050] It should be noted that the above description is of specific embodiments of the present invention and the technical principles used. Any changes made in accordance with the concept of the present invention that do not exceed the spirit of the specification and drawings should be included within the scope of the present invention.

Claims

1. A self-generating AI game engine system, characterized in that, It consists of: a designer terminal, a user terminal, an AI game generation system, and a logic parsing system; The designer terminal is an electronic device, such as a tablet, computer, or mobile phone, used to connect with the AI ​​game generation system. The designer terminal is used to input modeling keywords and provide the aforementioned keywords to the AI ​​game generation system. The designer terminal also receives and displays the preliminary model generated collaboratively by the AI ​​game generation system and the logic parsing system. The client is also an electronic device, such as a tablet, computer, or mobile phone. The user operates the game through the client and is connected to a game control center. The game control center transmits the key operation process of the client to the AI ​​game generation system. The AI ​​game generation system consists of a large text model, a large image model, a large physics and chemistry model, a large state reasoning model, and a custom large model. Based on the keyword requirements of the designer, the large text model automatically generates a game script. The designer then selects to delete or retain certain plots and elements and regenerates the game script. Based on the keyword requirements of the designer, the large image model generates relevant characters, objects and scenes, and any deficiencies are determined by the large image model itself. Based on the keyword requirements of the designer, this physical and chemical model assigns attributes related to characters, objects, and scenes. Based on the keyword requirements of the designer, this state reasoning model adjusts the attributes of characters, objects, and scenes through the state reasoning module. Through the AI ​​automatic learning principle, the state reasoning model generates dynamic state balance and then generates corresponding balance coefficients for characters, objects, and scenes. The designer can initially set these balance coefficients and set unique attributes for specific characters, objects, or scenes. Furthermore, without the designer's intervention, the state reasoning model automatically balances the game script, characters, objects, and images through the text model, image model, and physical / chemical model. Furthermore, after processing the aforementioned large text model, large image model, large physicochemical model, and large state reasoning model, the AI ​​game generation system generates corresponding object numbers, scene numbers, game script numbers, and character numbers for each object, scene, game script, and character. In addition, the AI ​​game generation system further generates 3D coordinates, depth data, color materials, physical and chemical simulations, small AI behavior models, state and sampling sticker sequences, and then integrates and packages the aforementioned 7 data into a PW point. The logic parsing system generates game models based on the various PW points, game scripts, characters, objects, scenes, related attributes, and balance coefficients generated by the AI ​​game generation system. After receiving the key operation process from the user, it sends the data back to the AI ​​game generation system according to the key weight ratio to correct the game script, characters, objects, scenes, related attributes, and balance coefficients.

2. The self-generating AI game engine system according to claim 1, characterized in that, The AI ​​game generation system includes a custom large model, which allows for direct changes or modifications to the parameters of the AI ​​game generation system, enabling overall adjustments and modifications to the game.

3. A method for generating a game using the system of claim 1, characterized in that, Includes: Step 1: Using the AI ​​game generation system's prompts, the designer inputs the game outline in sequence; Step 2: The AI ​​game generation system calculates the random pattern, outline, and structure. Based on the prompts from the AI ​​game generation system, the designer creates characters, objects, scenes, and related attributes. The AI ​​game generation system also generates multiple PW points used in the game. The designer then adjusts the detailed parameters and weight ratios of the AI ​​game generation system. Step 3: The logic parsing system generates models of characters, objects, scenes, and related attributes according to the AI ​​game generation system settings. These models are then sent to the designer's end for initial model testing. After subsequent editing and adjustment of the attributes, the models of characters, objects, scenes, and related attributes are created. Finally, the game is coded and packaged to complete the game production.

4. A method for generating a game using the system of claim 1, characterized in that, By operating the user terminal to play the game, the game control center detects the key operation process of the user terminal and transmits it to the AI ​​game generation system. The state reasoning model of the AI ​​game generation system corrects the game script, characters, objects, scenes and related attributes. Then, the logic parsing system generates the model based on the correction and provides the user terminal with the game to play.

Images