Novel movie and television split mirror generation method based on large language model

Abstract

The invention relates to a novel movilization split mirror generation method based on a large language model, which comprises the following steps of: identifying roles, scenes and key props appearing for the first time in a text, and extracting specific descriptions of the roles, the scenes and the key props; roles, scenes and key props are updated and supplemented; the recognized content, the extracted description and the updated and supplemented setting information are structurally stored in a setting library to form a dynamically-updated and globally-shared knowledge source, so that a dynamic setting library is constructed and maintained; the large language model receives the original text content of the current chapter and all setting information related to the chapter and retrieved from the dynamic setting library; the large language model follows the description in the dynamic setup library and converts the narrative text into a structured split script to generate a movie and television split. According to the method, the novel text can be automatically converted into the movie and television play split script, and the production efficiency of movie and television works is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence film and television production technology, and in particular to a method for generating storyboards for novel film and television adaptations based on a large language model. Background Art

[0002] Currently, adapting novels into film and television productions relies primarily on extensive manual re-creation by screenwriters and storyboard artists, a time-consuming, expensive, and highly specialized process. Directly using large language models for processing long texts like novels presents significant shortcomings: they lack the ability to retain the core settings of the entire text (such as character imagery and scene features). When models process text in segments, inconsistencies can easily arise, such as conflicting descriptions of the same character's appearance in different chapters. This issue results in a lack of coherence in the generated work, making it unsuitable for direct use in professional film and television production.

[0003] Therefore, it is necessary to provide a method for generating storyboards for novels and films based on a large language model, so as to automatically convert novel texts into storyboard scripts for film and television dramas and improve the production efficiency of film and television works. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for generating storyboards for film and television adaptations of novels based on a large language model, so as to automatically convert novel texts into storyboard scripts for film and television dramas and improve the production efficiency of film and television works.

[0005] In order to solve the problems existing in the prior art, the present invention provides a method for generating storyboards for a novel film and television adaptation based on a large language model, characterized by comprising the following steps:

[0006] Identify the first appearance of characters, scenes, and key props in the text and extract their detailed descriptions;

[0007] Update and supplement characters, scenes and key props;

[0008] The identified content, extracted descriptions, updated and supplemented setting information are stored in a structured manner in the setting library, forming a dynamically updated and globally shared knowledge source, thereby building and maintaining a dynamic setting library;

[0009] The large language model receives the original content of the current chapter and all the setting information related to this chapter retrieved from the dynamic setting library;

[0010] The large language model follows the descriptions in the dynamic setting library to convert narrative text into structured storyboards to generate film and television storyboards.

[0011] Optionally, in the novel film and television storyboard generation method based on the large language model, when constructing and maintaining the dynamic setting library, the novel text is processed sequentially in chapters.

[0012] Optionally, in the novel film and television storyboard generation method based on the large language model,

[0013] The character's detailed description includes appearance, clothing, personality, and character's personality;

[0014] The detailed description of the scene includes the atmosphere and layout of the environment;

[0015] The detailed description of key props includes the shape of the props and the function of the props.

[0016] Optionally, in the novel film and television storyboard generation method based on the large language model,

[0017] Updates and additions to characters include changing clothes, changing hairstyles, and adjusting appearances according to age;

[0018] The updating and supplementation of scenes include changing the environment according to the seasons and updating the scenes due to the destruction of the plot;

[0019] The update and supplement of key props include the upgrade of key props.

[0020] Optionally, in the novel film and television storyboard generation method based on the large language model, each storyboard includes scene size, camera movement, picture content, character actions, dialogues and sound effects.

[0021] Compared with the prior art, the present invention has the following advantages:

[0022] (1) Ensure consistency: The dynamic setting library mechanism fundamentally solves the problem of context forgetting in long text processing, ensuring the coherence and unity of visual elements such as characters and scenes throughout the entire work.

[0023] (2) Improved adaptation efficiency and reduced costs: The time-consuming and labor-intensive manual steps in the traditional adaptation process were automated, which greatly shortened the pre-production cycle and reduced labor costs.

[0024] (3) Achieved structuring and usability: The generated storyboard format is clear and complete in elements, which is highly practical and can be seamlessly connected to subsequent AI generation processes or manual production links.

[0025] (4) The present invention automatically converts novel texts into storyboards for film and television dramas, thereby improving the production efficiency of film and television works. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flowchart of a method for generating storyboards for a film or television adaptation of a novel provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following is a more detailed description of the specific embodiments of the present invention with reference to schematic diagrams. The advantages and features of the present invention will become more apparent from the following description. It should be noted that the drawings are greatly simplified and not to exact scale, and are only used for the purpose of conveniently and clearly illustrating the embodiments of the present invention.

[0028] In the description of the present application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0029] Hereinafter, if the method described herein includes a series of steps, the order in which the steps are presented herein is not necessarily the only order in which the steps may be performed, and some of the steps described may be omitted and / or some other steps not described herein may be added to the method.

[0030] Currently, adapting novels into film and television productions relies primarily on extensive manual re-creation by screenwriters and storyboard artists, a time-consuming, expensive, and highly specialized process. Directly using large language models for processing long texts like novels presents significant shortcomings: they lack the ability to retain the core settings of the entire text (such as character imagery and scene features). When models process text in segments, inconsistencies can easily arise, such as conflicting descriptions of the same character's appearance in different chapters. This issue results in a lack of coherence in the generated work, making it unsuitable for direct use in professional film and television production.

[0031] In order to solve the problems existing in the prior art, the present invention provides a method for generating storyboards for novels based on a large language model. Figure 1 As shown, the storyboard generation method includes the following steps:

[0032] Build and maintain a dynamic setting library. The construction and maintenance of the dynamic setting library runs through the entire processing process. This invention processes the novel text sequentially by chapter. When processing each chapter, the large language model first plays the role of "setting analyst" in the following way:

[0033] Identification and extraction: Identify the characters, scenes, and key props that appear for the first time in the text and extract their detailed descriptions. The detailed description of the characters includes appearance, clothing, personality, and character setting; the detailed description of the scenes includes the environment atmosphere and layout; the detailed description of the key props includes the shape and function of the props.

[0034] Updates and Supplements: Updates and supplements to characters, scenes, and key props. Character updates and supplements include changing clothes, changing hairstyles, and adjusting appearances according to age; scene updates and supplements include changing environments according to seasons, and updating scenes due to plot destruction; key prop updates and supplements include upgrading key props.

[0035] Structured storage: Identified content, extracted descriptions, and updated and supplemented setting information are structured and stored in a setting library, forming a dynamically updated, globally shared knowledge source; this setting library ensures that all core visual elements have consistent definitions throughout the story.

[0036] After extracting and storing the settings for a chapter, storyboards are generated based on the dynamic settings library. The storyboard generation phase begins. At this point, the large language model plays the role of both "director" and "storyboard artist" as follows:

[0037] Joint input: The large language model receives the original content of the current chapter and all the setting information related to this chapter retrieved from the dynamic setting library (such as the complete appearance of the characters in this chapter, the detailed layout of the scene, etc.);

[0038] Constrained Generation: When generating storyboards, the large language model is forced to strictly adhere to the descriptions in the dynamic setting library. For example, if a character in the dynamic setting library is defined as having a scar on the right corner of the eye, then even if the text in the current chapter does not mention this feature again, the large language model must include "scar on the right corner of the eye" in the storyboard description when generating close-ups or close-ups of that character.

[0039] Formatted output: Convert narrative text into a structured storyboard to generate film-like storyboards. Ideally, each storyboard should include key elements such as shot size, camera movement, image content, character actions, dialogue, and sound effects, allowing it to be directly used in subsequent AI video generation or as a blueprint for human directors.

[0040] This invention is a novel storyboard generation method that processes novels chapter by chapter. This method uses a large language model to process the input novel chapter by chapter, automatically extracting and maintaining a global library of character and scene settings, and then efficiently generating structured storyboards based on the content of each chapter to facilitate subsequent film and television production needs.

[0041] In summary, the present invention has the following advantages compared with the prior art:

[0042] (1) Ensure consistency: The dynamic setting library mechanism fundamentally solves the problem of context forgetting in long text processing, ensuring the coherence and unity of visual elements such as characters and scenes throughout the entire work.

[0043] (2) Improved adaptation efficiency and reduced costs: The time-consuming and labor-intensive manual steps in the traditional adaptation process were automated, which greatly shortened the pre-production cycle and reduced labor costs.

[0044] (3) Achieved structuring and usability: The generated storyboard format is clear and complete in elements, which is highly practical and can be seamlessly connected to subsequent AI generation processes or manual production links.

[0045] (4) The present invention automatically converts novel texts into storyboards for film and television dramas, thereby improving the production efficiency of film and television works.

[0046] The above description is merely a preferred embodiment of the present invention and does not limit the present invention in any way. Any person skilled in the art who, without departing from the scope of the present invention, makes any equivalent substitution, modification, or other changes to the technical solution and technical content disclosed in the present invention shall be deemed to be within the scope of the present invention and still fall within the scope of protection of the present invention.

Claims

1. A method for generating storyboards for film and television adaptations of novels based on a large language model, characterized in that: The following steps are involved: Identify the first appearance of characters, scenes, and key props in the text and extract their detailed descriptions; Update and supplement characters, scenes and key props; The identified content, extracted descriptions, updated and supplemented setting information are stored in a structured manner in the setting library, forming a dynamically updated and globally shared knowledge source, thereby building and maintaining a dynamic setting library; The large language model receives the original content of the current chapter and all the setting information related to this chapter retrieved from the dynamic setting library; The large language model follows the descriptions in the dynamic setting library to convert narrative text into structured storyboards to generate film and television storyboards.

2. The method for generating storyboards for a novel film and television adaptation based on a large language model according to claim 1, wherein: When building and maintaining a dynamic setting library, the novel text is processed sequentially in chapters.

3. The method for generating storyboards for a novel film and television adaptation based on a large language model as claimed in claim 1, wherein: The character's detailed description includes appearance, clothing, personality, and character's personality; The detailed description of the scene includes the atmosphere and layout of the environment; The detailed description of key props includes the shape of the props and the function of the props.

4. The method for generating storyboards for a novel film and television adaptation based on a large language model as claimed in claim 3, wherein: Updates and additions to characters include changing clothes, changing hairstyles, and adjusting appearances according to age; The updating and supplementation of scenes include changing the environment according to the seasons and updating the scenes due to the destruction of the plot; The update and supplement of key props include the upgrade of key props.

5. The method for generating storyboards for a novel film and television adaptation based on a large language model as claimed in claim 1, wherein: Each storyboard includes shot size, camera movement, picture content, character actions, dialogue and sound effects.

Images