Method and system for generating a tutorial video

Abstract

Embodiments of the present specification provide a teaching video generation method and system. In the method, the teaching video generation system can obtain text unit content based on a teaching original text, and perform complexity evaluation on the text unit content to obtain a complexity evaluation result. Then the teaching video generation system determines a target rendering mode from a plurality of preset rendering modes according to the complexity evaluation result, generates a corresponding target code according to the target rendering mode, and runs the target code through an animation engine to generate a video unit.

Description

Technical Field

[0001] This specification relates to the field of video generation technology, and in particular to a method and system for generating instructional videos. Background Technology

[0002] With the rapid development of information technology, the construction of digital educational resources has become an important part of the modern education system. As an intuitive and vivid carrier of knowledge, instructional videos provide students with a flexible and personalized learning experience, and have therefore received widespread attention and application.

[0003] Most methods for generating instructional videos in related technologies rely on manual production, requiring teachers or professional producers to invest a lot of time and energy in a series of complex processes such as scriptwriting, material preparation, animation design, video editing and compositing.

[0004] However, with the rapid updating of teaching resources, this method of generating teaching videos is inefficient and cannot meet the needs of large-scale, rapidly updated teaching resource development.

[0005] The information in the background section is merely information known only to the inventor and does not imply that such information had entered the public domain before the date of this application, nor does it imply that it can be considered prior art in this disclosure. Summary of the Invention

[0006] This manual provides a method and system for generating instructional videos. After obtaining the content of a text unit, the system determines the corresponding target rendering method based on the complexity evaluation result of the text unit content, and calls an animation engine based on the target rendering method to automatically generate video units corresponding to the text unit content.

[0007] Firstly, this specification provides a method for generating instructional videos. The method includes: acquiring text unit content, the text unit content being derived from original instructional text; performing a complexity assessment on the text unit content to obtain a complexity assessment result; determining a target rendering method from multiple preset rendering methods based on the complexity assessment result of the text unit content; wherein different complexity assessment results correspond to different preset rendering methods, and different preset rendering methods have different visual display strategies; generating target code corresponding to the text unit content based on the target rendering method, the target code being executable by an animation engine; and calling the animation engine to run the target code to generate a video unit corresponding to the text unit content.

[0008] In some embodiments, the visual display strategy includes at least one of the following dimensions: animation usage dimension, element display dimension, time control dimension, and information complexity dimension.

[0009] In some embodiments, the complexity evaluation results are divided into multiple levels. The target rendering method corresponding to low complexity has higher animation complexity in the animation usage dimension, more elements displayed in the element display dimension, shorter pause time in the time control dimension, and / or higher information complexity in the information complexity dimension than the target rendering method corresponding to high complexity.

[0010] In some embodiments, the step of evaluating the complexity of the text unit content to obtain a complexity evaluation result includes: obtaining the intrinsic complexity of the text unit content, wherein the intrinsic complexity is used to characterize the content complexity of the text unit content; and / or obtaining the extrinsic complexity of the text unit content, wherein the extrinsic complexity is used to characterize the expression complexity of the text unit content; and determining the complexity evaluation result based on the intrinsic complexity and / or the extrinsic complexity.

[0011] In some embodiments, obtaining the intrinsic complexity of the text unit content includes: obtaining the intrinsic complexity of the text unit content based on at least one of the following: the type and number of mathematical symbols contained in the text unit content; the correspondence between conceptual nouns in the text unit content and preset levels; and the structural complexity of the formulas contained in the text unit content.

[0012] In some embodiments, obtaining the external complexity of the text unit content includes: obtaining the external complexity of the text unit content based on at least one of the following: the text density contained in the text unit content; the term density contained in the text unit content; and the information organization complexity of the text unit content.

[0013] In some embodiments, obtaining text unit content includes: obtaining original teaching text; performing structured parsing on the original teaching text to segment the original teaching text into structured data containing at least one chapter, wherein the chapter includes: chapter title and chapter content; and obtaining text unit content based on the chapters in the structured data.

[0014] In some embodiments, the step of performing structured parsing on the original teaching text and segmenting the original teaching text into structured data containing at least one chapter includes: calling a large language model to analyze the original teaching text to obtain the directory structure or logical structure of the original teaching text; and performing structured parsing on the original teaching text based on the directory structure or logical structure to segment the original teaching text into structured data containing at least one chapter.

[0015] In some embodiments, obtaining text unit content based on chapters in the structured data includes: classifying the chapter content of the chapters into explanation stages to obtain a preset number of explanation stages corresponding to the chapters; and determining a target explanation stage among the preset number of explanation stages, and obtaining the text content of the target explanation stage as the text unit content.

[0016] In some embodiments, the method further includes: performing a quality evaluation on a preset number of explanation stages corresponding to the chapter; and modifying the preset number of explanation stages when the quality evaluation result indicates failure, until the quality evaluation result indicates success; wherein the quality evaluation includes: text coherence evaluation and knowledge point coverage evaluation.

[0017] In some embodiments, the method further includes: obtaining video units corresponding to each explanation stage in the chapter and splicing them together to obtain the teaching video corresponding to the chapter.

[0018] In some embodiments, the method further includes: inputting the teaching video and preset scoring guidance words into a large language model to obtain the evaluation result output by the large language model; when the evaluation result indicates pass, outputting the teaching video to the user; and when the evaluation result indicates fail, inputting the evaluation result into the large language model so that the large language model regenerates the target code based on the evaluation result until the evaluation result of the teaching video corresponding to the modified target code indicates pass.

[0019] In some embodiments, generating target code corresponding to the text unit content based on the target rendering method includes: obtaining rendering prompt words corresponding to the target rendering method; and inputting the text unit content and the rendering prompt words into a large language model, and obtaining the target code generated by the large language model.

[0020] In some embodiments, the method further includes: obtaining modification guidance information from the user for the video unit; The modification guidance information is input into the large language model to obtain the modified target code output by the large language model; and the animation engine is called to run the modified target code to update the video unit corresponding to the text unit content.

[0021] Secondly, this specification also provides a system for generating instructional videos, including at least one storage medium and at least one processor. The at least one storage medium stores at least one instruction set for generating instructional videos. The at least one processor is communicatively connected to the at least one storage medium, wherein the at least one processor reads the at least one instruction set during operation and executes the method described in any of the first aspects above according to the instructions of the at least one instruction set.

[0022] As can be seen from the above technical solutions, the teaching video generation method and system provided in this specification achieve standardized processing of teaching materials by acquiring text unit content based on the original teaching text. Next, the teaching video generation system performs complexity assessment on the text unit content to quantitatively analyze its complexity. Then, based on the complexity assessment results, the system determines the target rendering method from multiple preset rendering methods, thereby establishing a mapping relationship between text unit content and visual display strategies. Subsequently, the system generates target code executable by an animation engine based on the target rendering method, then calls the animation engine to run the target code and generate video units, thus forming an automated conversion from text unit content to video units. In the above solution, the teaching video generation system, by establishing a dynamic correlation between the complexity of text unit content and the rendering method, improves teaching effectiveness while automatically generating video units corresponding to text unit content, thereby solving the problem of low efficiency in manual production in related technologies and achieving rapid generation of teaching videos.

[0023] The methods for generating instructional videos and other functions of the system provided in this manual will be partially listed in the following description. The inventive aspects of the methods and systems for generating instructional videos provided in this manual can be fully explained through practice or by using the methods, apparatus, and combinations described in the detailed examples below. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A schematic diagram illustrating an application scenario of an instructional video generation system provided according to an embodiment of this specification is shown. Figure 2 A schematic diagram of the hardware structure of a computing device provided according to some embodiments of this specification is shown; Figure 3 A flowchart illustrating a method for generating an instructional video according to an embodiment of this specification is shown. Figure 4 A flowchart illustrating the process of obtaining text unit content according to an embodiment of this specification is shown; Figure 5 A flowchart illustrating the process of obtaining text unit content according to another embodiment of this specification is shown; and Figure 6 A schematic diagram of a complexity assessment process provided according to an embodiment of this specification is shown. Detailed Implementation

[0026] The following description provides specific application scenarios and requirements for this specification, intended to enable those skilled in the art to make and use the contents of this specification. Various partial modifications to the disclosed embodiments will be apparent to those skilled in the art, and the general principles defined herein can be applied to other embodiments and applications without departing from the spirit and scope of this specification. Therefore, this specification is not limited to the embodiments shown, but rather to the widest scope consistent with the claims.

[0027] The terminology used herein is for the purpose of describing particular exemplary embodiments only and is not restrictive. For example, unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” used herein may also include the plural forms. When used in this specification, the terms “comprising,” “including,” and / or “containing” mean that the associated integers, steps, operations, elements, and / or components are present, but do not exclude the presence of one or more other features, integers, steps, operations, elements, components, and / or groups, or that other features, integers, steps, operations, elements, components, and / or groups may be added to the system / method.

[0028] Considering the following description, these and other features of this specification, as well as the operation and function of the related components of the structure, and the economy of assembly and manufacture of the parts, can be significantly improved. All of these form part of this specification with reference to the accompanying drawings. However, it should be clearly understood that the drawings are for illustrative and descriptive purposes only and are not intended to limit the scope of this specification. It should also be understood that the drawings are not drawn to scale.

[0029] The flowcharts used in this specification illustrate operations implemented according to some embodiments of this specification. It should be clearly understood that the operations in the flowcharts may not be implemented in a sequential order. Instead, the operations may be implemented in reverse order or simultaneously. Furthermore, one or more additional operations may be added to the flowcharts. One or more operations may be removed from the flowcharts.

[0030] In this specification, "X includes at least one of A, B, or C" means that X includes at least A, or X includes at least B, or X includes at least C. That is, X may include only one of A, B, and C, or any combination of A, B, and C, as well as other possible content / elements. The arbitrary combination of A, B, and C can be A, B, C, AB, AC, BC, or ABC.

[0031] In this specification, unless explicitly stated otherwise, the relationships between structures can be direct or indirect. For example, when describing "A is connected to B," unless it is explicitly stated that A and B are directly connected, it should be understood that A can be directly connected to B or indirectly connected to B. Similarly, when describing "A is on top of B," unless it is explicitly stated that A is directly above B (AB is adjacent and A is above B), it should be understood that A can be directly above B or indirectly above B (AB is separated by other elements, and A is above B). And so on.

[0032] It should be noted that the user data obtained in this manual is authorized by the user and does not involve user privacy.

[0033] For ease of description, the terms that will appear later in this manual will be explained first.

[0034] Large language models are commonly used in the field of artificial intelligence in natural language processing (NLP), especially referring to large machine learning models with a large number of parameters and computational resources. Large language models are named for their huge number of parameters and complex network structures. They have powerful feature representation and feature understanding capabilities, and can better capture patterns and regularities in data when dealing with complex tasks. They are designed and trained to better understand and generate natural language.

[0035] Optical Character Recognition (OCR) is the process of analyzing and recognizing textual information from image files to obtain text and layout information. In this process, various types of images, such as handwritten and printed text, are converted into machine-encoded text. The core of OCR technology is recognizing textual information in images and converting it into a computer-editable and processable text format.

[0036] Manim is a high-precision animation engine designed for STEM fields such as mathematics and physics. Driven by Python code, it allows for precise control of animation elements (graphics, formulas, etc.) and motion logic. It natively adapts to mathematical derivations and geometric displays. The current mainstream version is the community-maintained Manim Community Edition, which is widely used in academic presentations, popular science videos, STEM teaching, and other scenarios to help users transform abstract mathematical concepts into intuitive dynamic visuals.

[0037] SurroundingRectangle: A Manim function that can add a rectangular border to a formula or text box, which is the most direct technical means to achieve "single focus emphasis".

[0038] The method provided in this manual is applicable to scenarios where video units are automatically generated from textual content in original teaching materials, and the rendering method is dynamically determined based on the complexity of the textual unit content. Examples include: production of math courses, physics courses, or popular science videos.

[0039] However, current methods for generating instructional videos often rely on fixed templates to visually transform text content or render content based on uniform rules. These fixed templates or uniform rules are ill-suited to the actual difficulty of the instructional content and the cognitive load required by users. The higher the actual difficulty (complexity) of the instructional content, the greater the cognitive load required by users; conversely, the lower the actual difficulty, the lower the cognitive load required. Therefore, instructional videos generated using these static methods may suffer from problems such as insufficient presentation of simple content, cognitive overload of complex content, or low learning efficiency in real-world teaching scenarios.

[0040] To address this, this specification provides a method for generating instructional videos. The system can acquire text unit content based on the original instructional text and perform a complexity assessment on the text unit content to obtain a complexity assessment result. Then, based on the complexity assessment result, the system determines a target rendering method from multiple preset rendering methods, generates corresponding target code according to the target rendering method, and generates video units through an animation engine.

[0041] Using the method provided in this manual, the instructional video generation system can determine the target rendering method from multiple preset rendering modes based on the complexity assessment results of the text unit content. The higher the actual difficulty of the text unit content, the higher its corresponding complexity assessment result; conversely, the lower the actual difficulty of the text unit content, the lower its corresponding complexity assessment result. That is, when different text unit content has different complexity assessment results, the determined target rendering method may differ. For example, a high animation complexity rendering method may be used for low-complexity text unit content; or, a low animation complexity rendering method may be used for high-complexity text unit content. The visual display strategies in the final generated video units may differ, thus enabling the instructional video generation system to dynamically adjust the content expression method of the video units according to the complexity of the text unit content. This overcomes the limitations of a single rendering method and improves the adaptability between video generation quality and teaching effectiveness.

[0042] It should be noted that the above description of application scenarios is only one of the many usage scenarios provided in this specification. Those skilled in the art should understand that when the method and system for generating instructional videos provided in this specification are applied to other usage scenarios, the implementation methods and technical effects are similar.

[0043] Figure 1 A schematic diagram of an application scenario 100 of an instructional video generation system 130 provided according to an embodiment of this specification is shown. For example... Figure 1 As shown, the application scenario may include: a teaching video generation system 130 (hereinafter referred to as: generation system 130).

[0044] In some embodiments, the generation system 130 may be a system that provides video generation services. The generation system 130 may obtain the original teaching text for the video to be generated from the database 110; or, the generation system 130 may obtain the original teaching text uploaded by the user through the user terminal 120. Furthermore, the generation system 130 is equipped with preset rendering methods, which may include multiple rendering methods, or a mapping relationship between complexity evaluation results and multiple rendering methods.

[0045] Next, the generation system 130 evaluates the complexity of the text unit content and determines the target rendering method from multiple preset rendering methods based on the complexity evaluation results. Then, the generation system generates the target code corresponding to the text unit content based on the target rendering method, and executes the target code through the animation engine to generate the video unit corresponding to the text unit content.

[0046] The generation system 130 is a computing system with a certain computing capability. The generation system 130 can correspond to a single computing device or a computing cluster composed of multiple computing devices. The generation system 130 can be deployed locally or remotely. In this case, the physical device corresponding to the generation system 130 can store data or instructions for executing the data processing methods described in this specification, and can execute or be used to execute said data or instructions.

[0047] In some embodiments, the physical device corresponding to the generation system 130 may include a hardware device with data information processing capabilities and the necessary programs required to drive the hardware device to work.

[0048] It should be understood that Figure 1 The number of generation system 130, database 110, and user equipment 120 shown is merely illustrative. Depending on implementation needs, any number of generation system 130, database 110, and user equipment 120 can be included.

[0049] It should be noted that all user data obtained in this manual has been authorized by the user and does not involve user privacy.

[0050] Figure 2 A schematic diagram of the hardware structure of a computing device 200 according to some embodiments of this specification is shown. This computing device 200 can be used as... Figure 1 The generation system 130 is described in some embodiments. When the generation system 130 employs a device cluster, the computing device 200 can be any one of the devices in the generation system 130.

[0051] like Figure 2 As shown, the computing device 200 includes at least one storage medium 230 and at least one processor 220. In some embodiments, the computing device 200 may further include an internal communication bus 210. In some embodiments, the computing device 200 may further include a communication port 250. In some embodiments, the computing device 200 may further include I / O components 260.

[0052] The internal communication bus 210 can connect different system components, including storage medium 230 and processor 220. I / O component 260 supports input / output between computing device 200 and other components.

[0053] Communication port 250 is used for data communication between computing device 200 and the outside world. For example, computing device 200 can connect to a network through communication port 250.

[0054] Storage medium 230 may include a data storage device. The data storage device may be a non-transitory storage medium or a temporary storage medium. For example, the data storage device may include one or more of a disk 232, a read-only storage medium (ROM) 234, or a random access storage medium (RAM) 236. Storage medium 230 also includes at least one instruction set stored in the data storage device. The instruction set is computer program code, which may include programs, routines, objects, components, data structures, procedures, modules, etc., that execute the method for generating instructional videos provided in this specification.

[0055] At least one processor 220 is communicatively connected to at least one storage medium 230 via an internal communication bus 210. The at least one processor 220 is used to execute at least one instruction set. When the system 130 is running, the at least one processor 220 reads at least one instruction set and executes the method for generating instructional videos provided in this specification according to the instructions of the at least one instruction set.

[0056] Processor 220 can execute all the steps included in the method for generating instructional videos. Processor 220 can be in the form of one or more processors. Processor 220 can issue execution instructions. Processor 220 may include one or more hardware processors, such as microcontrollers, microprocessors, reduced instruction set computers (RISC), application-specific integrated circuits (ASICs), application-specific instruction set processors (ASIPs), central processing units (CPUs), graphics processing units (GPUs), physical processing units (PPUs), microcontroller units, digital signal processors (DSPs), field-programmable gate arrays (FPGAs), advanced RISC machines (ARMs), programmable logic devices (PLDs), any circuit or processor capable of performing one or more functions, or any combination thereof.

[0057] For illustrative purposes only, only one processor 220 is shown in the accompanying drawings of the computing device 200. However, it should be noted that the computing device 200 may also include multiple processors. Therefore, the operation and / or method steps disclosed herein may be executed by a single processor or by multiple processors in combination, as described herein. For example, if processor 220 of the computing device 200 in this specification executes steps A and B, it should be understood that steps A and B may also be executed jointly or separately by two different processors 220 (e.g., a first processor executes step A, a second processor executes step B, or the first and second processors jointly execute steps A and B).

[0058] Figure 3A flowchart illustrating a method for generating an instructional video according to an embodiment of this specification is shown; this instructional video generation method P300 can be executed by the generation system 130. Figure 3 As shown, the method P300 provided in this specification may include S310-S390, wherein: S310: Obtain the content of the text unit, which is obtained based on the original teaching text.

[0059] In some embodiments, the content of a text unit may be a portion of the original teaching text, or it may be the entire content of the original teaching text.

[0060] For example, the content of a text unit can be a portion of the original teaching text selected by the user; or, the content of a text unit can be a portion obtained by the generation system after segmenting the original teaching text using a preset parsing method. The specific scope of content included in the text unit and the method of obtaining it can be flexibly adjusted according to the user's needs and are not limited to the examples given above.

[0061] The original teaching text can be a user-uploaded original teaching file received by the system from a specified file path, a database, or via a network interface. The original teaching file can be any format, such as plain text textbook format, textbook image format, textbook portable document format, textbook presentation document format, or audio format.

[0062] When the format of the original teaching file obtained by the generation system is not text, the generation system needs to convert the format of the original teaching file to obtain a plain text version of the original teaching text. The plain text version of the original teaching text contains the logical structure information of the original teaching file before conversion, such as the chapter number and title position of each chapter.

[0063] For example, the conversion method could be: the system generates the original teaching text in plain text format through character recognition; or, a speech-to-text tool can be used to convert the speech format into the original teaching text in plain text format. It should be understood that the specific method for obtaining the original teaching text in text format can be flexibly adjusted according to user needs and is not limited to the methods given in the above embodiments.

[0064] S330: Perform a complexity assessment on the content of the text unit to obtain the complexity assessment result.

[0065] In some embodiments, the complexity evaluation of text unit content by the generation system may include: obtaining the intrinsic complexity of the text unit content, and / or obtaining the extrinsic complexity of the text unit content. Intrinsic complexity characterizes the content complexity of the text unit content, such as the content complexity of the knowledge involved in the text unit content itself. Extrinsic complexity characterizes the complexity of the way the text unit content is expressed.

[0066] For example, different text unit contents may correspond to different complexity evaluation results. For instance, some text unit contents may only include pure formula content without textual description; in this case, the complexity evaluation result will only include intrinsic complexity. For text unit contents that include both textual content and formula content, the complexity evaluation result may include both intrinsic and extrinsic complexity. Alternatively, for some text unit contents that only contain textual description, the complexity evaluation result may only include extrinsic complexity, or it may include both intrinsic and extrinsic complexity. The specific content included in the complexity evaluation result can be flexibly adjusted according to user needs and is not limited to the examples given above.

[0067] In some embodiments, taking the complexity evaluation result as including intrinsic complexity and extrinsic complexity as an example, the generation system can perform weighted fusion of intrinsic complexity and extrinsic complexity based on preset weights to obtain a comprehensive complexity score as the complexity evaluation result of the text unit content.

[0068] S350: Based on the complexity assessment results of the text unit content, determine the target rendering method from multiple preset rendering methods; different complexity assessment results correspond to different preset rendering methods, and different preset rendering methods have different visual display strategies.

[0069] In some embodiments, the generation system may have a pre-defined correspondence between complexity scores and multiple pre-defined rendering methods. After obtaining the complexity score, the generation system can determine the target rendering method corresponding to the current complexity score based on the correspondence.

[0070] Alternatively, the generation system may have a pre-defined correspondence between complexity scores and multiple preset complexity levels. Furthermore, it may also have a pre-defined correspondence between multiple complexity levels and preset rendering methods. After obtaining the complexity score, the generation system can determine the complexity level corresponding to the complexity score based on the correspondence between the complexity score and multiple preset complexity levels, and then determine the target rendering method from the preset rendering methods based on the complexity level.

[0071] For example, the preset complexity levels may include low complexity, medium complexity, and high complexity. Different complexities have corresponding preset complexity score ranges. When the current complexity score falls within the corresponding target complexity score range, the complexity level corresponding to the target complexity score range is determined as the target complexity level for the text unit content. Of course, the preset complexity levels may also include only two complexity levels, or four, five, or even more complexity levels. The number of complexities included in each complexity level can be flexibly adjusted according to user needs and is not limited to the examples given above.

[0072] For example, a visual presentation strategy may include at least one of the following dimensions: animation usage, element display, timing control, and information complexity.

[0073] The dimensions for animation usage can include: the frequency and complexity of animation usage. The dimensions for element display can include: the number and style of elements displayed. The dimensions for time control can include: the total duration of each video unit and the pause time for each video frame. The dimensions for information complexity can include: the complexity of content displayed simultaneously within the same video frame.

[0074] In some embodiments, the complexity evaluation results are divided into multiple levels. The target rendering method corresponding to low complexity has higher animation complexity in the animation usage dimension, more elements displayed in the element display dimension, shorter pause time in the time control dimension, and / or higher information complexity in the information complexity dimension compared to the target rendering method corresponding to high complexity.

[0075] Specifically, when the complexity level of the text unit content is low, meaning the cognitive pressure on the user is low, the corresponding video unit can accommodate more elements, more complex animations, and denser information. Simultaneously, the pause time between each video frame can be shortened to avoid user waiting. Conversely, when the complexity level of the text unit content is high, meaning the cognitive pressure on the user is high, the corresponding video unit can reduce the number of displayed elements, simplify animations, and extend the pause time between each video frame to simplify rendering, reduce unnecessary interference, and lower the risk of user cognitive overload.

[0076] The method described above for determining the target rendering method based on complexity assessment results enables the generation system to flexibly determine the corresponding visual display strategy in the video unit according to the complexity of the text unit content, thus achieving adaptive and accurate matching between the text unit content and the visual presentation format.

[0077] S370: Based on target rendering, it generates target code corresponding to the content of text units, and the target code can be executed by the animation engine.

[0078] In some embodiments, the animation engine can be a mathematical animation engine, and the target code generated by the generation system can be Python code. Based on the target code, the animation engine can be directly driven to execute the corresponding instructions within the target code. Of course, the animation engine can also be other rendering libraries or platforms capable of executing target code, such as 2D animation engines or 3D animation engines. The specific animation engine selection can be flexibly adjusted according to user needs and is not limited to the embodiments given above.

[0079] Taking Python code as an example again, Python code, with its clear logical structure and rich mathematical operation library support, can accurately describe the dynamic changes of complex mathematical relationships, ensuring accurate visualization of professional content such as formula derivation and geometric transformations. Secondly, the Python code generated by the system can directly drive the mathematical animation engine to execute rendering instructions, eliminating the intermediate format conversion step, thus reducing the complexity of the generation system and improving code execution efficiency.

[0080] S390: Calls the animation engine to run the target code and generates video units corresponding to the text unit content.

[0081] For example, still using Python code as the target code, the generation system directly loads the Python code containing graphics instructions, timeline control, and / or mathematical operation logic into the mathematical animation engine. The mathematical animation engine, by parsing parameters such as object creation instructions, attribute change functions, and animation timeline settings in the code, renders the corresponding dynamic graphics frame by frame, converting the code into a video frame sequence, and finally synthesizing the teaching animation video corresponding to the text unit content. This end-to-end code-driven mechanism realizes the automated generation of video units, which not only improves the accuracy of video unit content visualization but also increases the efficiency of video unit generation.

[0082] In summary, the instructional video generation system provided in this manual acquires text unit content based on the original instructional text, evaluates the complexity of the text unit content, and quantifies its complexity. Then, based on the complexity evaluation results, the system determines a target rendering method from multiple preset rendering methods, thereby establishing a mapping relationship between text unit content and visual display strategies. Furthermore, the system generates target code executable by an animation engine based on the target rendering method, then calls the animation engine to run the target code and generate video units, thus achieving an automated conversion from text unit content to video units and enabling rapid generation of instructional videos. In this solution, the instructional video generation system improves teaching effectiveness while automatically generating video units corresponding to text unit content by dynamically linking the complexity of text unit content with the rendering method.

[0083] In some embodiments, before obtaining the text unit content based on the original teaching text, the generation system may first perform format conversion and text cleaning on the original teaching text.

[0084] Figure 4 A flowchart illustrating the process of obtaining text unit content according to an embodiment of this specification is shown, such as... Figure 4 As shown, after the generation system converts the original teaching files into plain text format to obtain the plain text format teaching original text, it also needs to perform text cleaning and standardization on the plain text format teaching original text.

[0085] The text cleaning process includes, but is not limited to, at least one of the following: removing redundant whitespace characters and special symbols, standardizing paragraph formatting, handling line breaks and indentation, and standardizing irregular text formats.

[0086] In some embodiments, the text unit content can also be a portion obtained by the generation system after segmenting the original teaching text based on a preset parsing method, as an example. The generation system can obtain the original teaching text, perform structured parsing on the original teaching text, and segment the original teaching text into structured data containing at least one chapter, wherein the chapter includes: chapter title and chapter content. Subsequently, the generation system obtains the text unit content based on the chapters in the structured data.

[0087] In other words, the text unit content can be structured data of at least one chapter after the original teaching text has been broken down by the generation system. The number of chapters included in the text unit content can be preset by the generation system, for example, retrieving structured data of 1, 2, or 3 chapters each time. Alternatively, it can be preset by the user, who can pre-set the number of chapters of structured data to be retrieved. The specific number of chapters included in the text unit content can be preset by the generation system or specified by the user.

[0088] Continue as Figure 4 As shown, the generation system can call a large language model to analyze the original teaching text to obtain its directory structure or logical structure. Then, based on the directory structure or logical structure, the generation system performs structured parsing of the original teaching text, dividing it into structured data containing at least one chapter.

[0089] In some embodiments, the large language model first determines whether the original instructional text contains a table of contents structure. If the original instructional text contains a table of contents structure, the large language model can directly obtain the table of contents structure within the original instructional text.

[0090] When the original teaching text does not contain a table of contents, the large language model can parse it using its natural language understanding and context awareness capabilities to identify its logical structure, such as heading levels (e.g., first-level headings, second-level headings), chapter divisions, and logical relationships between different parts (e.g., general-to-specific, parallel, or progressive). The large language model then generates the table of contents based on this logical structure.

[0091] Continue as Figure 4 As shown, the generation system performs structured parsing of the original teaching text based on the directory structure, so as to divide the original teaching text into structured data containing at least one chapter, and store the chapter content of each chapter in a structured manner.

[0092] For example, the generation system can locate the corresponding chapter title in the original teaching text based on the directory structure, and segment the text into corresponding chapter content blocks according to the order and hierarchical relationship of the chapter titles. For multi-level headings (such as chapters, sections, and subsections), a tree structure is used for hierarchical division to ensure the accuracy of the hierarchy in the subsequent chapter content segmentation process.

[0093] In some embodiments, the generation system can store the segmented chapter content into a preset database according to a directory structure. The directory structure includes information such as the hierarchical relationship between chapters, chapter titles, and chapter content.

[0094] For example, the generation system can use a tree structure to store the hierarchical relationships between chapters, facilitating subsequent inspection and management. Furthermore, the generation system can also store the mapping relationship between the original teaching text and the segmented chapter content, allowing for backtracking when needed later.

[0095] Figure 5 A flowchart illustrating the process of obtaining text unit content according to another embodiment of this specification is shown, such as... Figure 5 As shown, after obtaining the segmented and structured content of each chapter, the generation system performs data preprocessing on a chapter-by-chapter basis. This preprocessing includes, but is not limited to: standardizing the format of the chapter content, removing redundant punctuation, special characters, and blank paragraphs, and standardizing the annotation of mathematical formulas and technical terms to ensure the accuracy of subsequent large language model processing. Furthermore, when the chapter content includes images or image references, the generation system also needs to provide detailed descriptions of the images using the large language model.

[0096] Furthermore, based on the preprocessed data, the generation system uses a large language model to classify the content of each chapter into explanation stages, resulting in a preset number of explanation stages for each chapter. Then, the generation system identifies a target explanation stage from these preset stages and obtains the text content of that target explanation stage as the text unit content. For example, an explanation stage may include: introduction, core content, summary; or: introduction, core content, examples, summary; or: background, analysis, derivation, examples, etc.

[0097] like Figure 5 As shown, taking the explanation stage, which includes background, analysis, derivation, and examples, as an example, the generation system uses the preprocessed data and a large language model to classify the content of each chapter according to the above four explanation stages.

[0098] Among these methods, a large language model can be used to analyze the chapter content, uncover the historical background of the chapter content, its position in the teaching system of the corresponding subject, and its relevant content in practical applications, thereby generating a knowledge background introduction.

[0099] Furthermore, the large language model can filter through a pre-set database to select representative and intuitive examples corresponding to chapter content, and then modify them appropriately based on the chapter content. For example, when the chapter content is about probability and statistics, the large language model can select examples such as coin tossing and dice rolling from the pre-set database, and describe them mathematically to demonstrate the process of probability and statistics.

[0100] The large language model can also perform step-by-step analysis and parsing based on theorems, algorithms, or formulas within chapter content. For example, it can analyze the origin and derivation of formulas, breaking down the derivation process into multiple steps using logical reasoning and knowledge graphs. In each step, the meaning and function of key steps are explained using natural language. For instance, when deriving the general term formula for an arithmetic sequence, the large language model can start from the first term and common difference, deriving the general term formula through inductive reasoning, and explaining the application of induction in this process.

[0101] The large language model can also generate example problems corresponding to the chapter content. When there are no example problems in the chapter content, the large language model can create corresponding example problems based on the chapter content to generate corresponding problem-solving process instances. Simultaneously, the large language model can perform operations such as parameter adjustments and condition changes on the original example problems to generate a series of variant example problems, thereby increasing the richness of the example problems.

[0102] Continue as Figure 5 As shown, the generation system can perform quality evaluations on a preset number of explanation stages corresponding to a chapter. If the quality evaluation result indicates that the preset number of explanation stages are not satisfactory, the system will modify the preset number of explanation stages until the quality evaluation result indicates that the stage is satisfactory. The quality evaluation includes: text coherence evaluation and knowledge point coverage evaluation.

[0103] For example, the generation system can concatenate the content of a preset number of explanation stages to obtain the concatenated content. Subsequently, the generation system uses a large language model to score the text coherence and knowledge point coverage of the concatenated content, obtaining a text coherence score (Scoherence) and a knowledge point coverage score (Scovery). A comprehensive score for the concatenated content is then determined based on these two scores. S = αScoherence + βScoverage Where α and β are weighting coefficients. If the overall score of the spliced ​​content is less than or equal to the preset score threshold τ, the generation system modifies a preset number of explanation stages until the quality evaluation result indicates that it has passed (the overall score is greater than the preset score threshold), and outputs the content of each explanation stage that has passed the quality evaluation result.

[0104] Through the above process, the generation system can break down the content of each chapter into a preset number of explanation stages, providing a complete content framework for the automated generation of subsequent video units.

[0105] Figure 6 A flowchart illustrating a complexity evaluation process according to an embodiment of this specification is shown, such as... Figure 6As shown, the generation system can obtain the intrinsic complexity of the text unit content, which characterizes the content complexity of the text unit content; and / or, obtain the extrinsic complexity of the text unit content, which characterizes the expression complexity of the text unit content. Furthermore, based on the intrinsic and / or extrinsic complexity, the complexity evaluation result is determined.

[0106] like Figure 6 As shown, taking the explanation stage, which includes background, analysis, derivation, and examples, as an example, the generation system can evaluate the complexity of the content in multiple explanation stages separately, determining the complexity evaluation result for each stage. Then, based on the complexity evaluation result of each explanation stage, the generation system determines the target rendering method corresponding to each stage. That is, the target rendering methods for different explanation stages within the same chapter may be the same or different.

[0107] In some embodiments, the generation system may obtain the intrinsic complexity of the text unit content from at least one of the following: the type and number of mathematical symbols contained in the text unit content; the correspondence between conceptual nouns in the text unit content and preset levels; and the structural complexity of the formulas contained in the text unit content.

[0108] For example, the generation system can pre-define an intrinsic complexity evaluation algorithm, such as defining a quantitative evaluation method for intrinsic complexity through intrinsic complexity calculation code.

[0109] In some embodiments, the calculation of intrinsic complexity includes three dimensions: the type and number of mathematical symbols, the correspondence between conceptual nouns and preset levels, and the structural complexity of formulas. The generation system needs to be able to calculate the complexity of mathematical symbols based on symbol complexity calculation code, the complexity of conceptual levels based on concept level calculation code, and the structural complexity of formulas based on formula structure complexity calculation code. Then, based on a comprehensive assessment of these three types of complexity, the intrinsic complexity of the text unit content is determined, achieving a multi-dimensional and quantifiable intrinsic complexity evaluation.

[0110] The generation system categorizes and quantifies mathematical symbols using a pre-defined weighting system within the symbolic complexity calculation code. Taking a four-level weighting system as an example, basic operators correspond to weight 1, algebraic symbols to weight 2, advanced operators to weight 3, and specialized mathematical symbols to weight 4. The generation system uses regular expressions to identify the frequency of each type of mathematical symbol in the text unit content, calculates the mathematical symbol complexity through weighted summation, and finally maps the result to a pre-defined range, such as 0-10, through normalization. Of course, it can also be 0-1, 0-100, or any other range, which can be flexibly adjusted according to user needs. This method allows the generation system to quantify the depth of use of mathematical symbols contained in the text unit content, thereby distinguishing the differences in mathematical symbol complexity between basic arithmetic expressions and advanced mathematical formulas.

[0111] Secondly, the generation system can perform hierarchical analysis of the conceptual nouns included in the text unit content using a preset concept hierarchy system in the concept hierarchy complexity calculation code. Taking a five-level concept hierarchy system as an example, the generation system can use a keyword matching algorithm to traverse the text unit content through the concept hierarchy complexity calculation code, identify the highest-level conceptual nouns appearing in the text unit content, and determine that the highest level is the target concept level corresponding to the text unit content. To make the target concept level more obvious, the generation system can amplify the target concept level based on a preset coefficient (greater than 1).

[0112] Furthermore, the generation system can quantify the formula complexity within text units using formula structure complexity calculation code. This code detects the maximum nesting level (reflecting formula structure complexity) within formulas in the text unit by checking the bracket depth and calculating the operator density within the formulas. Then, the code employs a weighted combination strategy to calculate formula complexity. Similarly, the generation system needs to normalize the formula complexity to a preset range.

[0113] In some embodiments, the generation system may obtain the external complexity of the text unit content based on at least one of the following: the text density contained in the text unit content; the term density contained in the text unit content; and the information organization complexity of the text unit content.

[0114] For example, the generation system can pre-define the external complexity evaluation algorithm, such as by defining the quantitative evaluation method of external complexity through external complexity calculation code.

[0115] In some embodiments, the calculation of external complexity includes, for example, the calculation of the three dimensions of text density, term density, and information organization complexity of the text unit content. The generation system can determine the text density based on the text density calculation code, determine the term density based on the term density calculation code, calculate the information organization complexity based on the information organization complexity calculation code, and then determine the external complexity of the text unit content based on the calculation results of the above three dimensions.

[0116] In some embodiments, the generation system can calculate the average sentence length in the text unit content based on text density calculation code to quantify the text density of the text unit content. For example, the generation system can use text density calculation code to segment the text in the text unit content using periods as the dividing criteria, thereby dividing the text unit content into an array of sentences, and counting the total number of characters and sentences in the sentence array. Subsequently, the generation system determines the average number of characters corresponding to each sentence based on the number of sentences and the total number of characters. Furthermore, the generation system uses the ratio between the average number of characters and a preset character count threshold as the text density calculation result. To improve the stability of the text density calculation result, a preset maximum ratio can be set; if the current ratio exceeds the preset maximum ratio, the preset maximum ratio is determined as the text density calculation result.

[0117] For example, let's take a preset character count threshold of 20 and a preset maximum ratio of 5. If the average number of characters per sentence in text unit 1 is 30, then the ratio between the average number of characters in text unit 1 and the preset character count threshold is 1.5 (30 / 20=1.5). Since 1.5 < 5, the text density of text unit 1 is determined to be 1.5. Similarly, if the average number of characters per sentence in text unit 2 is 150, then the ratio between the average number of characters in text unit 2 and the preset character count threshold is 7.5 (150 / 20=7.5). Since 7.5 > 5 (the preset maximum ratio), the text density of text unit 2 is determined to be the preset maximum ratio of 5.

[0118] Secondly, the term density calculation code includes a pre-defined term dictionary, allowing the generation system to determine the frequency of terms appearing in text units. Furthermore, the system quantifies the term density value within a text unit by calculating the ratio of the number of terms to the total number of words in that unit. Since term density values ​​are generally small, directly using this value would result in a relatively small weighting in the calculation of external complexity. Therefore, to ensure comparability between term density and other dimensions (text density, information organization complexity), the generation system amplifies the term density value using a pre-defined amplification factor.

[0119] Furthermore, to avoid extreme cases and improve the stability of the generation system, the term density is also constrained by the maximum density value. When the amplified term density value is greater than the maximum density value, the maximum density value is determined to be the term density value corresponding to the content of the current text unit.

[0120] Furthermore, the generation system uses regular expressions to identify structured elements (numerical numbers, Chinese numbers, bullet points, etc. used to represent lists or steps) contained in text units based on the information organization complexity calculation code, and determines the number of paragraphs in the text unit content. Then, the generation system uses preset weights to perform a weighted summation of the structured elements and the number of paragraphs to calculate the organizational structure complexity.

[0121] The aforementioned method of calculating intrinsic and extrinsic complexity based on code improves the determinism of the computation algorithm and the consistency of complexity evaluation results for different text unit contents. During the calculation process, the evaluation dimensions can be flexibly adjusted according to the needs of the current scenario; for example, some evaluation dimensions can be added or deleted.

[0122] In some embodiments, after obtaining the complexity evaluation result corresponding to the text unit content and determining the target rendering method based on the complexity evaluation result, the generation system can obtain the rendering prompt words corresponding to the target rendering method, input the text unit content and the rendering prompt words into the large language model, and obtain the target code generated by the large language model.

[0123] For example, using mathematical teaching texts as examples, with complexity assessment results including high complexity, medium complexity, and low complexity, and with Manim as the animation engine and Manim LaTeX code as the target code, this will be explained.

[0124] For rendering methods corresponding to low complexity, the generation system can preset the corresponding rendering hints as follows: You are a professional math teaching video production expert who needs to generate Manim LaTeX code for math content with low complexity.

[0125] **Target Audience:** Learners with a basic understanding of this mathematical concept and the ability to process complex visual information. **Generation Requirements**: 1. **Complex Animation Sequences:** Use rich animation effects to make videos more vivid and interesting. - Multiple animation types can be combined (Write animation, FadeIn animation, Transform animation, Create animation, etc.) - Complex path animations and deformation effects - 3D effects and perspective changes (where appropriate) 2. **Multi-element parallel display:** - Simultaneously display multiple related concepts or formulas - Parallel animation sequences, where multiple objects move simultaneously. - Split-screen or multi-region simultaneous display 3. **Enrich visual effects:** - Color gradient and highlight effect - Particle effects and special transitions - Dynamic backgrounds and decorative elements 4. **Time Control**: - Base display time: T_display = t_base - The animation is fast-paced and the transitions are smooth. **Code structure requirements:** - Use Scene class inheritance - Includes a rich collection of animations - Use animation groups and successions appropriately. - Add appropriate sound cue points (#AUDIO_CUE) **Input content**: {Text cell content} Please generate complete Manim Python code and ensure that the code can be run directly.

[0126] For rendering methods corresponding to medium complexity, the generation system can preset the corresponding rendering hints as follows: You are a professional math instruction video production expert who needs to generate Manim LaTeX code for moderately complex math content.

[0127] **Target Audience:** Learners who are currently learning this mathematical concept and need to understand it step by step. **Generation Requirements**: 1. **Step-by-step, progressive demonstration**: - Break down complex concepts into multiple simple steps - Each step is clear and independent, making it easy to understand. - Logical progression from simple to complex 2. **Appropriate animation effects:** - Use basic animations (Write animation, FadeIn animation, Transform animation) - Avoid overly complex visual effects - Emphasize key transformation processes 3. **Logical sequence display**: - Elements appear sequentially according to the teaching logic - Each new element establishes a connection with the preceding content. - Use arrows, lines, etc. to guide the eye. 4. **Time Control**: - Display time: T_display = 1.2 * t_base - Give learners ample time to think - Appropriate pauses between steps **Code structure requirements:** - Clear step division (step 1, step 2, etc.) - Use the Wait() function to control the pace - Appropriate highlighting and emphasis effects - Add step markers and progress indicators **Input content**: {Text cell content} Please generate well-structured and clearly defined Manim Python code to ensure learners can keep up with the teaching pace.

[0128] For rendering methods that correspond to high complexity, the generation system can preset the corresponding rendering hints as follows: You are a professional math teaching video production expert who needs to generate Manim LaTeX code for complex mathematical content.

[0129] **Target Audience:** Learners who need ample time to understand and derive complex mathematical concepts. **Generation Requirements**: 1. **Primarily static charts:** - Centered on clear static displays - Minimize animation effects to avoid distractions - Emphasis on using simple animations such as Write and FadeIn. 2. **Single Focus Emphasis**: - Highlight only one key concept or formula each time. - Use borders, highlights, and other techniques to emphasize key points. - Avoid multiple elements competing for attention simultaneously 3. **Minimize visual disturbances:** - Simple background and layout - A unified color scheme (primarily using black, white, and gray). - Remove unnecessary decorative elements 4. **Time Control**: - Base display time: T_display = 1.5 * t_base - Critical pause time: t_pause = 2 * t_pause_base - Allow ample time for reading and comprehension. **Code structure requirements:** - Extensive use of the Wait() function increases pauses - Highlight key content using SurroundingRectangle - The formula derivation process is shown line by line. - Add sufficient time markers for thinking (#THINKING_TIME) **Special Requirements**: - For complex formulas, a step-by-step construction approach is adopted. - Key conclusions need to be highlighted separately. - Provide sufficient visual white space **Input content**: {Text cell content} Please generate concise and clear Manim Python code that highlights the key points, ensuring learners have ample time to understand complex concepts.

[0130] Continue as Figure 6 As shown, the generation system obtains the video units corresponding to each explanation stage in the chapter and splices them together to obtain the teaching video corresponding to the chapter.

[0131] For example, when splicing multiple explanation stages, the generation system can add transition effects between adjacent explanation stages, such as fade-in / fade-out, dissolve, and slide transition, so that the overall transition of the teaching video corresponding to the chapter is natural and smooth.

[0132] In some embodiments, to facilitate subsequent retrieval and management, the generation system can uniformly archive the teaching videos corresponding to the chapters, the text content of multiple explanation stages, the code used to generate the videos, and related information.

[0133] To improve the quality of instructional videos, the generation system can also input the instructional video and preset scoring prompts into a large language model to obtain the evaluation results output by the large language model. If the evaluation result indicates approval, the instructional video is output to the user. If the evaluation result indicates disapproval, the evaluation result is input into the large language model, allowing the model to regenerate the target code based on the evaluation result, until the evaluation result for the instructional video corresponding to the modified target code indicates approval.

[0134] For example, the evaluation of instructional videos can cover multiple dimensions, such as content accuracy and teaching standardization. Content accuracy checks whether the knowledge points and formula derivations explained in the instructional video are correct and consistent with the textbook content. For instance, the generation system uses a large language model to perform semantic analysis on the text and formulas in the instructional video, comparing it with a pre-set mathematical knowledge graph to determine the accuracy of the content.

[0135] Secondly, the system checks whether the teaching videos conform to educational and teaching standards. For example, it checks whether the teaching objectives in the videos are clear and whether the key points or difficulties are highlighted. These educational and teaching standards are, for example, pre-set and related to the application scenario and content of the teaching videos; different application scenarios or different teaching content may correspond to different educational and teaching standards.

[0136] In some embodiments, the evaluation results may include a score and evaluation content. The score is calculated by the large language model based on the review results for each review dimension and preset weights. The evaluation content includes the specific score for each review dimension and the problems identified.

[0137] Furthermore, after generating the instructional video, the system can send the video, along with corresponding evaluation results and chapter content, to the user for review. For example, the user can view the instructional video, evaluation results, and related data. During the review process, the user can provide feedback on the instructional video, such as adding knowledge points, adjusting the order of explanations, or correcting visual or audio content. The system records these feedbacks and inputs the feedback into a large language model, obtaining the modified target code output by the model. The system then calls the animation engine to run the modified target code, updating the video units corresponding to the text unit content until the modified instructional video meets the user's requirements.

[0138] This specification, in another aspect, provides a computer-readable non-transitory storage medium storing at least one set of instructions executable for generating instructional videos. When the at least one set of instructions is executed by a processor, it instructs the processor to implement the steps of the instructional video generation method P300 of this specification. In some possible embodiments, various aspects of this specification can also be implemented as a program product comprising program code. When the program product is run on the generation system 130, the program code causes the generation system 130 to perform the steps of the instructional video generation method P300 described in this specification. The program product for implementing the above method may employ a portable compact disc read-only memory (CD-ROM) containing program code and may run on the generation system 130. However, the program product of this specification is not limited thereto. In this specification, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system. The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Computer-readable storage media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A readable storage medium may also be any readable medium other than a readable storage medium that can send, propagate, or transmit programs for use by or in connection with an instruction execution system, apparatus, or device. Program code contained on a readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination of the foregoing. Program code for performing the operations described herein may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar programming languages.

[0139] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0140] In summary, after reading this detailed disclosure, those skilled in the art will understand that the foregoing detailed disclosure may be presented by way of example only and may not be restrictive. Although not explicitly stated herein, those skilled in the art will understand that this specification requires various reasonable changes, improvements, and modifications to the embodiments. These changes, improvements, and modifications are intended to be made by this specification and are within the spirit and scope of the exemplary embodiments described herein.

[0141] Furthermore, certain terms in this specification have been used to describe embodiments of this specification. For example, "an embodiment," "an embodiment," and / or "some embodiments" mean that a particular feature, structure, or characteristic described in connection with that embodiment may be included in at least one embodiment of this specification. Therefore, it is to be emphasized and understood that two or more references to "an embodiment" or "an embodiment" or "alternative embodiment" in various parts of this specification do not necessarily refer to the same embodiment. Moreover, specific features, structures, or characteristics may be suitably combined in one or more embodiments of this specification.

[0142] It should be understood that in the foregoing description of the embodiments in this specification, various features are combined in a single embodiment, drawing, or description for the purpose of simplifying the description and to aid in understanding a feature. However, this does not mean that the combination of these features is necessary, and those skilled in the art, upon reading this specification, may readily identify some of the devices as separate embodiments. That is, the embodiments in this specification can also be understood as an integration of multiple secondary embodiments. And the content of each secondary embodiment is valid even if it contains fewer than all the features of a single foregoing disclosed embodiment.

[0143] Every patent, patent application, publication of a patent application, and other documents, such as articles, books, specifications, publications, documents, and literature (excluding any related historical examination documents), cited in this disclosure are incorporated herein by reference for all purposes relating to this disclosure, such as in the specification and claims of this disclosure. However, in the event of any inconsistency or conflict between the descriptions, definitions, and / or terms in the foregoing documents and the descriptions, definitions, and / or terms used in this disclosure, the descriptions, definitions, and / or terms used in this disclosure shall prevail.

[0144] Finally, it should be understood that the embodiments disclosed herein are illustrative of the principles of the embodiments described in this specification. Other modified embodiments are also within the scope of this specification. Therefore, the embodiments disclosed in this specification are merely examples and not limitations. Those skilled in the art can implement the applications described in this specification using alternative configurations based on the embodiments in this specification. Therefore, the embodiments in this specification are not limited to the embodiments precisely described in the applications.

Claims

1. A method for generating instructional videos, wherein, The method includes: The content of the text unit is obtained based on the original teaching text; The complexity of the text unit content is evaluated to obtain the complexity evaluation result; Based on the complexity assessment results of the text unit content, a target rendering method is determined from multiple preset rendering methods; wherein, different complexity assessment results correspond to different preset rendering methods, and different preset rendering methods have different visual display strategies. Based on the target rendering method, target code corresponding to the content of the text unit is generated, and the target code can be executed by the animation engine; and The animation engine is invoked to run the target code, generating video units corresponding to the text unit content.

2. The method according to claim 1, wherein, The visual presentation strategy includes at least one of the following dimensions: animation usage dimension, element display dimension, time control dimension, and information complexity dimension.

3. The method according to claim 2, wherein, The complexity evaluation results are divided into multiple levels. The target rendering method corresponding to low complexity has higher animation complexity in the animation usage dimension, more elements displayed in the element display dimension, shorter pause time in the time control dimension, and / or higher information complexity in the information complexity dimension compared to the target rendering method corresponding to high complexity.

4. The method according to claim 1, wherein, The process of evaluating the complexity of the text unit content to obtain the complexity evaluation result includes: Obtain the intrinsic complexity of the text unit content, which characterizes the content complexity of the text unit content; and / or, obtain the extrinsic complexity of the text unit content, which characterizes the expression complexity of the text unit content; and The complexity assessment result is determined based on the intrinsic complexity and / or the extrinsic complexity.

5. The method according to claim 4, wherein, The inherent complexity of obtaining the content of the text unit includes: The inherent complexity of obtaining the content of the text unit is based on at least one of the following: The types and number of mathematical symbols contained in the text unit content; The correspondence between conceptual nouns and preset levels in the text unit content; The structural complexity of the formulas contained in the text unit content.

6. The method according to claim 4, wherein, The external complexity of obtaining the content of the text unit includes: The external complexity of the text unit content is obtained based on at least one of the following: The text density contained in the text unit content; The terminology density contained in the text unit content; The complexity of information organization within the text unit content.

7. The method according to claim 1, wherein, The acquisition of text unit content includes: Obtain the original teaching text; The original teaching text is subjected to structured parsing, dividing it into structured data containing at least one chapter, wherein the chapter includes: chapter title and chapter content, and Based on the chapters in the structured data, the content of the text unit is obtained.

8. The method according to claim 7, wherein, The step of performing structured parsing on the original teaching text, dividing it into structured data containing at least one chapter, includes: The original teaching text is analyzed using a large language model to obtain its directory structure or logical structure; and Based on the directory structure or logical structure, the original teaching text is structured and parsed to divide it into structured data containing at least one chapter.

9. The method according to claim 7, wherein, The step of obtaining text unit content based on the chapters in the structured data includes: The chapter content is categorized into explanation stages to obtain a preset number of explanation stages corresponding to each chapter; and Among the preset number of explanation stages, a target explanation stage is determined, and the text content of the target explanation stage is obtained as the text unit content.

10. The method according to claim 9, wherein, The method further includes: A quality evaluation is performed on a preset number of explanation stages corresponding to the chapter. If the quality evaluation result indicates that the preset number of explanation stages are not satisfactory, the process is modified until the quality evaluation result indicates that the process is satisfactory. The quality evaluation includes: text coherence evaluation and knowledge point coverage evaluation.

11. The method according to claim 10, wherein, The method further includes: The video units corresponding to each explanation stage in the chapter are obtained and spliced ​​together to obtain the teaching video corresponding to the chapter.

12. The method according to claim 11, wherein, The method further includes: The teaching video and preset scoring guide words are input into the large language model to obtain the evaluation results output by the large language model; When the evaluation result indicates that the evaluation is passed, the instructional video will be output to the user; and When the evaluation result indicates failure, the evaluation result is input into the large language model so that the large language model can regenerate the target code based on the evaluation result until the evaluation result of the teaching video corresponding to the modified target code indicates success.

13. The method according to claim 1, wherein, The step of generating target code corresponding to the content of the text unit based on the target rendering method includes: Obtain the rendering prompt text corresponding to the target rendering method; and The text unit content and the rendering prompt words are input into the large language model, and the target code generated by the large language model is obtained.

14. The method according to claim 13, wherein, The method further includes: Obtain user modification guidance information for the video unit; The modification guidance information is input into the large language model to obtain the modified target code output by the large language model; and The animation engine is invoked to run the modified target code, updating the video unit corresponding to the text unit content.

15. A system for generating instructional videos, comprising: At least one storage medium storing at least one instruction set for generating instructional videos; as well as At least one processor is communicatively connected to the at least one storage medium, wherein the at least one processor reads the at least one instruction set during operation and executes the method of any one of claims 1-14 according to the instructions of the at least one instruction set.

Images