Method and device for generating mathematical physics explanation videos based on multi-agent collaboration

Abstract

The present application relates to the field of computer technology, and in particular to a method and device for generating a mathematical physics explanation video based on multi-agent collaboration, wherein the method comprises: obtaining actual task requirement information of the mathematical physics explanation video; based on the actual task requirement information, controlling the script master agent to generate an animation frame script, so that the animator agent generates an animation video corresponding to the animation frame script, and generates synchronized subtitles and explanation audio corresponding to the animation video through the explainer agent; synchronously integrating the animation frame script, the animation video, the synchronized subtitles and the explanation audio to construct an initial explanation video, and performing a picture analysis on the initial explanation video through the quality inspector agent to obtain a picture analysis result, so as to correct the initial explanation video and generate a final mathematical physics explanation video. Thus, the existing teaching video production technology solves the problems of poor content accuracy and visual expression effect, high video production cost, and difficulty in meeting actual teaching needs.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method and device for generating a mathematical physics explanation video based on multi-agent collaboration. Background Art

[0002] Mathematical physics explanation videos are animated videos used for teaching. They can dynamically display physical processes and mathematical formulas, vividly and concretely explain various concepts and principles in mathematics and physics, and make complex theories and formulas more intuitive and easy to understand.

[0003] In the traditional video production process, animation and editing work mainly rely on experienced professionals to operate professional video production software. The threshold for using such tools is high, and new users need to invest a lot of time and energy to become proficient in operating them. In addition, the production of high-quality videos requires extremely high computing resources, which significantly increases production costs. When a completed video needs to be modified or updated, complex editing and rendering must be performed again, which limits the flexibility of the production process and the iteration speed of the project.

[0004] The generation technology based on the video big model can generate videos directly according to the user's text description. The technology parses the text, identifies the key scene description, and converts it into pixel points in the video to generate coherent dynamic images and scenes. Although this technology significantly improves the efficiency of video production and allows non-professional users to quickly generate videos, there are still some problems in practical applications. For example, it is difficult to accurately control the generation effect of complex geometric shapes and mathematical formulas, resulting in the video failing to achieve the expected accuracy; it is unable to generate real physical processes and lacks scientific accuracy; it is highly dependent on user input and requires users to provide specific and accurate descriptions; the technology integration is complex and costly, requiring powerful computing resources and complex algorithm support, increasing hardware and software costs, and prolonging the video generation time, especially when processing long videos or highly complex content.

[0005] The limitations of the above technologies are particularly obvious when making teaching videos in mathematics and physics. In addition, fields such as teaching video production require extremely high content accuracy and clear visual expression, but current video production technology often cannot fully meet strict content and visualization standards, which greatly affects the teaching quality and wide application of videos.

[0006] To sum up, the existing teaching video production technology has poor content accuracy and visual expression effects, high video production costs, and is difficult to meet actual teaching needs, which needs to be solved urgently. Summary of the invention

[0007] The present application provides a method and device for generating mathematical physics explanation videos based on multi-agent collaboration, in order to solve the problems of existing teaching video production technologies, such as poor content accuracy and visual expression effects, high video production costs, and difficulty in meeting actual teaching needs.

[0008] The first aspect of the present application is a method for generating a mathematical physics explanation video based on multi-agent collaboration, wherein the multi-agent includes a scriptwriter agent, an animator agent, an explainer agent and a quality inspector agent, wherein the method for generating a mathematical physics explanation video based on multi-agent collaboration includes the following steps: obtaining actual task requirement information of the mathematical physics explanation video; based on the actual task requirement information, controlling a preset scriptwriter agent to generate an animation frame script, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generates synchronized subtitles and explanation audio corresponding to the animation video through a preset explainer agent; synchronously integrating the animation frame script, the animation video, the synchronized subtitles and the explanation audio to construct an initial explanation video corresponding to the mathematical physics explanation video, and performing picture analysis on the initial explanation video through a preset quality inspector agent to obtain a picture analysis result, and correcting the initial explanation video according to the picture analysis result to generate a final mathematical physics explanation video.

[0009] Optionally, in one embodiment of the present application, based on the actual task requirement information, controlling a preset scriptwriter agent to generate an animation frame script includes: performing a preprocessing operation on the actual task requirement information to generate standard task requirement information corresponding to the actual task requirement information, wherein the preprocessing operation includes at least one of text cleaning processing, word segmentation processing and grammatical parsing processing; using a preset embedding model to convert the standard task requirement information into a task requirement vector, and analyzing the vector similarity between the task requirement vector and a preset requirement database; determining whether the similarity is less than a preset similarity threshold, and if the similarity is less than the preset similarity threshold, storing the actual task requirement information; if the similarity analysis result is greater than or equal to the preset similarity threshold, obtaining a target task requirement vector corresponding to the task requirement vector in the requirement database, and generating the animation frame script according to the target task requirement vector through the scriptwriter agent.

[0010] Optionally, in one embodiment of the present application, based on the actual task requirement information, controlling a preset scriptwriter agent to generate an animation frame script, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generating synchronized subtitles and commentary audio corresponding to the animation video through a preset commentator agent, including: performing syntax, content and logic self-inspection operations on the animation frame script through the scriptwriter agent to obtain a self-inspection result, and when the self-inspection result meets the preset animation frame script self-inspection requirement, allowing the quality inspector agent to perform syntax, content and logic preliminary review operations on the animation frame script to generate a preliminary review result; judging whether the animation frame script meets a first preset creativity and quality standard based on the preliminary review result, if the animation frame script does not meet the first preset creativity and quality standard, revising the animation frame script; when the animation frame script meets the first preset creativity and quality standard, parsing the animation frame script through the animator agent to obtain a parsing result. result; generating an animation automation execution script corresponding to the animation frame script according to the parsing result, and performing a grammatical and logical review operation on the animation automation execution script by the quality inspector agent to obtain a review result corresponding to the animation automation execution script; when the review result meets the preset grammatical and logical standards, the animator agent calls a preset video construction tool to execute the animation automation execution script to generate multiple animation frames corresponding to the animation automation execution script; based on the preset arrangement requirements, the multiple animation frames are arranged, and the arranged multiple animation frames are compressed and formatted according to the preset frame rate and resolution by the target codec to generate the animation video; the quality inspector agent determines whether the animation video meets the second preset creativity and quality standards, wherein if the animation video does not meet the second preset creativity and quality standards, the animation video is corrected, otherwise the narrator agent generates the corresponding synchronized subtitles and narration audio for the animation video.

[0011] Optionally, in one embodiment of the present application, the generation of synchronized subtitles and explanation audio corresponding to the animation video by a preset narrator agent includes: generating corresponding synchronized subtitles according to the animation video by the narrator agent, and using the quality inspector agent to perform a subtitle review operation on the synchronized subtitles to obtain a subtitle review result; judging whether the synchronized subtitles meet a first preset quality standard according to the subtitle review result, and if the synchronized subtitles do not meet the first preset quality standard, correcting the synchronized subtitles; if the synchronized subtitles meet the first preset quality standard, inputting the synchronized subtitles into a preset explanation audio generation tool by the narrator agent to generate the explanation audio corresponding to the animation video, and performing an audio review operation on the explanation audio by the quality inspector agent to obtain an audio review result; judging whether the audio review result meets a second preset quality standard according to the audio review result, and if the audio review result does not meet the second preset quality standard, correcting the explanation audio

[0012] Optionally, in one embodiment of the present application, the initial explanation video is analyzed by a preset quality inspector agent to obtain a picture analysis result, and the initial explanation video is corrected according to the picture analysis result to generate a final mathematics and physics explanation video, including: decomposing the initial explanation video into multiple video frames, and performing video frame quality assessment on each of the multiple video frames through the spatial perception enhancement strategy of the quality inspector agent to obtain a quality assessment result corresponding to each video frame; judging whether the initial explanation video meets the preset explanation video quality standard according to the quality assessment result, if the initial explanation video meets the explanation video quality standard, generating the final mathematics and physics explanation video, otherwise correcting the animation frame script, the animation video, the synchronized subtitles and the explanation audio, and re-synchronizing and integrating the revised animation frame script, animation video, synchronized subtitles and explanation audio and performing video frame quality assessment operations to generate the corresponding final mathematics and physics explanation video.

[0013] The second aspect of the present application provides a device for generating a mathematical physics explanation video based on multi-agent collaboration, wherein the multi-agents include a scriptwriter agent, an animator agent, an explainer agent and a quality inspector agent, wherein the device for generating a mathematical physics explanation video based on multi-agent collaboration includes: an acquisition module, used to acquire actual task requirement information of the mathematical physics explanation video; a control module, used to control a preset scriptwriter agent to generate an animation frame script based on the actual task requirement information, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generates synchronized subtitles and explanation audio corresponding to the animation video through a preset explainer agent; an explanation video generation module, used to synchronously integrate the animation frame script, the animation video, the synchronized subtitles and the explanation audio to construct an initial explanation video corresponding to the mathematical physics explanation video, and perform picture analysis on the initial explanation video through a preset quality inspector agent to obtain a picture analysis result, and correct the initial explanation video according to the picture analysis result to generate a final mathematical physics explanation video.

[0014] Optionally, in one embodiment of the present application, the control module includes: a preprocessing unit, used to perform preprocessing operations on the actual task requirement information to generate standard task requirement information corresponding to the actual task requirement information, wherein the preprocessing operation includes at least one of text cleaning processing, word segmentation processing and grammatical parsing processing; a conversion unit, used to convert the standard task requirement information into a task requirement vector using a preset embedding model, and analyze the vector similarity between the task requirement vector and a preset requirement database; a first judgment unit, used to judge whether the similarity is less than a preset similarity threshold, and if the similarity is less than the preset similarity threshold, store the actual task requirement information; a comparison unit, used to obtain a target task requirement vector corresponding to the task requirement vector in the requirement database if the similarity analysis result is greater than or equal to the preset similarity threshold, and generate the animation frame script according to the target task requirement vector through the scriptwriter agent.

[0015] Optionally, in one embodiment of the present application, the control module further includes: a self-checking unit, which is used to perform grammar, content and logic self-checking operations on the animation frame script through the scriptmaster agent to obtain a self-checking result, and when the self-checking result meets the preset animation frame script self-checking requirements, the quality inspector agent performs grammar, content and logic preliminary review operations on the animation frame script to generate a preliminary review result; a second judgment unit, which is used to judge whether the animation frame script meets the first preset creativity and quality standards according to the preliminary review results, and if the animation frame script does not meet the first preset creativity and quality standards, the animation frame script is corrected; a parsing unit, which is used to parse the animation frame script through the animator agent to obtain a parsing result when the animation frame script meets the first preset creativity and quality standards; a first review unit, which is used to generate an animation automation execution script corresponding to the animation frame script according to the parsing result, and the quality inspector agent performs a preliminary review operation on the animation frame script according to the preliminary review results. The intelligent agent performs syntax and logic review operations on the animation automation execution script to obtain a review result corresponding to the animation automation execution script; an execution unit is used to call a preset video construction tool through the animator agent to execute the animation automation execution script when the review result meets the preset syntax and logic standards to generate multiple animation frames corresponding to the animation automation execution script; an arrangement unit is used to arrange the multiple animation frames based on preset arrangement requirements, and compress and format the arranged multiple animation frames according to a preset frame rate and resolution through a target codec to generate the animation video; a third judgment unit is used to judge whether the animation video meets a second preset creativity and quality standard through the quality inspector agent, wherein if the animation video does not meet the second preset creativity and quality standard, the animation video is corrected, otherwise the corresponding synchronized subtitles and commentary audio of the animation video are generated through the commentator agent.

[0016] Optionally, in one embodiment of the present application, the control module also includes: a second review unit, which is used to generate corresponding synchronized subtitles according to the animation video through the narrator agent, and use the quality inspector agent to perform a subtitle review operation on the synchronized subtitles to obtain a subtitle review result; a fourth judgment unit, which is used to judge whether the synchronized subtitles meet a first preset quality standard according to the subtitle review result, and if the synchronized subtitles do not meet the first preset quality standard, the synchronized subtitles are corrected; a third review unit, which is used to input the synchronized subtitles into a preset explanation audio generation tool through the narrator agent to generate an explanation audio corresponding to the animation video when the synchronized subtitles meet the first preset quality standard, and perform an audio review operation on the explanation audio through the quality inspector agent to obtain an audio review result; a fifth judgment unit, which is used to judge whether the audio review result meets the second preset quality standard according to the audio review result, and if the audio review result does not meet the second preset quality standard, the explanation audio is corrected.

[0017] Optionally, in one embodiment of the present application, the explanation video generation module includes: a decomposition unit, used to decompose the initial explanation video into multiple video frames, and perform video frame quality assessment on each of the multiple video frames through the spatial perception enhancement strategy of the quality inspector intelligent agent to obtain a quality assessment result corresponding to each video frame; a correction unit, used to determine whether the initial explanation video meets a preset explanation video quality standard based on the quality assessment result, if the initial explanation video meets the explanation video quality standard, then generate the final mathematics and physics explanation video, otherwise correct the animation frame script, the animation video, the synchronized subtitles and the explanation audio, and re-synchronize and integrate the revised animation frame script, animation video, synchronized subtitles and explanation audio and perform video frame quality assessment operations to generate the corresponding final mathematics and physics explanation video.

[0018] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for generating mathematical physics explanation videos based on multi-agent collaboration as described in the above embodiment.

[0019] The fourth aspect of the present application provides a computer-readable storage medium, which stores a computer program. When the program is executed by a processor, it implements the above-mentioned method for generating mathematical physics explanation videos based on multi-agent collaboration.

[0020] The fifth aspect of the present application provides a computer program product, including a computer program, which is executed to implement the above-mentioned method for generating mathematical physics explanation videos based on multi-agent collaboration.

[0021] Therefore, the embodiments of the present application have the following beneficial effects:

[0022] The embodiments of the present application can obtain the actual task requirement information of the mathematical physics explanation video; based on the actual task requirement information, control the preset scripter intelligent agent to generate the animation frame script, so that the preset animator intelligent agent generates the animation video corresponding to the animation frame script, and generates the synchronous subtitles and explanation audio corresponding to the animation video through the preset explainer intelligent agent; the animation frame script, animation video, synchronous subtitles and explanation audio are synchronously integrated to construct the initial explanation video corresponding to the mathematical physics explanation video, and the initial explanation video is analyzed by the preset quality inspector intelligent agent to obtain the picture analysis result, and the initial explanation video is corrected according to the picture analysis result to generate the final mathematical physics explanation video. The present application improves the generation efficiency and content quality of mathematical physics explanation videos through intelligent technology, and can provide high-quality video teaching content for the education field. Thus, the existing teaching video production technology solves the problems of poor content accuracy and visual expression effect, high video production cost, and difficulty in meeting actual teaching needs.

[0023] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0025] Figure 1 A flowchart of a method for generating a mathematical physics explanation video based on multi-agent collaboration according to an embodiment of the present application;

[0026] Figure 2 A schematic diagram of a flow chart for generating an animation frame script according to user requirements provided in one embodiment of the present application;

[0027] Figure 3 A schematic diagram of an execution logic for multiple agents to generate synchronized subtitles and narration audio according to an embodiment of the present application;

[0028] Figure 4 (a) is a flow chart of generating an animation frame script provided by an embodiment of the present application;

[0029] Figure 4(b) is a schematic diagram of a process of synthesizing an animation video by automatically executing a script in an animation, provided by an embodiment of the present application;

[0030] Figure 4 (c) is a flow chart of generating synchronized subtitles and audio narration provided by an embodiment of the present application;

[0031] Figure 5 A schematic diagram of a process for constructing an initial explanation video provided for an embodiment of the present application;

[0032] Figure 6 A schematic diagram of a process for generating a high-quality mathematical physics explanation video using spatial perception enhancement technology provided in one embodiment of the present application;

[0033] Figure 7 A schematic diagram of the execution logic of a method for generating a mathematical physics explanation video based on multi-agent collaboration provided for one embodiment of the present application;

[0034] Figure 8 A schematic diagram of an application example effect provided for an embodiment of the present application;

[0035] Fig. 9 A schematic diagram of the logical architecture of a mathematical physics explanation video generation system based on multi-agent collaboration provided for one embodiment of the present application;

[0036] Fig.10 A schematic diagram of user-system role interaction of a mathematical physics explanation video generation system based on multi-agent collaboration provided for one embodiment of the present application;

[0037] Fig.11 A schematic diagram of a spatial perception enhancement technology workflow of a mathematical physics explanation video generation system based on multi-agent collaboration provided in one embodiment of the present application;

[0038] Fig.12 A schematic diagram of a B / S architecture corresponding to a mathematical physics explanation video generation system based on multi-agent collaboration provided for one embodiment of the present application;

[0039] Fig.13 This is an example diagram of a device for generating a mathematical physics explanation video based on multi-agent collaboration according to an embodiment of the present application;

[0040] Fig.14 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0041] Among them, 10-a device for generating mathematical physics explanation videos based on multi-agent collaboration; 100-acquisition module, 200-control module, 300-explanation video generation module; 1401-memory, 1402-processor, 1403-communication interface. DETAILED DESCRIPTION

[0042] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0043] The following describes the method and device for generating a mathematical physics explanation video based on multi-agent collaboration of the embodiment of the present application with reference to the accompanying drawings. In view of the problems mentioned in the above background technology, the present application provides a method for generating a mathematical physics explanation video based on multi-agent collaboration, in which the actual task requirement information of the mathematical physics explanation video is obtained; based on the actual task requirement information, the preset script master agent is controlled to generate an animation frame script, so that the preset animator agent generates an animation video corresponding to the animation frame script, and the preset explainer agent generates the synchronous subtitles and explanation audio corresponding to the animation video; the animation frame script, the animation video, the synchronous subtitles and the explanation audio are synchronously integrated to construct the initial explanation video corresponding to the mathematical physics explanation video, and the initial explanation video is analyzed by the preset quality inspector agent to obtain the picture analysis result, and the initial explanation video is corrected according to the picture analysis result to generate the final mathematical physics explanation video. The present application improves the generation efficiency and content quality of mathematical physics explanation videos through intelligent technology, and can provide high-quality video teaching content for the field of education. Thus, the existing teaching video production technology solves the problems of poor content accuracy and visual expression effect, high video production cost, and difficulty in meeting actual teaching needs.

[0044] In order to facilitate those skilled in the art to understand the execution logic of the method for generating mathematical physics explanation videos based on multi-agent collaboration of the present application, the types of agents designed in the present application and their functions are briefly introduced and explained as follows.

[0045] Scripter Agent: Generates animation frame scripts based on detailed requirements input by users, and plans the scene layout and visual elements of each frame in detail to ensure that the content meets user needs.

[0046] Animator Agent: Receives and converts animation frame scripts into animation automation execution scripts, and uses built-in video construction tools to generate videos to accurately show the process content.

[0047] Explainer Agent: Analyzes the completed animation content, automatically generates synchronized subtitles, and uses audio generation tools to produce audio to ensure the professionalism and consistency of audio and video content.

[0048] Quality Inspector Agent: Performs strict quality control and review on the generated animation frame scripts, animation automation execution scripts, videos, subtitles, and audio throughout the entire process to ensure that the content meets professional standards and user expectations, thereby improving the accuracy and quality of the video.

[0049] Each agent is capable of self-reflection, including self-checking its output results and immediately modifying and optimizing the original data when problems are found, ensuring that the video has high-quality content; through self-reflection and continuous quality review, each agent continuously improves the accuracy and quality of its generated content, thereby ensuring that the final product meets high quality requirements.

[0050] Specifically, Figure 1 A flowchart of a method for generating mathematical physics explanation videos based on multi-agent collaboration provided in an embodiment of the present application.

[0051] like Figure 1 As shown, the method for generating a mathematical physics explanation video based on multi-agent collaboration includes the following steps:

[0052] In step S101, actual task requirement information of the mathematics and physics explanation video is obtained.

[0053] In step S102, based on the actual task requirement information, the preset scriptwriter agent is controlled to generate an animation frame script, so that the preset animator agent generates an animation video corresponding to the animation frame script, and the preset commentator agent generates synchronized subtitles and commentary audio corresponding to the animation video.

[0054] The embodiments of the present application can firstly obtain the user's actual task requirement information (i.e., user demand) for the mathematical physics lecture video by inputting the user's demand for the mathematical physics lecture video, and pre-process the actual task requirement information, and perform vector analysis and comparison processing on the pre-processed actual task requirement information, so as to improve the efficiency and accuracy of content generation.

[0055] In addition, within the multi-agent collaboration framework, the scripter agent can generate animation frame scripts based on detailed requirements input by the user, including scene arrangement and other key visual elements; the animator agent receives the animation frame scripts and converts them into animation automation execution scripts, and then generates videos using the built-in video construction tool; the narrator agent analyzes the video content, generates synchronized subtitles, and uses the audio generation tool to produce audio to ensure the professionalism and consistency of the audio content; the quality inspector agent conducts strict quality control and review of the generated animation frame scripts, animation automation execution scripts, videos, subtitles, and audio to ensure that the content meets professional standards and user expectations, thereby improving the accuracy and quality of the video. Throughout the entire explanation video generation process, each agent will conduct self-reflection and evaluate the effectiveness and accuracy of its output to ensure continuous improvement and optimization of its performance.

[0056] At the same time, each agent can self-check its output results to improve the accuracy and quality of the generated content. If there are any problems, the agent will immediately modify and optimize the original content data to ensure that the video has high-quality content.

[0057] Optionally, in one embodiment of the present application, based on actual task requirement information, a preset scriptwriter agent is controlled to generate an animation frame script, including: performing preprocessing operations on the actual task requirement information to generate standard task requirement information corresponding to the actual task requirement information, wherein the preprocessing operations include at least one of text cleaning processing, word segmentation processing, and grammatical parsing processing; using a preset embedding model to convert the standard task requirement information into a task requirement vector, and analyzing the vector similarity between the task requirement vector and a preset requirement database; determining whether the similarity is less than a preset similarity threshold, and if the similarity is less than the preset similarity threshold, storing the actual task requirement information; if the similarity analysis result is greater than or equal to the preset similarity threshold, obtaining a target task requirement vector corresponding to the task requirement vector in the requirement database, and generating an animation frame script according to the target task requirement vector through the scriptwriter agent.

[0058] like Figure 2 As shown, the specific process of the embodiment of the present application obtaining the user's actual task requirement information and using vector analysis and comparison technology to control the scripter agent to generate an animation frame script is as follows:

[0059] 1. User demand input: Users describe the content information of mathematics and physics lecture videos in detail through the interactive interface.

[0060] 2. Receive user requirements and perform preprocessing: The system receives user requirements and performs preprocessing operations such as text cleaning, word segmentation, and grammatical analysis on the detailed requirements entered by the user to obtain the actual task requirement information after preprocessing (i.e., standard task requirement information), thereby improving the accuracy and efficiency of subsequent processing.

[0061] 3. Vectorized data conversion: Use the embedding model to convert the preprocessed text (i.e., standard task requirement information) into a vector representation (i.e., task requirement vector) for subsequent database vector similarity comparison and analysis.

[0062] 4. Vector similarity analysis: The system compares the user demand vector with the existing demand vectors stored in the database and retrieves the corresponding vector data (i.e., the target task demand vector).

[0063] 5. Determine the generation strategy and record the analysis results: Based on the results of the database comparison (i.e. the similarity analysis results), determine the best content generation strategy:

[0064] Specifically, the embodiments of the present application can perform vector similarity analysis and set the best content generation strategy from two aspects: high similarity matching, low similarity matching or no matching:

[0065] (1) High similarity matching: When highly similar matching vector data is found after vector similarity comparison (i.e., the similarity analysis result is greater than or equal to the preset similarity threshold), an animation frame script that matches this data will be generated and an animator agent will be assigned to generate the video.

[0066] (2) Low similarity match or no match: In the case of low similarity or no match found (i.e., the similarity analysis result is less than the preset similarity threshold), the user's original demand data (i.e., the actual task demand information) can be retained for more in-depth analysis and content generation processing.

[0067] Therefore, the embodiments of the present application enable the scriptwriter agent to generate animation frame scripts by preprocessing and performing vector similarity analysis on actual task requirement information, thereby providing reliable data support for the generation of subsequent mathematics and physics explanation videos.

[0068] Optionally, in one embodiment of the present application, based on actual task requirement information, a preset scriptwriter agent is controlled to generate an animation frame script, so that the preset animator agent generates an animation video corresponding to the animation frame script, and a preset commentator agent generates synchronized subtitles and commentary audio corresponding to the animation video, including: performing syntax, content and logic self-inspection operations on the animation frame script through the scriptwriter agent to obtain a self-inspection result, and when the self-inspection result meets the preset animation frame script self-inspection requirement, the quality inspector agent performs syntax, content and logic preliminary review operations on the animation frame script to generate a preliminary review result; judging whether the animation frame script meets the first preset creativity and quality standard according to the preliminary review result, if the animation frame script does not meet the first preset creativity and quality standard, the animation frame script is corrected; when the animation frame script meets the first preset creativity and quality standard, the animation frame script is parsed by the animator agent to obtain a parsing result. result; generate an animation automation execution script corresponding to the animation frame script according to the parsing result, and perform syntax and logic review operations on the animation automation execution script through the quality inspector agent to obtain the review result corresponding to the animation automation execution script; when the review result meets the preset syntax and logic standards, call the preset video construction tool through the animator agent to execute the animation automation execution script to generate multiple animation frames corresponding to the animation automation execution script; arrange the multiple animation frames based on the preset arrangement requirements, and compress and format the arranged multiple animation frames according to the preset frame rate and resolution through the target codec to generate an animation video; determine whether the animation video meets the second preset creativity and quality standards through the quality inspector agent, wherein if the animation video does not meet the second preset creativity and quality standards, modify the animation video, otherwise generate the corresponding synchronized subtitles and explanation audio of the animation video through the narrator agent.

[0069] like Figure 3 As shown, the embodiment of the present application carefully compiles an animation frame script including a detailed scene description and logical structure of each frame of animation according to the specific needs of the user (i.e., the actual task requirement information), such as Figure 4 As shown in (a) in the figure, the specific process is as follows:

[0070] 1. Receive and parse system-processed data: The embodiment of the present application can accurately receive and parse the actual task requirement information that has been pre-processed, that is, the standard task requirement information, which is the text description input by the user, after cleaning, detailed word segmentation and strict grammar parsing;

[0071] 2. The Scripter Agent generates animation frame scripts according to user needs: When generating animation frame scripts, the Scripter Agent first conducts deep self-reflection and self-checks the syntax, content richness and logical coherence of the script (i.e., performs syntax, content and logic self-check operations) to obtain self-check results. If the self-check results meet the preset animation frame script self-check requirements and the self-check is confirmed to be correct, the animation frame script will be submitted to the Quality Inspector Agent for preliminary review;

[0072] 3. Quality inspector agent preliminary review of script: The generated animation frame script is carefully reviewed and evaluated by the quality inspector agent (i.e., preliminary review of grammar, content and logic). The review content includes a detailed check of the script's grammar, content creativity and logical consistency;

[0073] 4. Judgment of preliminary review results: Based on the preliminary review results of the quality inspector intelligent agent, judge whether the script meets the predetermined creativity and quality standards (i.e., the first preset creativity and quality standards); if it passes the preliminary review, the animation frame script can enter the next process stage, that is, the animation frame script is parsed by the animator intelligent agent to obtain the parsing results; if it fails, further processing is required and the animation frame script must be corrected.

[0074] Afterwards, if Figure 4 As shown in (b), in the embodiment of the present application, the specific steps of parsing the animation frame script through the animator agent to generate and execute the animation automation execution script and synthesize the animation video according to the parsing result are as follows:

[0075] 1. The animator agent generates an animation automation execution script based on the animation frame script: The animator agent receives and parses the animation frame script provided by the scriptwriter to determine the scene change instructions. The animation automation execution script is generated based on the parsing results, and the layout parameters of the graphic content and visual elements required to be displayed in each frame of the image are specified in detail; during the script generation process, the agent will reflect on itself and evaluate the accuracy of the generated script to ensure that the animation automation execution script can accurately match the description content and visual requirements;

[0076] 2. The quality inspector agent conducts preliminary review of the animation automation execution script: The generated animation automation execution script is carefully reviewed and evaluated by the quality inspector agent (i.e., grammar and logic review operations are performed). The review content includes a detailed check of the script's grammar and logic consistency;

[0077] 3. Call the video construction tool to execute the animation automation execution script to generate animation frames: The animator agent calls the video construction tool to execute the approved animation automation execution script, and generates images frame by frame (i.e., multiple animation frames) according to the script instructions;

[0078] 4. Call the video construction tool to synthesize the animation video: The animator agent uses the video construction tool to synthesize the previously generated frame-by-frame images into a continuous video. The process specifically includes:

[0079] (1) Ensure that all frame images are arranged in the correct order and time interval;

[0080] (2) Select a codec that is suitable for the target playback platform and video quality requirements, such as H.264 or H.265;

[0081] (3) Set the appropriate frame rate (e.g. 24fps, 30fps, or 60fps) and resolution to ensure video clarity and smoothness;

[0082] (4) Compress and format the video to optimize its storage size and transmission efficiency to achieve high-quality visual performance.

[0083] 5. Initial video review by the quality inspector agent: The generated video is carefully reviewed and evaluated by the quality inspector agent, including a detailed check of the video's picture quality, animation smoothness, and consistency of visual effects;

[0084] 6. Judgment of preliminary review results: Based on the preliminary review results of the quality inspector intelligent agent, judge whether the video meets the predetermined creativity and quality standards (i.e., the second preset creativity and quality standards); if it fails the preliminary review, the animation video is corrected, otherwise the narrator intelligent agent generates the corresponding synchronized subtitles and narration audio for the animation video.

[0085] Therefore, the embodiments of the present application enable the scriptwriter agent to generate animation frame scripts through actual task requirement information, and generate animation videos corresponding to the animation frame scripts through the animator agent, thereby providing data guidance and basis for the narrator agent to generate reliable synchronized subtitles and narration audio corresponding to the animation videos.

[0086] Optionally, in one embodiment of the present application, synchronized subtitles and explanation audio corresponding to the animation video are generated by a preset narrator agent, including: generating corresponding synchronized subtitles according to the animation video by the narrator agent, and performing a subtitle review operation on the synchronized subtitles by a quality inspector agent to obtain a subtitle review result; judging whether the synchronized subtitles meet a first preset quality standard according to the subtitle review result, and if the synchronized subtitles do not meet the first preset quality standard, correcting the synchronized subtitles; when the synchronized subtitles meet the first preset quality standard, inputting the synchronized subtitles into a preset explanation audio generation tool by the narrator agent to generate an explanation audio corresponding to the animation video, and performing an audio review operation on the explanation audio by the quality inspector agent to obtain an audio review result; judging whether the audio review result meets a second preset quality standard according to the audio review result, and if the audio review result does not meet the second preset quality standard, correcting the explanation audio.

[0087] In the actual implementation process, Figure 4 As shown in (c), the specific process of generating synchronized subtitles and explanation audio based on the animation video in the embodiment of the present application is as follows:

[0088] 1. The narrator agent generates synchronized subtitles based on the animation video: The narrator agent generates synchronized subtitle text based on the animation video. This process ensures the consistency and accuracy of the subtitles. The narrator agent conducts self-reflection and evaluates the accuracy of the subtitles to ensure that the subtitles can accurately and effectively convey the information and intent of the video;

[0089] 2. Initial review of subtitles by the quality inspector agent: The generated subtitles are initially reviewed and evaluated by the quality inspector agent, including the accuracy and completeness of the subtitles;

[0090] 3. Judgment of preliminary review results: According to the preliminary review results of the quality inspector agent (i.e., the subtitle review results), it is judged whether the synchronized subtitles meet the predetermined quality standards (i.e., the first preset quality standards); if they pass the preliminary review, the synchronized subtitles can enter the next process stage; if they fail, they need to be further processed and corrected;

[0091] 4. Calling an audio generation tool to generate audio: After the synchronized subtitles pass the preliminary review, the narrator agent in the embodiment of the present application can input the generated synchronized subtitles into the audio generation tool to generate an audio explanation that matches the animation video content;

[0092] 5. The quality inspector agent conducts preliminary review of the explanation audio: The generated explanation audio is initially reviewed and evaluated by the quality inspector agent, including the sound quality of the audio and the naturalness of the speech expression;

[0093] 6. Judgment of preliminary review results: Based on the preliminary review results of the quality inspector intelligent body, judge whether the explanation audio meets the predetermined quality standards (i.e., the second preset quality standards); if it passes the preliminary review, the explanation audio can enter the next process stage; if it fails, it needs to be further processed and the explanation audio needs to be corrected.

[0094] Therefore, the embodiments of the present application improve the efficiency and accuracy of video content generation by enabling each agent to collaborate to complete the workflow according to task requirements within a multi-agent collaboration framework, thereby ensuring the accuracy and high quality of the generated content.

[0095] In step S103, the animation frame script, animation video, synchronized subtitles and explanation audio are synchronously integrated to construct an initial explanation video corresponding to the mathematical physics explanation video, and the initial explanation video is analyzed by a preset quality inspector agent to obtain a picture analysis result, and the initial explanation video is corrected according to the picture analysis result to generate a final mathematical physics explanation video.

[0096] After generating the synchronized subtitles and explanation audio of the animation video, the embodiment of the present application further needs to synchronously integrate the animation frame script, animation video, synchronized subtitles and explanation audio to construct an initial explanation video; the quality inspector intelligent agent uses spatial perception enhancement technology to perform a detailed analysis of the picture of the initial video to obtain the picture analysis results, and performs high-precision quality assessment and necessary corrections on the initial explanation video based on the picture analysis results, and finally generates a high-quality mathematics and physics explanation video (i.e., the final mathematics and physics explanation video).

[0097] Therefore, the embodiments of the present application provide high-quality video teaching content for the education field through an intelligent, efficient and easily expandable mathematics and physics explanation video generation strategy.

[0098] Optionally, in one embodiment of the present application, a preset quality inspector agent is used to perform image analysis on the initial explanation video to obtain an image analysis result, and the initial explanation video is corrected according to the image analysis result to generate a final mathematics and physics explanation video, including: decomposing the initial explanation video into multiple video frames, and performing video frame quality assessment on each of the multiple video frames through the spatial perception enhancement strategy of the quality inspector agent to obtain a quality assessment result corresponding to each video frame; judging whether the initial explanation video meets the preset explanation video quality standard according to the quality assessment result, if the initial explanation video meets the explanation video quality standard, then generating the final mathematics and physics explanation video, otherwise correcting the animation frame script, animation video, synchronized subtitles and explanation audio, and re-synchronizing and integrating the revised animation frame script, animation video, synchronized subtitles and explanation audio and performing video frame quality assessment operations to generate the corresponding final mathematics and physics explanation video.

[0099] It should be noted that if Figure 5 As shown, the specific process of synchronously integrating the animation frame script, the animation automation execution script, the animation video, the synchronous subtitles and the explanation audio to construct the initial explanation video in the embodiment of the present application is as follows:

[0100] 1. Collect and summarize optimized animation frame scripts, animation automation execution scripts, animation videos, synchronized subtitles and explanation audio;

[0101] 2. Integrate animation frame scripts and other data: Use video editing and timeline management tools to synchronously integrate animation frame scripts, animation automation execution scripts, animation videos, synchronized subtitles and audio narration to ensure video logic coherence and coordination between audio and video; if integration fails, analyze and handle the reasons for the failure and reintegrate;

[0102] 3. Generate an initial explanation video: Generate an initial explanation video based on the synchronized and integrated data.

[0103] like Figure 6 As shown, in the embodiment of the present application, the quality inspector intelligent agent uses the spatial perception enhancement technology to perform a detailed picture analysis on the initial explanation video, and performs a high-precision quality assessment and necessary corrections on the initial explanation video according to the picture analysis results, and finally generates a high-quality mathematics and physics explanation video. The specific process is as follows:

[0104] 1. The quality inspector agent receives the initial explanation video: The quality inspector agent receives the generated initial explanation video for detailed quality assessment and analysis;

[0105] 2. Divide the initial explanation video into frames: The video processing module decomposes the received video into multiple video frames to prepare for subsequent evaluation and processing;

[0106] 3. The quality inspector agent uses spatial perception enhancement technology to conduct a detailed evaluation of the video frames: The quality inspector agent uses spatial perception enhancement technology to conduct an in-depth evaluation of the visual effects and scene spatial logic of each video frame to obtain the quality evaluation results corresponding to each video frame;

[0107] 4. The quality inspector agent generates a detailed quality assessment report: The quality inspector agent generates a detailed quality report based on the assessment results, which includes analysis of various indicators and improvement suggestions;

[0108] 5. Determine whether it passes the quality standards: Determine whether the initial explanation video meets the predetermined explanation video quality standards based on the quality assessment results; if the initial explanation video passes the quality standards, generate the final high-quality video; if it fails, make necessary corrections to the initial explanation video content for each agent based on the quality assessment results, and then integrate the new initial explanation video for re-examination.

[0109] To sum up, with the continuous advancement of the digital age, video has become an important carrier of education and information dissemination. The embodiments of the present application can complete the corresponding mathematics and physics explanation video generation tasks through the collaboration of multiple intelligent agents; specifically, the embodiments of the present application can be responsible for generating detailed animation frame scripts through the scriptwriter intelligent agent; the animator intelligent agent converts the animation frame script into an animation automation execution script and generates an animation video; the lecturer intelligent agent generates the corresponding synchronized subtitles and explanation audio for the animation video; the quality inspector intelligent agent conducts strict quality control and review of the generated animation frame scripts, animation automation execution scripts, animation videos, synchronized subtitles and explanation audio throughout the process to ensure that all content meets professional standards and user expectations, thereby improving the accuracy and quality of the teaching content.

[0110] It can be understood that the embodiments of the present application not only emphasize the technical innovation of current commentary content generation, but also focus on user experience and the audio-visual effects of the final product, thereby effectively meeting the needs of modern users for high-quality and diversified audio-visual content.

[0111] The following is a detailed description of the execution process of the method for generating mathematical physics explanation videos based on multi-agent collaboration of the present application through a specific embodiment and in combination with the accompanying drawings.

[0112] Figure 7 This is a schematic diagram of the execution logic of the method for generating a mathematical physics explanation video based on multi-agent collaboration in this application. Figure 7 As shown, the execution process of the method for generating a mathematical physics explanation video based on multi-agent collaboration of the present application is as follows:

[0113] S701: Obtaining the user's demand information for mathematics and physics explanation videos, and using vector analysis and comparison technology to improve the efficiency and accuracy of content generation;

[0114] S702: Within the multi-agent collaboration framework, each agent collaborates to complete the workflow according to task requirements to ensure the accuracy and high quality of the generated content;

[0115] S703: Integrate and synchronize all generated data to construct an initial explanation video;

[0116] S704: The quality inspector agent uses spatial perception enhancement technology to conduct a detailed analysis of the images of the initial explanation video, perform high-precision quality assessment and necessary corrections, and ultimately generate a high-quality mathematics and physics explanation video.

[0117] In a specific embodiment of the present application, Python programming language and Manim Python library are used as examples to demonstrate how to implement the above-mentioned method for generating mathematical physics explanation videos.

[0118] It should be noted that the specific embodiments of the present application may use the Python programming language and the Manim Python library to demonstrate the flexibility and openness of the solution. However, the technical implementation is not limited to these specific tools, software or programming languages. Those skilled in the art can select tools, software or programming languages ​​according to actual conditions during the specific implementation process to generate the above-mentioned mathematical physics explanation video, thereby reflecting the inclusiveness of the present application for a wide range of technical options and the pursuit of diversified technical solutions.

[0119] Specifically, users can make a mathematical physics explanation video about the proof of the Pythagorean theorem, and by adopting vector analysis and comparison technology, the efficiency and accuracy of content generation can be significantly improved; within the multi-agent collaboration framework, the scriptwriter agent generates animation frame scripts, and plans the scene layout and visual elements of each frame in detail to ensure that the content meets user needs; the animator agent receives the animation frame script, uses the Python programming language to write the animation automation execution script, and uses the Manim Python library in the video construction tool to execute the animation automation execution script to generate an animation video; the commentator agent generates synchronized subtitles based on the generated animation video, and uses audio tools to generate explanation audio based on the synchronized subtitles, thereby improving the professionalism and viewing experience of the video; the quality inspector agent performs quality control on the entire process, reviews grammar, content innovation, and visual and auditory effects to ensure that the video meets the highest quality standards.

[0120] In addition, each intelligent agent has the ability of self-reflection, can evaluate the effectiveness and accuracy of its output, and make necessary adjustments and optimizations, so as to continuously improve the accuracy and quality of the content; then, synchronously integrate all generated data to construct an initial explanation video; then, the initial video is divided into video frames, and the quality inspector intelligent agent uses spatial perception enhancement technology to conduct a detailed analysis of the initial explanation video, conduct a high-precision quality assessment, and make necessary corrections, finally generating a high-quality mathematics and physics explanation video.

[0121] FIG8 is a schematic diagram of the effect of the application example of the present application. As shown in FIG8, it shows the three stages of the original video frame, grid processing and corrected video frame of the spatial perception enhancement technology, as described below:

[0122] 1. Original video frame: displays basic graphics and annotations;

[0123] 2. Grid processing: Adding grids and coordinate systems to the original graphics can more accurately identify and measure, and help understand the relationship between graphics;

[0124] 3. Corrected video frame: After analysis and correction using spatial perception enhancement technology, the grid and coordinate system are removed to clearly display the corrected graphics.

[0125] Therefore, the embodiments of the present application not only ensure the visual effects of teaching videos through sophisticated image processing, but also reflect the important role of spatial perception enhancement technology in improving the quality of visual presentation and educational content.

[0126] The following is a corresponding mathematical physics explanation video generation system based on multi-agent collaboration according to the mathematical physics explanation video generation method based on multi-agent collaboration of the present application, and the mathematical physics explanation video generation system based on multi-agent collaboration of the present application is introduced in detail in conjunction with the accompanying drawings.

[0127] Fig. 9 This is a schematic diagram of the logical architecture of the mathematical physics explanation video generation system based on multi-agent collaboration in this application. Fig. 9 As shown, the mathematical physics explanation video generation system based on multi-agent collaboration of the present application mainly includes a user module, a system module, a multi-agent role module based on a large model and a video processing module.

[0128] Among them, the user module: This module provides an interactive platform that enables users to directly input video requirements; users can describe video requirements in detail and convert these requirements into clear task instructions; this module focuses on user experience, ensuring that the input process is both intuitive and efficient, laying the foundation for the subsequent content generation process;

[0129] System module: As the core hub for processing and converting user input, this module receives data provided by the user module and converts it into a data format suitable for system processing through pre-processing steps such as text cleaning, word segmentation, syntax parsing and vectorization. The main task of this module is to ensure the accuracy and consistency of input data to correctly guide the subsequent content generation and analysis process;

[0130] Multi-agent role module based on large models: This module includes a series of specialized agents, each responsible for different tasks in the process. These agents use artificial intelligence technology to generate specific content elements based on the data provided by the system modules. This module design allows efficient collaboration and information exchange between agents to ensure that the video content meets both user needs and high quality standards;

[0131] Video processing module: This module is responsible for integrating the scripts, audio, video and subtitles generated by each intelligent agent into a coherent video file; this module uses the timeline tool to ensure the temporal and logical consistency of audio, video and subtitles; after the integration is completed, the video processing module performs the final video rendering, compression and formatting to ensure that the output video meets industry standards and the quality expected by users; in addition, this module also performs final quality control, including adjusting the color, brightness and sound balance of the video to provide the best viewing experience.

[0132] Fig.10 This is a user-system role interaction diagram of the multi-agent collaborative mathematical physics video generation system of this application. Fig.10 As shown, the user inputs the requirements through the user module, and the system module preprocesses the requirement data and passes it to the scriptmaster agent in the multi-agent role module to generate an animation frame script; the script is passed to the animator agent, which generates and executes the animation script to generate an animation video; the animation video is passed to the quality inspector agent for quality inspection and feedback; the commentator agent generates synchronized subtitles and commentary audio based on the video, which are reviewed again by the quality inspector agent; finally, the video processing module integrates all elements to generate high-quality mathematics and physics explanation videos, and each agent works closely together to ensure that the generated content is efficient and high-quality.

[0133] Fig.11 This is a schematic diagram of the spatial perception enhancement technology workflow of the multi-agent collaborative mathematical physics explanation video generation system of this application. Fig.11 As shown, the spatial perception enhancement technology workflow of the embodiment of the present application is as follows:

[0134] 1. In the image preprocessing stage, a variety of techniques are used to optimize the image:

[0135] Improve the sharpness and detail of the image through clarity enhancement technology to ensure the accurate communication of visual information; use color correction technology to adjust the color balance and contrast of the image to make it more in line with the real scene or specific needs; remove noise and interference in the image through denoising processing to improve the overall quality of the image; finally, apply super-resolution technology to improve the image resolution and enhance the detail expression through deep learning algorithms or interpolation methods;

[0136] 2. Add precise grid structure to the preprocessed image:

[0137] By superimposing a regular grid on the image, a reference frame and coordinate reference system are provided, providing an accurate data basis for subsequent spatial perception processing;

[0138] 3. Spatial perception processing:

[0139] First, a multimodal large model is used to perform preliminary coordinate positioning of the image with a grid. This usually involves using a pre-trained object detector or semantic segmentation network to preliminarily identify and locate the main objects or regions in the image. Subsequently, the preliminarily located objects or regions are iteratively enlarged or reduced through scaling and refinement processing to further extract and refine spatial information, thereby improving the precision and accuracy of positioning.

[0140] 4. Evaluate the exact coordinates:

[0141] First, check whether the final coordinate positioning meets the predetermined accuracy requirements and application requirements. If the positioning result meets the accuracy requirements and application requirements, enter the reflection and optimization stage, which may include adjusting the positioning parameters to further improve the performance and stability of the spatial perception technology; if it does not meet the accuracy requirements and application requirements, continue to iterate the scaling and refinement processing until the required accuracy is achieved;

[0142] Therefore, the embodiments of the present application integrate image processing and multimodal large models through spatial perception enhancement technology, which can significantly improve the positioning accuracy of targets and areas in complex scenes, enhance the ability to understand the environment, and ensure efficient target positioning and accurate environmental perception in various application scenarios, thereby meeting complex practical needs.

[0143] Fig.12 This is a schematic diagram of the B / S architecture corresponding to the mathematical physics explanation video generation system based on multi-agent collaboration in this application. Fig.12 As shown, the B / S architecture of the embodiment of the present application is composed of a client browser and a server, wherein the server includes an intelligent agent and a video rendering tool.

[0144] Specifically, users can interact with the agent through the browser and submit video production requests; the agent receives and interprets user requirements on the server side and is responsible for sending specific task instructions to the video rendering tool. The video rendering tool is deployed on the server side and is responsible for video generation and processing. It can efficiently handle complex video rendering tasks and ensure that the generated video is of high quality and good effect. After rendering, the generated video file is received by the agent and fed back to the user's browser. The user can directly watch or download the final video file in the browser. The architecture ensures the flexibility, reliability and cost-effectiveness of the system through the scalability and high availability provided by the cloud platform.

[0145] According to the method for generating a mathematical physics explanation video based on multi-agent collaboration proposed in the embodiment of the present application, the actual task requirement information of the mathematical physics explanation video is obtained; based on the actual task requirement information, the preset scripter agent is controlled to generate an animation frame script, so that the preset animator agent generates an animation video corresponding to the animation frame script, and the preset explainer agent generates the synchronized subtitles and explanation audio corresponding to the animation video; the animation frame script, the animation video, the synchronized subtitles and the explanation audio are synchronously integrated to construct an initial explanation video corresponding to the mathematical physics explanation video, and the initial explanation video is analyzed by the preset quality inspector agent to obtain the picture analysis result, and the initial explanation video is corrected according to the picture analysis result to generate the final mathematical physics explanation video. This application improves the generation efficiency and content quality of mathematical physics explanation videos through intelligent technology, and can provide high-quality video teaching content for the education field.

[0146] Secondly, the apparatus for generating mathematical physics explanation videos based on multi-agent collaboration proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0147] Fig.13 It is a block diagram of a device for generating a mathematical physics explanation video based on multi-agent collaboration according to an embodiment of the present application.

[0148] like Fig.13 As shown, the mathematical physics explanation video generation device 10 based on multi-agent collaboration includes: an acquisition module 100, a control module 200 and an explanation video generation module 300.

[0149] Among them, the acquisition module 100 is used to obtain the actual task requirement information of the mathematics and physics explanation video.

[0150] The control module 200 is used to control the preset scriptwriter agent to generate an animation frame script based on actual task requirement information, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generates synchronized subtitles and commentary audio corresponding to the animation video through the preset commentator agent.

[0151] The explanation video generation module 300 is used to synchronously integrate the animation frame script, animation video, synchronized subtitles and explanation audio to construct an initial explanation video corresponding to the mathematics and physics explanation video, and perform picture analysis on the initial explanation video through a preset quality inspector intelligent agent to obtain a picture analysis result, and correct the initial explanation video according to the picture analysis result to generate a final mathematics and physics explanation video.

[0152] Optionally, in one embodiment of the present application, the control module 200 includes: a preprocessing unit, a conversion unit, a first judgment unit and a comparison unit.

[0153] Among them, the preprocessing unit is used to perform preprocessing operations on the actual task requirement information to generate standard task requirement information corresponding to the actual task requirement information, wherein the preprocessing operation includes at least one of text cleaning processing, word segmentation processing and grammar parsing processing.

[0154] The conversion unit is used to convert the standard task requirement information into a task requirement vector by using a preset embedding model, and analyze the vector similarity between the task requirement vector and a preset requirement database.

[0155] The first judgment unit is used to judge whether the similarity is less than a preset similarity threshold, and if the similarity is less than the preset similarity threshold, the actual task requirement information is stored.

[0156] The comparison unit is used to obtain the target task requirement vector corresponding to the task requirement vector in the requirement database if the similarity analysis result is greater than or equal to the preset similarity threshold, and generate an animation frame script according to the target task requirement vector through the scriptwriter intelligent agent.

[0157] Optionally, in one embodiment of the present application, the control module 200 further includes: a self-checking unit, a second judgment unit, a parsing unit, a first review unit, an execution unit, an arrangement unit and a third judgment unit.

[0158] Among them, the self-check unit is used to perform syntax, content and logic self-check operations on the animation frame script through the scriptwriter intelligent agent to obtain a self-check result, and when the self-check result meets the preset animation frame script self-check requirements, the quality inspector intelligent agent performs a preliminary review operation on the syntax, content and logic of the animation frame script to generate a preliminary review result.

[0159] The second judgment unit is used to judge whether the animation frame script meets the first preset creativity and quality standards according to the preliminary review result, and if the animation frame script does not meet the first preset creativity and quality standards, the animation frame script is corrected.

[0160] The parsing unit is used to parse the animation frame script through the animator intelligent agent to obtain a parsing result when the animation frame script meets the first preset creativity and quality standards.

[0161] The first review unit is used to generate an animation automation execution script corresponding to the animation frame script according to the analysis result, and perform syntax and logic review operations on the animation automation execution script through the quality inspector intelligent agent to obtain a review result corresponding to the animation automation execution script.

[0162] The execution unit is used to call a preset video construction tool through an animator agent to execute the animation automation execution script when the review result meets the preset syntax and logic standards, so as to generate multiple animation frames corresponding to the animation automation execution script.

[0163] The arrangement unit is used to arrange multiple animation frames based on preset arrangement requirements, and compress and format the arranged multiple animation frames according to preset frame rate and resolution through a target codec to generate an animation video.

[0164] The third judgment unit is used to judge whether the animation video meets the second preset creativity and quality standards through the quality inspector intelligent agent, wherein if the animation video does not meet the second preset creativity and quality standards, the animation video is corrected, otherwise the corresponding synchronized subtitles and commentary audio of the animation video are generated through the commentator intelligent agent.

[0165] Optionally, in one embodiment of the present application, the control module 200 further includes: a second review unit, a fourth judgment unit, a third review unit and a fifth judgment unit.

[0166] Among them, the second review unit is used to generate corresponding synchronized subtitles according to the animation video through the narrator intelligent agent, and use the quality inspector intelligent agent to perform subtitle review operations on the synchronized subtitles to obtain subtitle review results.

[0167] The fourth judgment unit is used to judge whether the synchronized subtitles meet the first preset quality standard according to the subtitle review result, and if the synchronized subtitles do not meet the first preset quality standard, the synchronized subtitles are corrected.

[0168] The third review unit is used to input the synchronized subtitles into a preset explanation audio generation tool through the narrator intelligent agent when the synchronized subtitles meet the first preset quality standard, so as to generate the explanation audio corresponding to the animation video, and perform an audio review operation on the explanation audio through the quality inspector intelligent agent to obtain the audio review result.

[0169] The fifth judgment unit is used to judge whether the audio review result meets the second preset quality standard based on the audio review result, wherein if the audio review result does not meet the second preset quality standard, the explanation audio is corrected.

[0170] Optionally, in one embodiment of the present application, the explanation video generation module 300 includes: a decomposition unit and a correction unit.

[0171] Among them, the decomposition unit is used to decompose the initial explanation video into multiple video frames, and perform video frame quality assessment on each of the multiple video frames through the spatial perception enhancement strategy of the quality inspector intelligent agent to obtain the quality assessment result corresponding to each video frame.

[0172] The correction unit is used to determine whether the initial explanation video meets the preset explanation video quality standard based on the quality assessment result. If the initial explanation video meets the explanation video quality standard, a final mathematics and physics explanation video is generated. Otherwise, the animation frame script, animation video, synchronized subtitles and explanation audio are corrected, and the corrected animation frame script, animation video, synchronized subtitles and explanation audio are re-synchronized and the video frame quality assessment operation is performed to generate the corresponding final mathematics and physics explanation video.

[0173] It should be noted that the aforementioned explanation of the embodiment of the method for generating a mathematical physics explanation video based on multi-agent collaboration is also applicable to the apparatus for generating a mathematical physics explanation video based on multi-agent collaboration of this embodiment, and will not be repeated here.

[0174] According to the mathematical physics explanation video generation device based on multi-agent collaboration proposed in the embodiment of the present application, it includes an acquisition module 100 for acquiring the actual task requirement information of the mathematical physics explanation video; a control module 200 for controlling the preset script master agent to generate an animation frame script based on the actual task requirement information, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generates the synchronous subtitles and explanation audio corresponding to the animation video through the preset explainer agent; an explanation video generation module 300 for synchronously integrating the animation frame script, the animation video, the synchronous subtitles and the explanation audio to construct an initial explanation video corresponding to the mathematical physics explanation video, and performing picture analysis on the initial explanation video through the preset quality inspector agent to obtain the picture analysis result, and correcting the initial explanation video according to the picture analysis result to generate the final mathematical physics explanation video. The present application improves the generation efficiency and content quality of mathematical physics explanation videos through intelligent technology, and can provide high-quality video teaching content for the education field.

[0175] Fig.14 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0176] A memory 1401 , a processor 1402 , and a computer program stored in the memory 1401 and executable on the processor 1402 .

[0177] When the processor 1402 executes the program, it implements the method for generating mathematical physics explanation videos based on multi-agent collaboration provided in the above embodiment.

[0178] Furthermore, the electronic device further comprises:

[0179] The communication interface 1403 is used for communication between the memory 1401 and the processor 1402 .

[0180] The memory 1401 is used to store computer programs that can be executed on the processor 1402 .

[0181] The memory 1401 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0182] If the memory 1401, the processor 1402 and the communication interface 1403 are implemented independently, the communication interface 1403, the memory 1401 and the processor 1402 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig.14 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0183] Optionally, in a specific implementation, if the memory 1401, the processor 1402 and the communication interface 1403 are integrated on a chip, the memory 1401, the processor 1402 and the communication interface 1403 can communicate with each other through an internal interface.

[0184] The processor 1402 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0185] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for generating a mathematical physics explanation video based on multi-agent collaboration.

[0186] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed, is used to implement the above-mentioned method for generating mathematical physics explanation videos based on multi-agent collaboration.

[0187] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0188] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0189] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0190] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or N wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways as necessary and then storing it in a computer memory.

[0191] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0192] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0193] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0194] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for generating mathematical physics explanation videos based on multi-agent collaboration, characterized in that: The multi-agent comprises a script writer agent, an animator agent, a lecturer agent and a quality inspector agent, wherein the method for generating a mathematical physics explanation video based on multi-agent collaboration comprises the following steps: Get the actual task requirement information of mathematics and physics explanation videos; Based on the actual task requirement information, control the preset scripter agent to generate an animation frame script, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generates synchronized subtitles and explanation audio corresponding to the animation video through the preset commentator agent; Synchronously integrating the animation frame script, the animation video, the synchronized subtitles and the explanation audio to construct an initial explanation video corresponding to the mathematical physics explanation video, and performing screen analysis on the initial explanation video through a preset quality inspector agent to obtain a screen analysis result, and correcting the initial explanation video according to the screen analysis result to generate a final mathematical physics explanation video; Wherein, based on the actual task requirement information, controlling the preset scripter agent to generate the animation frame script includes: Performing a preprocessing operation on the actual task requirement information to generate standard task requirement information corresponding to the actual task requirement information, wherein the preprocessing operation includes at least one of text cleaning processing, word segmentation processing, and grammar parsing processing; Converting the standard task requirement information into a task requirement vector using a preset embedding model, and analyzing the vector similarity between the task requirement vector and a preset requirement database; Determining whether the similarity is less than a preset similarity threshold, and if the similarity is less than the preset similarity threshold, storing the actual task requirement information; If the similarity analysis result is greater than or equal to the preset similarity threshold, the target task requirement vector corresponding to the task requirement vector in the requirement database is obtained, and the animation frame script is generated according to the target task requirement vector by the scriptwriter agent.

2. The method for generating a mathematical physics explanation video based on multi-agent collaboration according to claim 1, characterized in that: The method of controlling a preset scripter agent to generate an animation frame script based on the actual task requirement information, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generates synchronized subtitles and explanation audio corresponding to the animation video through a preset commentator agent, includes: The scripter agent performs a syntax, content and logic self-check operation on the animation frame script to obtain a self-check result, and when the self-check result meets the preset animation frame script self-check requirement, the quality inspector agent performs a syntax, content and logic preliminary review operation on the animation frame script to generate a preliminary review result; Determining whether the animation frame script meets a first preset creativity and quality standard according to the preliminary review result, and if the animation frame script does not meet the first preset creativity and quality standard, revising the animation frame script; When the animation frame script meets the first preset creativity and quality standard, the animator agent parses the animation frame script to obtain a parsing result; Generate an animation automation execution script corresponding to the animation frame script according to the parsing result, and perform syntax and logic review operations on the animation automation execution script by the quality inspector agent to obtain a review result corresponding to the animation automation execution script; When the review result meets the preset grammar and logic standards, the animator agent calls a preset video construction tool to execute the animation automation execution script to generate a plurality of animation frames corresponding to the animation automation execution script; Arranging the plurality of animation frames based on a preset arrangement requirement, and compressing and formatting the arranged plurality of animation frames according to a preset frame rate and resolution through a target codec to generate the animation video; The quality inspector agent is used to determine whether the animation video meets the second preset creativity and quality standards, wherein if the animation video does not meet the second preset creativity and quality standards, the animation video is corrected, otherwise the commentator agent generates synchronized subtitles and commentary audio corresponding to the animation video.

3. The method for generating a mathematical physics explanation video based on multi-agent collaboration according to claim 2, characterized in that: The generating of the synchronous subtitles and explanation audio corresponding to the animation video by a preset narrator agent includes: Generate corresponding synchronized subtitles according to the animation video by the narrator agent, and perform subtitle review operation on the synchronized subtitles by the quality inspector agent to obtain subtitle review results; determining whether the synchronized subtitles meet a first preset quality standard according to the subtitle review result, and if the synchronized subtitles do not meet the first preset quality standard, modifying the synchronized subtitles; In the case where the synchronized subtitles meet the first preset quality standard, the synchronized subtitles are input into a preset explanation audio generation tool through the explainer agent to generate the explanation audio corresponding to the animation video, and the quality inspector agent performs an audio review operation on the explanation audio to obtain an audio review result; Determine whether the audio review result meets a second preset quality standard based on the audio review result, wherein if the audio review result does not meet the second preset quality standard, modify the explanation audio.

4. The method for generating a mathematical physics explanation video based on multi-agent collaboration according to claim 1, characterized in that: The method of performing screen analysis on the initial explanation video by a preset quality inspector agent to obtain a screen analysis result, and correcting the initial explanation video according to the screen analysis result to generate a final mathematical physics explanation video includes: Decomposing the initial explanation video into multiple video frames, and performing video frame quality assessment on each of the multiple video frames through the spatial perception enhancement strategy of the quality inspector agent to obtain a quality assessment result corresponding to each video frame; According to the quality assessment result, it is determined whether the initial explanation video meets the preset explanation video quality standard. If the initial explanation video meets the explanation video quality standard, the final mathematics and physics explanation video is generated. Otherwise, the animation frame script, the animation video, the synchronized subtitles and the explanation audio are corrected, and the corrected animation frame script, animation video, synchronized subtitles and explanation audio are re-synchronized and the video frame quality assessment operation is performed to generate the corresponding final mathematics and physics explanation video.

5. A device for generating mathematical physics explanation videos based on multi-agent collaboration, characterized in that: The multi-agents include a script writer agent, an animator agent, a lecturer agent, and a quality inspector agent, wherein the apparatus for generating a mathematical physics explanation video based on multi-agent collaboration includes: An acquisition module is used to obtain the actual task requirement information of the mathematics and physics explanation video; A control module is used to control a preset scripter agent to generate an animation frame script based on the actual task requirement information, so that the preset animator agent generates an animation video corresponding to the animation frame script, and generates synchronized subtitles and explanation audio corresponding to the animation video through a preset commentator agent; An explanation video generation module is used to synchronously integrate the animation frame script, the animation video, the synchronized subtitles and the explanation audio to construct an initial explanation video corresponding to the mathematical physics explanation video, and to perform screen analysis on the initial explanation video through a preset quality inspector agent to obtain a screen analysis result, and to correct the initial explanation video according to the screen analysis result to generate a final mathematical physics explanation video; Wherein, the control module comprises: A preprocessing unit, configured to perform a preprocessing operation on the actual task requirement information to generate standard task requirement information corresponding to the actual task requirement information, wherein the preprocessing operation includes at least one of text cleaning processing, word segmentation processing, and grammar parsing processing; A conversion unit, used to convert the standard task requirement information into a task requirement vector using a preset embedding model, and analyze the vector similarity between the task requirement vector and a preset requirement database; A first judging unit, configured to judge whether the similarity is less than a preset similarity threshold, and if the similarity is less than the preset similarity threshold, storing the actual task requirement information; A comparison unit is used to obtain a target task requirement vector corresponding to the task requirement vector in the requirement database if the similarity analysis result is greater than or equal to the preset similarity threshold, and generate the animation frame script according to the target task requirement vector through the scriptwriter agent.

6. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for generating mathematical physics explanation videos based on multi-agent collaboration as described in any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the method for generating mathematical physics explanation videos based on multi-agent collaboration as described in any one of claims 1 to 4.

8. A computer program product, comprising a computer program, characterized in that The computer program is executed to implement the method for generating mathematical physics explanation videos based on multi-agent collaboration as described in any one of claims 1 to 4.

Images