An automated video clip method and system for clip single recommendation

Abstract

The application discloses an automatic video clipping method and system for film list recommendation, comprising: acquiring film data and preprocessing the film data; generating a recommendation list and basic data of a film list by using a preset large model and a prompt word according to the preprocessed film data; generating a script corresponding to the film list data according to the recommendation list and the basic data, and selecting corresponding film resources and background music; performing video synthesis and clipping according to the script, the film resources and the background music, and outputting an automatic clipping video for film list recommendation. The application can dynamically adjust material combination logic according to a film list theme, deeply excavate and structureally apply film list recommendation data, and realize efficient and large-scale generation of a film list recommendation video.

Description

Technical Field

[0001] This invention relates to the field of video processing technology, and in particular to an automated video editing method and system for recommending video lists. Background Technology

[0002] With the surge in demand for personalized video playlist recommendations on short video platforms (e.g., film and television commentary compilations, themed mashups), automated video editing technology is gradually shifting from single-scene editing to intelligent production that integrates multiple materials. Current mainstream technologies primarily focus on processing single video footage (e.g., shot segmentation, filter adaptation), but when generating playlist-based recommended content, they still rely on manual material selection, script writing, and manual editing and splicing, resulting in low efficiency and difficulty in scaling. For example, Douyin's "film and television mashup" template only supports matching background music to fixed shot durations and cannot dynamically adjust the material combination logic based on the playlist theme; while automated tools commonly used by broadcasters on some websites (e.g., a certain AI editing tool) can generate basic clips, they lack in-depth mining and structured application of playlist recommendation data (e.g., user review keywords, film emotional tags).

[0003] In summary, existing automated video editing technologies still rely on manual material selection, script writing, and manual editing and splicing when generating recommended content for video lists, which is inefficient and difficult to scale. Therefore, existing technologies need to be improved. Summary of the Invention

[0004] The technical problem this invention aims to solve is to address the shortcomings of existing technologies by providing an automated video editing method and system for video list recommendation. This addresses the issue that existing automated video editing technologies, when generating video list-based recommendation content, still rely on manual material selection, script writing, and manual editing and splicing, resulting in low efficiency and difficulty in scaling.

[0005] The technical solution adopted by this invention to solve the technical problem is as follows:

[0006] In a first aspect, the present invention provides an automated video editing method for recommending video lists, including:

[0007] Acquire video data and preprocess the video data;

[0008] Based on the pre-processed video data, a recommended list and basic data for the video list are generated using a preset large model and prompt words;

[0009] Based on the recommended list and basic data, generate the text corresponding to the movie list data, and select the corresponding movie resources and background music;

[0010] Based on the text, the video resources, and the background music, the video is synthesized and edited to output an automated edited video recommended by the playlist.

[0011] In one implementation, video data is acquired and preprocessed, including:

[0012] The film data is obtained from a pre-set media website; wherein, the film data includes: film rating, film introduction, actor information, film reviews, and viewing experience;

[0013] The video data is cleaned, videos that do not meet the conditions are filtered out, and videos of the same type are put into the same dataset to obtain the preprocessed video data.

[0014] In one implementation, the step of generating a recommended list and basic data for the film list based on the preprocessed film data using a preset large model and prompt words includes:

[0015] The preprocessed video data is input into the preset large model;

[0016] The prompt words are generated based on preference tags, scene descriptions, number of videos, and themes / types.

[0017] Based on the preset large model, the prompt words are used to generate a recommendation list containing movie titles and basic data containing movies and background music.

[0018] In one implementation, generating the copy corresponding to the video list data based on the recommendation list and basic data includes:

[0019] The recommended list and the basic data are input into the preset large model. Based on the editing operation of the film list recommendation, corresponding introductory text containing the film is generated. A fixed sentence pattern is added after the introductory text to obtain the text corresponding to the film list data.

[0020] In one implementation, selecting the corresponding video resources and background music includes:

[0021] Download the film resources corresponding to each film list from the media resource library corresponding to the basic data, obtain the introductory narration, add other prompt words to the introductory narration, and select film clips as the screen content when the introductory narration is explained based on the preset big model;

[0022] Download the music corresponding to each video playlist from the media resource library corresponding to the basic data, and select the key parts of the music based on the preset big model. Use the key parts of the music as the background music for the key shots in the recommended videos of the video playlist.

[0023] In one implementation, the step of performing video synthesis and editing based on the text, the video resources, and the background music to output an automated edited video for a recommended video list includes:

[0024] Based on the text, the video resources, and the background music, a video synthesis and editing process is performed using a preset editing template and a recommended video list structure, outputting the automatically edited video recommended by the video list.

[0025] In one implementation, the step of performing video synthesis and editing based on the text, the video resources, and the background music using a preset editing template and a recommended video list structure, and outputting the automatically edited video recommended by the video list, includes:

[0026] Set the first emoji on the opening screen of the recommended videos in the playlist, along with a greeting and author information;

[0027] Based on the video resources, set the corresponding high-energy scenes of the videos recommended in the video list, and configure the MP3 version of the introductory text, the corresponding text information, and the background music when the key scenes are reached.

[0028] Set a second emoji at the end of the recommended videos in the playlist, and set a preset ending phrase;

[0029] Output the automatically edited videos recommended by the aforementioned video list.

[0030] Secondly, the present invention provides an automated video editing system for recommending video lists, comprising:

[0031] The data acquisition module is used to acquire video data and preprocess the video data;

[0032] The recommendation list generation module is used to generate a recommendation list and basic data for the film list based on the preprocessed film data, using a preset large model and prompt words;

[0033] The video material module is used to generate text corresponding to the video list data based on the recommendation list and basic data, and to select the corresponding video resources and background music;

[0034] The compositing and editing module is used to perform video compositing and editing based on the text, the video resources, and the background music, and output an automated edited video recommended by the video list.

[0035] Thirdly, the present invention provides a terminal, comprising: a processor and a memory, wherein the memory stores an automated video editing program for recommending video lists, and the automated video editing program for recommending video lists, when executed by the processor, is used to implement the automated video editing method for recommending video lists as described in the first aspect.

[0036] Fourthly, the present invention also provides a medium, which is a computer-readable storage medium, storing an automated video editing program for recommending video lists, which, when executed by a processor, is used to implement the automated video editing method for recommending video lists as described in the first aspect.

[0037] The present invention, by employing the above technical solution, has the following effects:

[0038] This invention acquires and preprocesses video data; based on the preprocessed video data, it generates a recommendation list and basic data for video playlists using a preset large model and prompts; it then generates corresponding text for the playlist data based on the recommendation list and basic data, and selects corresponding video resources and background music; thereby, it performs video synthesis and editing based on the text, video resources, and background music, outputting an automated edited video of the playlist recommendation. This invention can dynamically adjust the material combination logic according to the playlist theme, and deeply mine and structure the playlist recommendation data, achieving highly efficient and large-scale playlist recommendation video generation. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0040] Figure 1 This is a flowchart of the automated video editing method for recommending video lists in this invention.

[0041] Figure 2 This is a functional schematic diagram of the terminal in one implementation of the present invention.

[0042] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0044] Exemplary methods

[0045] With the surge in demand for personalized video playlist recommendations on short video platforms (e.g., film and television commentary compilations, themed mashups), automated video editing technology is gradually shifting from single-scene editing to intelligent production that integrates multiple materials. Current mainstream technologies primarily focus on processing single video footage (e.g., shot segmentation, filter adaptation), but when generating playlist-based recommended content, they still rely on manual material selection, script writing, and manual editing and splicing, resulting in low efficiency and difficulty in scaling. For example, Douyin's "film and television mashup" template only supports matching background music to fixed shot durations and cannot dynamically adjust the material combination logic based on the playlist theme; while automated tools commonly used by broadcasters on some websites (e.g., a certain AI editing tool) can generate basic clips, they lack in-depth mining and structured application of playlist recommendation data (e.g., user review keywords, film emotional tags).

[0046] In summary, existing automated video editing technologies still rely on manual material selection, script writing, and manual editing and splicing when generating recommended content for video lists, which is inefficient and difficult to scale. Therefore, existing technologies need to be improved.

[0047] To address the above technical problems, this invention provides an automated video editing method for movie list recommendation. The method includes: acquiring movie data and preprocessing the movie data; generating a recommendation list and basic data for the movie list based on the preprocessed movie data using a preset large model and prompts; generating text corresponding to the movie list data based on the recommendation list and basic data, and selecting corresponding movie resources and background music; performing video synthesis and editing based on the text, the movie resources, and the background music, and outputting an automated edited video for movie list recommendation. This invention can dynamically adjust the material combination logic according to the movie list theme, deeply mine and structure the movie list recommendation data, and achieve highly efficient and large-scale movie list recommendation video generation.

[0048] like Figure 1 As shown, this embodiment of the invention provides an automated video editing method for recommending video lists, including the following steps:

[0049] Step S100: Obtain video data and preprocess the video data.

[0050] In this embodiment, to achieve the automated video editing method for recommending video lists, an automated video editing system for recommending video lists is proposed, whose technical innovations and advantages are as follows:

[0051] 1) Dynamic list generation and data optimization:

[0052] This embodiment designs a multi-level data cleaning pipeline, combining film review sentiment analysis (e.g., BERT model) with user behavior profiles to dynamically expand film tags (e.g., "suspense index", "tearjerker density", etc.) to improve the theme matching of the film list.

[0053] 2) Build a prompt word optimization engine to automatically generate fine-grained instructions based on the type of movie list (e.g., "Generate a list of mind-bending sci-fi movies that include the concept of 'time loop,' focusing on movies mentioned in user reviews as having 'plot twists'"), guiding the large model to output a structured recommendation list.

[0054] 3) End-to-end high-efficiency processing architecture:

[0055] A distributed task scheduling framework is adopted: integrating web crawlers, large model inference, material download and editing engines, and optimizing resource allocation through priority queues (e.g., urgency of the video list, user level) to reduce end-to-end processing latency.

[0056] In this embodiment, the web crawling technology of the above system is used to obtain film data from third-party media websites (e.g., free movie viewing and material platforms), and these film data (metadata) are automatically organized to obtain the organized film dataset.

[0057] Specifically, in one implementation of this embodiment, step S100 includes the following steps:

[0058] Step S101: Obtain the film data from a preset media website; wherein, the film data includes: film rating, film introduction, actor information, film reviews, and viewing experience;

[0059] Step S102: Clean the video data, filter out videos that do not meet the conditions, and put videos of the same type into the same dataset to obtain the preprocessed video data.

[0060] Since third-party media websites (such as GuaTV and Video) have a large amount of film information, including film ratings, introductions, actor information, etc., and importantly, a large number of viewers' reviews and viewing experiences, this data can serve as the basis for supporting film list recommendations. Therefore, in this embodiment, the web crawler function built into the system is used to obtain film data containing film ratings, film introductions, actor information, film reviews, and viewing experiences from these third-party media websites. After obtaining a large amount of film data, it is processed through the multi-level data cleaning pipeline of the system and enters the corresponding media asset library for data processing. Films that do not meet the conditions (e.g., films that do not conform to the user's viewing habits) are filtered out, and the filtered films of the same type are put into the same dataset to obtain the preprocessed film data.

[0061] like Figure 1 As shown, this embodiment of the invention provides an automated video editing method for recommending video lists, including the following steps:

[0062] Step S200: Based on the preprocessed video data, generate a recommended list and basic data for the video list using a preset large model and prompt words.

[0063] In this embodiment, after the data is cleaned and processed through a multi-level data cleaning pipeline, the processed number of videos is input into a preset large model. Some pre-set prompts are used to enable the large model to generate a recommended list of high-quality videos and basic data based on the video information. The recommended list includes the video list name, and the basic data includes the video, background music, etc.

[0064] Specifically, in one implementation of this embodiment, step S200 includes the following steps:

[0065] Step S201: Input the preprocessed video data into the preset large model;

[0066] Step S202: Generate the prompt words based on preference tags, scene descriptions, number of videos, and theme / type;

[0067] Step S203: Based on the preset large model, use the prompt words to generate a recommendation list containing movie titles and basic data containing movies and background music.

[0068] In this embodiment, film review sentiment analysis (e.g., BERT model) and user behavior profiles are combined to dynamically expand film tags (e.g., "suspense index", "tearjerker density", etc.) to improve the theme matching of the film list, thereby generating a recommendation list containing film list names and basic data containing film and background music.

[0069] As an example, when selecting a large model, generative large language models (e.g., BERT model, GPT series models) or hybrid recommendation frameworks (e.g., SkyReels-V2's diffusion forcing framework) can be used. These large models support text generation and multimodal input. After selecting these large models, they can be trained by fine-tuning.

[0070] During the training phase of the pre-designed large model, based on the organized film data, multi-dimensional feature vectors are constructed using user behavior data and film metadata (e.g., genre, director, actors). Then, contextual information (e.g., viewing time, device type) is integrated to improve the dynamic recommendation effect. Finally, semantic understanding is enhanced by structurally representing the film content (e.g., shot composition, character expressions).

[0071] In this embodiment, during the prompt word design process, firstly, based on the user's historical preferences (preference tags) and the current scenario (scenario description), a basic template for prompt words is set for movies whose quantity does not match the theme / type. Then, reinforcement learning is used to dynamically adjust the prompt word weights to improve recommendation relevance. Finally, combined with the Elasticsearch movie library, the candidate set is retrieved in real time and input into the generation model to ensure the timeliness of the results. This embodiment uses a prompt word optimization engine to automatically generate fine-grained instructions based on the movie list type (e.g., "Generate a list of mind-bending sci-fi movies containing the concept of 'time loop,' focusing on movies mentioned in user reviews as having 'plot twists'"), guiding the large model to output a structured recommendation list.

[0072] In the process of generating a recommendation list containing movie titles and basic data including movies and background music, the model outputs JSON format data containing movie titles, reasons for recommendation, and matching scores (range 0-1) to generate the recommendation list; then, the model automatically extracts movie metadata (e.g., duration, year of release, awards), emotional tags (e.g., "suspenseful and mind-bending", "healing") and similar movie associations to generate basic data.

[0073] like Figure 1 As shown, this embodiment of the invention provides an automated video editing method for recommending video lists, including the following steps:

[0074] Step S300: Generate text corresponding to the movie list data based on the recommended list and basic data, and select the corresponding movie resources and background music.

[0075] In this embodiment, after generating a recommendation list containing movie titles and basic data containing movies and background music using the prompt words, the system automatically generates text corresponding to the movie title data based on the recommendation list and basic data. It also automatically downloads movie resources from the recommendation list and music corresponding to each movie title from the corresponding media resource library to obtain the materials needed for subsequent editing.

[0076] Specifically, in one implementation of this embodiment, step S300 includes the following steps:

[0077] Step S301: Input the recommendation list and the basic data into the preset large model, generate corresponding introductory text containing the movie based on the editing operation of the movie list recommendation, and add a fixed sentence after the introductory text to obtain the text corresponding to the movie list data;

[0078] Step S302: Download the film resources corresponding to each film list from the media resource library corresponding to the basic data, obtain the introductory narration, add other prompt words to the introductory narration, and select film clips as the screen content when the introductory narration is explained based on the preset big model;

[0079] Step S303: Download the music corresponding to each video list from the media resource library corresponding to the basic data, and select the key part of the music based on the preset large model, and use the key part of the music as the background music for the key shot of the recommended video list.

[0080] In this embodiment, after obtaining the movie list data (including a recommended list of movie names and basic data containing movies and background music), the subsequent operations mainly consist of three steps:

[0081] Large Model Text Generation: After inputting the video data from the video list into the large model, tell the large model that the operation being performed is the editing of the video list recommendations, and let the large model generate some introductory narration text containing the videos. After obtaining the introductory narration text, add some fixed sentence patterns and convert it into MP3 format (text-to-speech MP3), which will be added as background music to the finished video.

[0082] Downloading and selecting clips (AI clip selection): After obtaining the list of recommended movies, the corresponding movie resources can be downloaded from the media resource library. The introductory narration generated by the large model in the previous step is also obtained. Some other prompts are added and re-entered into the large model, allowing the large model to select some movie clips as the screen content for the introductory narration.

[0083] Download music and select the climax (AI music climax selection): After obtaining the background music to be used in the video playlist recommendation, download the music from the media asset database and let the large model select the climax of the song, which can be used as the background music when the video playlist recommendation reaches the key shot.

[0084] like Figure 1 As shown, this embodiment of the invention provides an automated video editing method for recommending video lists, including the following steps:

[0085] Step S400: Perform video synthesis and editing based on the text, the video resources, and the background music, and output an automatically edited video recommended by the video list.

[0086] In this embodiment, after automatically generating the script, AI shot selection, and AI music climax excerpts for the video list through a large model, the video can be synthesized and edited based on the script, AI-selected video resources, and background music. During the video synthesis and editing process, resource allocation can be optimized through priority queues (e.g., video list urgency, user level) to reduce end-to-end processing latency.

[0087] Specifically, in one implementation of this embodiment, step S400 includes the following steps:

[0088] Step S401: Based on the text, the video resources, and the background music, perform video synthesis and editing using a preset editing template and a recommended video list structure, and output the automatically edited video recommended by the video list.

[0089] Specifically, in one implementation of this embodiment, step S401 includes: setting a first emoticon at the beginning of the recommended video, and setting a greeting and author introduction information; setting corresponding high-energy scenes in the main body of the recommended video according to the video resources, and configuring the MP3 version of the introductory narration, corresponding text information, and background music for key scenes; setting a second emoticon at the end of the recommended video, and setting a preset ending phrase; and outputting the automatically edited video of the recommended video.

[0090] In this embodiment, based on a large model, a semantic-visual-audio ternary alignment scheme and a dynamic resource orchestration algorithm are used, combined with GPU-accelerated video preprocessing to perform efficient video synthesis and editing, thereby achieving high-quality automated video production and accurately outputting automated edited videos for recommended playlists.

[0091] As an example, in this embodiment, the steps for assembling the montage video are as follows:

[0092] 1. After obtaining all the necessary video materials, synthesize the materials according to the recommended structure of the video list.

[0093] 2. The opening screen is a 5-second emoticon, accompanied by a greeting and an introduction of the author.

[0094] 3. The main part includes some high-energy scenes from the film. The background audio is an MP3 version of the introductory narration provided by the large model. Each film's introduction and narration is about 40 seconds long. When the film introduction begins, some text information will be displayed on the screen for the first few seconds to help the audience obtain information about the film, such as the film title and actors.

[0095] 4. The ending part is an emoji lasting about 3 seconds, and it's also a fairly fixed closing phrase, such as: See you next time!

[0096] 5. Add video camera transitions and fade-in / fade-out effects for background music.

[0097] This embodiment integrates the capabilities of a large model to achieve a series of automated editing functions, including playlist generation, sound equalization, video motion effects, audio track analysis, text-to-speech, AI-generated audio, shot segmentation, automatic subtitles, optical flow analysis, and celebrity recognition.

[0098] This embodiment achieves the following technical effects through the above technical solution:

[0099] This embodiment acquires and preprocesses video data. Based on the preprocessed video data, it generates a recommendation list and basic data for a video playlist using a preset large model and prompts. It then generates corresponding text for the playlist based on the recommendation list and basic data, and selects appropriate video resources and background music. Finally, it performs video synthesis and editing based on the text, video resources, and background music, outputting an automated edited video recommending the playlist. This embodiment can dynamically adjust the material combination logic according to the playlist theme, and deeply mines and structures the playlist recommendation data, achieving highly efficient and large-scale playlist recommendation video generation.

[0100] Exemplary device

[0101] Based on the above embodiments, the present invention also provides an automated video editing system for recommending video lists, comprising:

[0102] The data acquisition module is used to acquire video data and preprocess the video data;

[0103] The recommendation list generation module is used to generate a recommendation list and basic data for the film list based on the preprocessed film data, using a preset large model and prompt words;

[0104] The video material module is used to generate text corresponding to the video list data based on the recommendation list and basic data, and to select the corresponding video resources and background music;

[0105] The compositing and editing module is used to perform video compositing and editing based on the text, the video resources, and the background music, and output an automated edited video recommended by the video list.

[0106] This embodiment achieves the following technical effects through the above technical solution:

[0107] This embodiment acquires and preprocesses video data. Based on the preprocessed video data, it generates a recommendation list and basic data for a video playlist using a preset large model and prompts. It then generates corresponding text for the playlist based on the recommendation list and basic data, and selects appropriate video resources and background music. Finally, it performs video synthesis and editing based on the text, video resources, and background music, outputting an automated edited video recommending the playlist. This embodiment can dynamically adjust the material combination logic according to the playlist theme, and deeply mines and structures the playlist recommendation data, achieving highly efficient and large-scale playlist recommendation video generation.

[0108] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 2 As shown.

[0109] The terminal includes: a processor, a memory, an interface, a display screen, and a communication module connected via a system bus; wherein, the processor of the terminal provides computing and control capabilities; the memory of the terminal includes a storage medium and internal memory; the storage medium stores the operating system and computer programs; the internal memory provides an environment for the operation of the operating system and computer programs in the storage medium; the interface is used to connect to external devices; the display screen is used to display relevant information; and the communication module is used to communicate with a cloud server or other devices.

[0110] When executed by the processor, this computer program is used to implement an automated video editing method for recommending movie lists.

[0111] It will be understood by those skilled in the art that Figure 2 The schematic diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0112] In one embodiment, a terminal is provided, comprising: a processor and a memory, the memory storing an automated video editing program for recommending video lists, which, when executed by the processor, is used to implement the operations of the automated video editing method for recommending video lists as described above.

[0113] In one embodiment, a storage medium is provided, wherein the storage medium stores an automated video editing program for recommending video lists, which, when executed by a processor, is used to implement the operations of the automated video editing method for recommending video lists as described above.

[0114] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, database, or other media used in the embodiments provided by this invention can include both non-volatile and volatile memory.

[0115] In summary, this invention provides an automated video editing method and system for movie list recommendation, comprising: acquiring movie data and preprocessing the movie data; generating a recommendation list and basic data for the movie list based on the preprocessed movie data using a preset large model and prompt words; generating text corresponding to the movie list data based on the recommendation list and basic data, and selecting corresponding movie resources and background music; performing video synthesis and editing based on the text, the movie resources, and the background music, and outputting an automated edited video for movie list recommendation. This invention can dynamically adjust the material combination logic according to the movie list theme, perform in-depth mining and structured application of movie list recommendation data, and achieve highly efficient and large-scale movie list recommendation video generation.

[0116] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. An automated video editing method for recommending video lists, characterized in that, include: Acquire video data and preprocess the video data; Based on the pre-processed video data, a recommended list and basic data for the video list are generated using a preset large model and prompt words; The preprocessed video data is input into the preset large model; The prompt words are generated based on preference tags, scene descriptions, number of videos, and themes / types. Based on the preset large model, the prompt words are used to generate a recommendation list containing movie titles and basic data containing movies and background music; Based on the recommended list and basic data, generate the text corresponding to the movie list data, and select the corresponding movie resources and background music; The recommended list and the basic data are input into the preset large model. Based on the editing operation of the film list recommendation, the corresponding introductory text containing the film is generated. A fixed sentence pattern is added after the introductory text to obtain the text corresponding to the film list data. Download the film resources corresponding to each film list from the media resource library corresponding to the basic data, obtain the introductory narration, add other prompt words to the introductory narration, and select film clips as the screen content when the introductory narration is explained based on the preset big model; Download the music corresponding to each video list from the media resource library corresponding to the basic data, and select the key part of the music based on the preset big model, and use the key part of the music as the background music when the recommended video of the video list reaches the key shot; Based on the text, the video resources, and the background music, the video is synthesized and edited to output an automated edited video recommended by the playlist.

2. The automated video editing method for recommending video lists according to claim 1, characterized in that, Acquire video data and preprocess the video data, including: The film data is obtained from a pre-set media website; wherein, the film data includes: film rating, film introduction, actor information, film reviews, and viewing experience; The video data is cleaned, videos that do not meet the conditions are filtered out, and videos of the same type are put into the same dataset to obtain the preprocessed video data.

3. The automated video editing method for recommending video lists according to claim 1, characterized in that, The automated video editing process, which combines the text, video resources, and background music to output a recommended video list, includes: Based on the text, the video resources, and the background music, a video synthesis and editing process is performed using a preset editing template and a recommended video list structure, outputting the automatically edited video recommended by the video list.

4. The automated video editing method for recommending video lists according to claim 3, characterized in that, The process of performing video synthesis and editing based on the text, the video resources, and the background music using a preset editing template and a recommended video list structure, and outputting the automatically edited video recommended by the video list, includes: Set the first emoji on the opening screen of the recommended videos in the playlist, along with a greeting and author information; Based on the video resources, set the corresponding high-energy scenes of the videos recommended in the video list, and configure the MP3 version of the introductory text, the corresponding text information, and the background music when the key scenes are reached. Set a second emoji at the end of the recommended videos in the playlist, and set a preset ending phrase; Output the automatically edited videos recommended by the aforementioned video list.

5. An automated video editing system for recommending video lists, characterized in that, include: The data acquisition module is used to acquire video data and preprocess the video data; The recommendation list generation module is used to generate a recommendation list and basic data for the film list based on the preprocessed film data, using a preset large model and prompt words; The preprocessed video data is input into the preset large model; The prompt words are generated based on preference tags, scene descriptions, number of videos, and themes / types. Based on the preset large model, the prompt words are used to generate a recommendation list containing movie titles and basic data containing movies and background music; The video material module is used to generate text corresponding to the video list data based on the recommendation list and basic data, and to select the corresponding video resources and background music; The recommended list and the basic data are input into the preset large model. Based on the editing operation of the film list recommendation, the corresponding introductory text containing the film is generated. A fixed sentence pattern is added after the introductory text to obtain the text corresponding to the film list data. Download the film resources corresponding to each film list from the media resource library corresponding to the basic data, obtain the introductory narration, add other prompt words to the introductory narration, and select film clips as the screen content when the introductory narration is explained based on the preset big model; Download the music corresponding to each video list from the media resource library corresponding to the basic data, and select the key part of the music based on the preset big model, and use the key part of the music as the background music when the recommended video of the video list reaches the key shot; The compositing and editing module is used to perform video compositing and editing based on the text, the video resources, and the background music, and output an automated edited video recommended by the video list.

6. A terminal, characterized in that, include: The processor and memory, wherein the memory stores an automated video editing program for recommending video lists, which, when executed by the processor, is used to implement the automated video editing method for recommending video lists as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an automated video editing program for recommending video lists, which, when executed by a processor, is used to implement the automated video editing method for recommending video lists as described in any one of claims 1-4.

Images