Video editing method and device

Abstract

The embodiment of the invention provides a video editing method. The video editing method comprises the following steps: acquiring an original video; determining a plurality of video clips according to the original video; inputting the plurality of video clips and the editing requirements into a pre-trained target visual language model, so that the target visual language model selects at least one target video clip from the plurality of video clips according to the editing requirements; and generating a target video according to the at least one target video clip. According to the technical scheme of the embodiment of the invention, tasks do not need to be created in advance according to editing requirements and algorithm combination does not need to be carried out, the working process of video editing can be simplified, and the working efficiency of video editing can be improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular, to a video editing method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Art

[0002] With the development of digital media technology, the creation and consumption of video content has become an important part of modern society. In recent years, the progress of artificial intelligence technology, especially the breakthrough of deep learning in the field of image and language understanding, has provided new possibilities for the automation and intelligence of video editing. Most of the existing automatic video editing technologies need to convert the requirements into specific task types according to the editing requirements, and then combine multiple algorithms to screen video frames or clips. The whole process is complicated, the degree of automation is not high, and algorithm compatibility problems are prone to occur. Moreover, the existing intelligent editing algorithms using related models need to describe the content of the video to form intermediate content such as text summaries, and there is also the problem of inappropriate clip screening.

[0003] It should be noted that the above content is not necessarily prior art, nor is it intended to limit the scope of patent protection of this application. Summary of the invention

[0004] The embodiments of the present application provide a video editing method, apparatus, computer device, computer-readable storage medium, and computer program product to solve or alleviate one or more of the technical problems raised above.

[0005] One aspect of an embodiment of the present application provides a video editing method, the method comprising: Get the original video; Determine a plurality of video clips according to the original video; Inputting the plurality of video clips and the editing requirements into a pre-trained target visual language model, and selecting a target video clip from the plurality of video clips according to the editing requirements through the target visual language model; and A target video is generated according to the target video segment.

[0006] Optionally, determining a plurality of video segments according to the original video includes: Determining at least one scene segmentation point of the original video; According to the at least one scene segmentation point, the original video is divided into a plurality of video segments.

[0007] Optionally, inputting the plurality of video clips and editing requirements into a pre-trained target visual language model comprises: Determining a text prompt word according to the editing requirement, wherein the text prompt word is used to guide the target visual language model to output according to the editing requirement; The plurality of video clips and the text prompt words are input into the target visual language model.

[0008] Optionally, the plurality of video clips respectively carry different video numbers; Selecting a target video segment from the plurality of video segments according to the editing requirement by using the target visual language model includes: A target video number sequence is returned through the target visual language model, where the target video number sequence includes video numbers of a plurality of the target video clips.

[0009] Optionally, generating a target video according to the target video segment includes: According to the target video number sequence, a plurality of the target video segments are synthesized to obtain the target video.

[0010] Optionally, the target visual language model includes InternVLx or MiniCPM-V.

[0011] Another aspect of an embodiment of the present application provides a video editing device, the device comprising: An acquisition module is used to acquire the original video; A determination module, used to determine a plurality of video clips according to the original video; An input module, configured to input the plurality of video clips and editing requirements into a pre-trained target visual language model, and select a target video clip from the plurality of video clips according to the editing requirements through the target visual language model; and The generating module is used to generate a target video according to the target video segment.

[0012] Another aspect of an embodiment of the present application provides a computer device, including: at least one processor; and a memory communicatively coupled to the at least one processor; Wherein: the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method as described above.

[0013] Another aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method described above is implemented.

[0014] Another aspect of an embodiment of the present application provides a computer program product, including a computer program, which implements the method described above when executed by a processor.

[0015] The embodiment of the present application adopts the above technical solution and may include the following advantages: directly using the visual language model to select the target video segment for generating the target video according to the editing requirements. Thus, there is no need to create tasks and combine algorithms in advance according to the editing requirements, which can simplify the workflow of video editing and improve the work efficiency of video editing. At the same time, the video is directly input to the large visual language model, and the content of the video is not described before the segment is screened. Thus, the problem of information loss due to content description and the problem of inappropriate segment screening due to inaccurate extracted video description can be avoided. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings exemplarily illustrate the embodiments and constitute a part of the specification, and together with the text description of the specification, are used to explain the exemplary implementation of the embodiments. The embodiments shown are for illustrative purposes only and do not limit the scope of the claims. In all drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0017] Figure 1 The following schematically shows an operating environment diagram of a video editing method according to Embodiment 1 of the present application; Figure 2 The flowchart of the video editing method according to the first embodiment of the present application is schematically shown; Figure 3 Schematically shows Figure 2 Flow chart of sub-steps of step S202; Figure 4 Schematically shows Figure 2 Flow chart of sub-steps of step S204; Figure 5 Schematically shows an example diagram of an application of a video editing method according to an embodiment of the present application; Figure 6 A block diagram schematically shows a video editing device according to the second embodiment of the present application; and Figure 7 The hardware architecture diagram of the computer device according to the third embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0019] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0020] In the description of the present application, it should be understood that the numerical labels before the steps do not indicate the order in which the steps are executed, but are only used to facilitate the description of the present application and to distinguish each step, and therefore should not be understood as a limitation on the present application.

[0021] First, the following terms are explained: Large Visual Language Models (VLMs) are a type of deep learning model that can process visual information (such as images) and text information simultaneously. They can understand image and text data at the same time and have the ability to perform tasks such as image / video description, image / video question and answer, and visual dialogue.

[0022] Scene segmentation: It is a concept in computer vision and image processing, which is used to divide a video into different segments, each of which has the same or similar scenes.

[0023] Prompt: Prompt is an input used to guide or instruct the model to generate specific output.

[0024] InternVLx (x is 1.5, 2, ...): It is a large multimodal model released by Shanghai Artificial Intelligence Laboratory. Taking InternVL2 as an example, it is an upgrade based on InternVL1.5, including model sizes of different sizes from 1B to 76B. InternVL2 combines vision and language models and can perform complex cross-modal tasks.

[0025] MiniCPM-V: is a series of multimodal large language models (MLLMs) designed for visual-language understanding. It can accept images and text as input and provide high-quality text output. It can handle multiple data types such as images and videos.

[0026] OpenCV: is an open source computer vision and machine learning software library that provides a programming interface for developing real-time computer vision applications, including but not limited to image processing, feature detection, video analysis, etc.

[0027] Classification: refers to the use of computer vision algorithms to identify and distinguish different objects and scenes in images or videos. It involves training a model to extract features from input image data and classify images into predefined categories based on these features.

[0028] Detection: refers to the process of using computer programs to identify and understand objects, scenes, or activities in images or videos. It includes extracting features from image data and using these features to train models so that different categories or objects can be accurately identified.

[0029] PySceneDetect library: is an open source Python library for automatically detecting scene changes in video content. It helps users segment videos or perform scene analysis by using image processing techniques to identify scene transition points.

[0030] ContentDetector algorithm: is a tool for detecting whether content is generated by artificial intelligence. It uses algorithms and machine learning techniques to analyze text to identify its style, tone, and vocabulary usage patterns.

[0031] Video caption: refers to the text description added to the video content, which is used to provide an overview, explanation or supplementary information about the video content.

[0032] Secondly, in order to facilitate those skilled in the art to understand the technical solutions provided in the embodiments of the present application, the relevant technologies are described below: In the process of automated editing, intelligent editing algorithms that do not use relevant models are mainly based on classic computer vision tasks such as classification and detection. These algorithms usually need to convert the requirements into specific task types based on the editing requirements, and then use the corresponding algorithms to screen the video frames or clips. The whole process relies on the prior task setting, and then the automated editing is achieved by analyzing the video frame by frame or clip by clip. The limitation of this type of algorithm is that it cannot flexibly handle complex scenarios and can only output results according to the predetermined tasks.

[0033] The intelligent editing algorithm using the relevant model needs to segment the video first, then use the corresponding model to generate summaries for each video segment, and finally filter out the clips based on the video summary and editing requirements. However, compared with the original video, the video summary loses information, and the subsequent clip selection based on these video summaries will lead to inaccurate clip selection. In addition, the screening of clips by similarity matching will inevitably involve manually set thresholds. This method of manually setting thresholds is not very universal and will cause problems such as multiple screening and missed screening.

[0034] To this end, an embodiment of the present application provides a video editing technical solution. In this technical solution: (1) The editing requirements are expressed in text form and written into prompt words so that the visual language model can directly understand these requirements and automatically filter video clips. In this way, cumbersome requirements analysis and modeling steps can be avoided, thereby improving editing efficiency; (2) The prompt words contain complete editing requirements, so that the visual language model can output a video that meets the requirements at one time, thereby avoiding the use of multiple different algorithms for combination, simplifying the complexity of algorithm implementation; (3) The video is directly input into a large visual language model, and the content of the video is not described before filtering the clips. In this way, problems such as information loss caused by content description and inappropriate clip screening caused by inaccurate extracted video description can be avoided; (4) Directly understanding the video and filtering the clips according to the editing requirements can avoid the process of matching first and then threshold screening, reduce the problem of multiple screening and missed screening of clips, and improve the versatility of the editing method. See below for details.

[0035] Finally, for ease of understanding, an exemplary operating environment is provided below.

[0036] like Figure 1 As shown, the operating environment diagram includes: service platform 2, network 4, client 6, where: The service platform 2 may be composed of a single or multiple computing devices. The multiple computing devices may include virtualized computing instances. Virtualized computing instances may include virtual machines, such as simulations of computer systems, operating systems, servers, etc. The computing device may load a virtual machine based on a virtual image and / or other data defining specific software (e.g., operating system, dedicated application, server) for simulation. As the demand for different types of processing services changes, different virtual machines may be loaded and / or terminated on one or more computing devices. A hypervisor may be implemented to manage the use of different virtual machines on the same computing device.

[0037] The service platform 2 may be configured to communicate with the client 6 or the like via a network 4. The network 4 includes various network devices, such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and / or the like. The network 4 may include physical links, such as coaxial cable links, twisted pair cable links, optical fiber links, combinations thereof, and the like, or wireless links, such as cellular links, satellite links, Wi-Fi links, and the like.

[0038] The service platform 2 can provide storage, reading, writing, querying, and deleting services, such as running a visual language model or providing video editing services to clients.

[0039] The client 6 may be an electronic device running an operating system such as Windows, Android™ or iOS, such as a smart phone, a tablet device, a laptop computer, a virtual reality device, a gaming device, a set-top box, a vehicle-mounted terminal, or a smart TV. Based on the above operating system, various applications may be run, such as an application for video editing.

[0040] The client 6 can provide / configure a user access page for controlling the service platform 2 or uploading objects (such as videos), etc.

[0041] It should be noted that the above devices are exemplary, and the number and type of devices are adjustable in different scenarios or according to different needs.

[0042] The following describes the technical solution of the present application through multiple embodiments, taking the service platform 2 or the client 6 as the execution subject. It should be noted that these embodiments can be implemented in a variety of different forms and should not be interpreted as being limited to the embodiments described here.

[0043] Embodiment 1 Figure 2 The flowchart of the video editing method according to the first embodiment of the present application is schematically shown.

[0044] like Figure 2 As shown, the video editing method may include steps S200 to S206, wherein: Step S200, obtaining the original video; Step S202, determining a plurality of video clips according to the original video; Step S204, inputting the plurality of video clips and the editing requirements into a pre-trained target visual language model, and selecting a target video clip from the plurality of video clips according to the editing requirements through the target visual language model; and Step S206: Generate a target video according to the target video segment.

[0045] The video editing method provided in this embodiment uses a visual language model to select a target video segment for generating a target video from multiple video segments according to the editing requirements. Therefore, there is no need to create tasks and perform algorithm combination according to the editing requirements in advance, which can simplify the workflow of video editing and improve the work efficiency of video editing.

[0046] The following combination Figure 2 , each step in steps S200~S204 and other optional steps are described in detail.

[0047] Step S200 , get the original video.

[0048] In specific implementation, the original video can be a movie, documentary, animation video, teaching video, etc. on the network platform. These videos can come from video sharing websites or other online resources. According to actual conditions, it can also be other available videos, such as news reports, talk shows, etc., to meet the video needs in different scenarios.

[0049] Step S202 , determining multiple video segments according to the original video.

[0050] In some embodiments, the original video can be divided into multiple video clips by scene segmentation. In other embodiments, the video can be segmented by setting a fixed time interval. According to actual needs, multiple video clips can also be determined by other available methods.

[0051] In this embodiment, segmenting the original video can improve the content relevance of each video segment, making it easier for the subsequent visual language model to extract and understand the video content, thereby improving the efficiency and accuracy of video editing.

[0052] In an optional embodiment, if Figure 3 As shown, step S202 includes: S300: Determine at least one scene segmentation point of the original video.

[0053] S302: Divide the original video into a plurality of video segments according to the at least one scene segmentation point.

[0054] In some embodiments, the scene segmentation point can be determined by the scene segmentation method provided by OpenCV. In other embodiments, the PySceneDetect library can be used to use the ContentDetector algorithm to detect scene changes in the video, and the scene switching point can be identified by calculating the difference in video data between the previous and next frames. According to actual conditions, other feasible methods can also be used to determine the scene segmentation point.

[0055] In this embodiment, the original video is segmented according to the scene to obtain video clips, which can improve the automation of video processing, thereby making the video editing process more efficient. At the same time, the difference in scenes can strengthen the content difference between different video clips, thereby achieving more accurate video editing.

[0056] Step S204 , the multiple video clips and editing requirements are input into a pre-trained target visual language model, and a target video clip is selected from the multiple video clips according to the editing requirements through the target visual language model.

[0057] In specific implementation, the target visual language model can be an open source third-party large visual language model, or a customized and fine-tuned large visual language model. Depending on the actual situation, it can also be other available visual language models.

[0058] In this embodiment, the video is directly input into the large visual language model, and the content of the video is not described before the clips are screened. This can avoid the problem of information loss due to content description and the problem of inappropriate clip screening due to inaccurate extracted video description. At the same time, directly understanding the video and screening the clips according to the editing requirements can avoid the process of matching first and then threshold screening, reduce the problem of multiple screening and missed screening of clips, and improve the versatility of the editing method.

[0059] In an optional embodiment, if Figure 4 As shown, step S204 includes: S400, determining text prompt words according to the editing requirement, wherein the text prompt words are used to guide the target visual language model to output according to the editing requirement.

[0060] S402: Input the multiple video clips and the text prompt words into the target visual language model.

[0061] Text prompt words can be generated by manual input or by adopting a model or algorithm. According to actual needs, text prompt words can also be generated by other means. In certain embodiments, the text prompt words may also include some additional guiding information, such as the style, emotional tone or specific visual element requirements of the video, to enhance the accuracy and relevance of the model output, in addition to including the editing requirements. In other embodiments, the text prompt words may not include other additional information.

[0062] In this embodiment, the editing requirements are expressed in text form and written into the prompt words so that the visual language model can directly understand these requirements and automatically filter video clips. In this way, cumbersome requirements analysis and modeling steps can be avoided, thereby improving editing efficiency. At the same time, the prompt words contain complete editing requirements, so that the visual language model can output a video that meets the requirements at one time, thereby avoiding the use of multiple different algorithms for combination and simplifying the complexity of the algorithm implementation.

[0063] In an optional embodiment, the plurality of video clips respectively carry different video numbers, and step S204 includes: A target video number sequence is returned through the target visual language model, where the target video number sequence includes video numbers of a plurality of the target video clips.

[0064] The video number can be a number, a letter, etc. In this embodiment, the video clips are identified by numbers, which can simplify the processing of the video clips and improve the automation of video editing. Through the numbering, each video clip can be quickly located and retrieved, making the editing process more efficient and ensuring the accuracy and consistency of the editing results.

[0065] Step S206 , generating a target video according to the target video segment.

[0066] In specific implementation, the target video segments can be spliced ​​to obtain the target video. According to actual needs, the target video segments can also be processed for audio and video before splicing, or the generated target video can be processed.

[0067] In some embodiments, the target video can be produced by using a related editing tool (such as FFmpeg). In other embodiments, the target video can also be produced by using a related AI tool (such as OpenCV) or algorithm.

[0068] In this embodiment, the target video is generated by combining the screened target video clips, which can ensure the consistency of the video content with the editing requirements, while improving the overall quality of the video and the viewing experience of the audience.

[0069] In an optional embodiment, step S206 includes: According to the target video number sequence, a plurality of the target video segments are synthesized to obtain the target video.

[0070] Producing target videos according to the target video number sequence can ensure that the playback order of the video clips is consistent with the editing requirements, thereby improving the narrative fluency and logical coherence of the video, while also improving the level of automation in video production.

[0071] In an optional embodiment, the target visual language model includes InternVLx or MiniCPM-V.

[0072] In the specific implementation, InternVL2 or MiniCPM-V is used as the target visual language model, and the text prompt words can be input into the model as the text part and the video clips can be input as the visual part to determine the target video clips that meet the editing requirements.

[0073] Depending on the actual situation, you can also use open source models such as VideoLLaMA (a multimodal large language model consisting of two core components: Vision-Language Branch and Audio-LanguageBranch, which can understand and respond to video- and audio-based instructions to achieve multimodal understanding of video content), Qwen2-VL (Qwen2-VL: a multimodal large language model series that can be used to process images of different resolutions and aspect ratios, and analyze dynamic video content in real time. The main functions include image understanding, video understanding, multi-language support, visual agent operation, dynamic resolution support, and multimodal rotation position embedding (M-ROPE) technology), or other fine-tuned models.

[0074] In order to make this application easier to understand, the following Figure 5 An exemplary application is provided. S11, obtaining an animation video A (ie, the original video) for editing.

[0075] S12, dividing the animation video A into video segments 1, 2, 3, and 4 through scene segmentation.

[0076] S13, determining text prompt words according to the editing requirements of this editing. The text prompt words are as follows: You are a video editor. Now you have cut a long video into 4 segments without overlap or omission, and they are arranged in order. Please understand the content of these 4 video segments, select 3 segments from them, output the sequence numbers of the selected segments, and connect these segments to form the editing results. The editing requirements are as follows: select segments suitable for children to watch, and ensure that the content is appropriate and attractive.

[0077] S14, input the text prompt words and video clips 1, 2, 3, 4 into the large visual language model InternVL2 (i.e., the target visual language model).

[0078] S15, the video number sequence returned by the large visual language model based on the text prompt words is 1, 3, 4 (i.e., the target video number sequence).

[0079] S16, performing audio and video processing on the video clips 1, 3, and 4, and splicing them together to obtain a finished video (ie, a target video) that meets the editing requirements.

[0080] In this exemplary application, the visual language model is used to select video segments 1, 3, and 4 for generating the target video from video segments 1, 2, 3, and 4 according to the editing requirements. As a result, there is no need to create tasks and combine algorithms in advance according to the editing requirements, which can simplify the workflow of video editing and improve the efficiency of video editing. At the same time, the video is directly input to the large visual language model without describing the content of the video before screening the segments. As a result, the problem of information loss caused by content description and the problem of inappropriate segment screening due to inaccurate extracted video description can be avoided.

[0081] Embodiment 2 Figure 6 The block diagram of the video editing device according to the second embodiment of the present application is schematically shown. The device can be divided into one or more program modules, one or more program modules are stored in a storage medium and executed by one or more processors to complete the embodiment of the present application. The program module referred to in the embodiment of the present application refers to a series of computer program instruction segments that can complete specific functions. The following description will specifically introduce the functions of each program module in this embodiment. Figure 6 As shown, the device 1000 may include: an acquisition module 110, a determination module 1200, an input module 1300, and a generation module 1400, wherein: An acquisition module 1100 is used to acquire an original video; A determination module 1200, configured to determine a plurality of video segments according to the original video; An input module 1300 is used to input the plurality of video clips and editing requirements into a pre-trained target visual language model, and select a target video clip from the plurality of video clips according to the editing requirements through the target visual language model; and The generating module 1400 is configured to generate a target video according to the target video segment.

[0082] As an optional embodiment, the determination module 1200 is further configured to: Determining at least one scene segmentation point of the original video; According to the at least one scene segmentation point, the original video is divided into a plurality of video segments.

[0083] As an optional embodiment, the input module 1300 is further used for: Determining a text prompt word according to the editing requirement, wherein the text prompt word is used to guide the target visual language model to output according to the editing requirement; The plurality of video clips and the text prompt words are input into the target visual language model.

[0084] As an optional embodiment, the plurality of video clips respectively carry different video numbers, and the input module 1300 is further configured to: A target video number sequence is returned through the target visual language model, where the target video number sequence includes video numbers of a plurality of the target video clips.

[0085] As an optional embodiment, the generating module 1400 is further configured to: According to the target video number sequence, a plurality of the target video segments are synthesized to obtain the target video.

[0086] As an optional embodiment, the target visual language model includes InternVLx or MiniCPM-V.

[0087] Embodiment 3 Figure 7 The schematic diagram of the hardware architecture of a computer device 10000 suitable for implementing a video editing method according to the third embodiment of the present application is shown schematically. In some embodiments, the computer device 10000 may be a terminal device such as a smart phone, a wearable device, a tablet computer, a personal computer, a vehicle-mounted terminal, a game console, a virtual device, a workbench, a digital assistant, a set-top box, a robot, etc. In other embodiments, the computer device 10000 may be a rack server, a blade server, a tower server, or a cabinet server (including an independent server, or a server cluster composed of multiple servers), etc. Figure 7 As shown, the computer device 10000 includes but is not limited to: a memory 10010, a processor 10020, and a network interface 10030 that can communicate with each other through a system bus. Among them: The memory 10010 includes at least one type of computer-readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 10010 can be an internal storage module of the computer device 10000, such as a hard disk or memory of the computer device 10000. In other embodiments, the memory 10010 can also be an external storage device of the computer device 10000, such as a plug-in hard disk equipped on the computer device 10000, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the memory 10010 can also include both the internal storage module of the computer device 10000 and its external storage device. In this embodiment, the memory 10010 is generally used to store an operating system and various application software installed in the computer device 10000, such as program codes of a video editing method, etc. In addition, the memory 10010 can also be used to temporarily store various data that have been output or are to be output.

[0088] In some embodiments, the processor 10020 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other chips. The processor 10020 is generally used to control the overall operation of the computer device 10000, such as performing control and processing related to data interaction or communication with the computer device 10000. In this embodiment, the processor 10020 is used to run the program code stored in the memory 10010 or process data.

[0089] The network interface 10030 may include a wireless network interface or a wired network interface, and the network interface 10030 is generally used to establish a communication link between the computer device 10000 and other computer devices. For example, the network interface 10030 is used to connect the computer device 10000 to an external terminal through a network, and to establish a data transmission channel and a communication link between the computer device 10000 and the external terminal. The network may be a wireless or wired network such as an intranet, the Internet, the Global System of Mobile communication (GSM), Wideband Code Division Multiple Access (WCDMA), 4G network, 5G network, Bluetooth, Wi-Fi, etc.

[0090] It should be pointed out that Figure 7 Only a computer device having components 10010 - 10030 is shown, but it should be understood that implementation of all of the components shown is not a requirement, and more or fewer components may alternatively be implemented.

[0091] In this embodiment, the video editing method stored in the memory 10010 can also be divided into one or more program modules and executed by one or more processors (such as processor 10020) to complete the embodiment of the present application.

[0092] Embodiment 4 An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, wherein when the computer program is executed by a processor, the steps of the video editing method in the embodiment are implemented.

[0093] In this embodiment, the computer-readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the computer-readable storage medium can be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium can also be an external storage device of a computer device, such as a plug-in hard disk equipped on the computer device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the computer-readable storage medium can also include both the internal storage unit of the computer device and its external storage device. In this embodiment, the computer-readable storage medium is generally used to store an operating system and various application software installed on the computer device, such as the program code of the video editing method in the embodiment. In addition, the computer-readable storage medium can also be used to temporarily store various types of data that have been output or are to be output.

[0094] Embodiment 5 An embodiment of the present application also provides a computer program product, including a computer program, which implements the method in the above embodiment when executed by a processor.

[0095] Obviously, those skilled in the art should understand that the modules or steps of the above-mentioned embodiments of the present application can be implemented by general-purpose computer devices, they can be concentrated on a single computer device, or distributed on a network composed of multiple computer devices, optionally, they can be implemented by executable program codes of computer devices, so that they can be stored in a storage device and executed by the computer device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0096] It should be noted that the above are only preferred embodiments of the present application, and the patent protection scope of the present application is not limited thereto. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A video editing method, characterized in that: The method comprises: Get the original video; Determine a plurality of video clips according to the original video; Inputting the plurality of video clips and the editing requirements into a pre-trained target visual language model, and selecting a target video clip from the plurality of video clips according to the editing requirements through the target visual language model; and A target video is generated according to the target video segment.

2. The method according to claim 1, characterized in that Determine a plurality of video segments according to the original video, including: Determining at least one scene segmentation point of the original video; According to the at least one scene segmentation point, the original video is divided into a plurality of video segments.

3. The method according to claim 1, characterized in that Inputting the plurality of video clips and editing requirements into a pre-trained target visual language model includes: Determining a text prompt word according to the editing requirement, wherein the text prompt word is used to guide the target visual language model to output according to the editing requirement; The plurality of video clips and the text prompt words are input into the target visual language model.

4. The method according to claim 1, characterized in that: The plurality of video clips respectively carry different video numbers; Selecting a target video segment from the plurality of video segments according to the editing requirement by using the target visual language model includes: A target video number sequence is returned through the target visual language model, where the target video number sequence includes video numbers of a plurality of the target video clips.

5. The method according to claim 4, characterized in that Generating a target video according to the target video segment includes: According to the target video number sequence, a plurality of the target video segments are synthesized to obtain the target video.

6. The method according to any one of claims 1 to 5, characterized in that: The target visual language model includes InternVLx or MiniCPM-V.

7. A video editing device, characterized in that: The device comprises: An acquisition module is used to acquire the original video; A determination module, used to determine a plurality of video clips according to the original video; An input module, configured to input the plurality of video clips and editing requirements into a pre-trained target visual language model, and select a target video clip from the plurality of video clips according to the editing requirements through the target visual language model; and The generating module is used to generate a target video according to the target video segment.

8. A computer device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein: The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to claims 1 to 6 are implemented.

Images