Video generation method and device, storage medium and electronic equipment

Abstract

The invention discloses a video generation method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining a plurality of game video clips of a to-be-promoted game, a game description text, and a user portrait of a target user of the to-be-promoted game, inputting the game description text and the user portrait into a large language model, and obtaining an advertisement copywriting of the to-be-promoted game; and respectively extracting key frames in each game video clip, determining the game video clip matched with the advertisement copywriting as a target video clip according to the key frames in each game video clip, and synthesizing the target video clip and the advertisement copywriting to obtain a promotion video of the to-be-promoted game. In the method, on the basis of the game video clip input by the user, the game description text and the user portrait of the target user, the promotion video of the to-be-promoted game can be automatically generated, manual operation in game advertisement putting is reduced, and the creation efficiency of the game promotion video is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a video generation method, device, storage medium and electronic device. Background Art

[0002] When placing game ads, professional designers are usually required to create promotional videos or images based on the characteristics of the game to be promoted as game ads. The game ads are then placed on various media platforms to attract users of the media platforms to become target users of the game to be promoted.

[0003] Because promotional videos can contain dynamic features such as the game's dynamic images, visual effects, and gameplay demonstrations, they are more likely to attract users, thereby achieving a higher conversion rate and better delivery results. However, compared to promotional images, the creation of promotional videos is more complicated and requires more work from professional designers. The efficiency of game advertising creation is low, which affects the timeliness of advertising delivery and is not conducive to game promotion.

[0004] Therefore, the present specification provides a video generation method. Summary of the invention

[0005] The present specification provides a video generation method, device, storage medium and electronic device to at least partially solve the above-mentioned problems existing in the prior art.

[0006] This manual adopts the following technical solutions:

[0007] This specification provides a video generation method, including:

[0008] Obtaining several game video clips, game description texts, and user portraits of target users of the game to be promoted;

[0009] Input the game description text and the user portrait into a large language model to obtain the advertising copy of the game to be promoted;

[0010] Extracting key frames from each game video clip respectively, and determining a game video clip matching the advertisement copy as a target video clip according to the key frames from each game video clip;

[0011] The target video clip and the advertising copy are synthesized to obtain a promotional video of the game to be promoted.

[0012] Optionally, extracting key frames from each game video clip separately includes:

[0013] For each game video clip, determining each frame of video image contained in the game video clip;

[0014] Clustering the video images according to the image features of the video images to obtain clusters;

[0015] For each cluster, a representative image of the cluster is determined in each frame of video image included in the cluster;

[0016] The representative images of the clusters are used as key frames corresponding to the game video clip.

[0017] Optionally, determining a representative image of the cluster among the frames of video images included in the cluster specifically includes:

[0018] Calculate the image information entropy of each frame of video image contained in the cluster;

[0019] In the cluster, the video image corresponding to the maximum image information entropy is determined as the representative image of the cluster.

[0020] Optionally, determining a game video clip matching the advertisement copy as a target video clip according to the key frames in each game video clip specifically includes:

[0021] For each game video clip, determine the video clip features of the game video clip according to the key frames corresponding to the game video clip; determine the copy features of the advertisement copy;

[0022] Determining the similarity between the video clip features of the game video clip and the text features;

[0023] According to the similarity between the video clip features of the game video clips and the text features, a target video clip is determined in the game video clips.

[0024] Optionally, there are multiple target video clips;

[0025] The target video clip and the advertising copy are synthesized to obtain a promotional video of the game to be promoted, specifically including:

[0026] According to the similarity between the video clip features of each target video clip and the copywriting features, the target video clips are sorted to obtain a video sequence;

[0027] The video sequence is synthesized with the advertising copy to obtain a promotional video of the game to be promoted.

[0028] Optionally, synthesizing the target video clip and the advertisement copy to obtain a promotional video of the game to be promoted specifically includes:

[0029] The background music is determined, and the background music, the target video clip and the advertising copy are synthesized to obtain a promotional video of the game to be promoted.

[0030] Optionally, background music is determined, specifically including:

[0031] Determine each pre-stored candidate music and the music emotion characteristics of each candidate music;

[0032] Inputting the advertisement copy into a text sentiment analysis model to obtain the sentiment polarity and sentiment category of the advertisement copy;

[0033] Determining text sentiment characteristics according to the sentiment polarity and sentiment category of the advertisement copy;

[0034] According to the similarity between the music emotion feature of each candidate music and the text emotion feature, the candidate music that matches the emotion of the advertisement copy is determined as the background music.

[0035] This specification provides a video generation device, the device comprising:

[0036] An acquisition module, which acquires several game video clips, game description texts, and user portraits of target users of the game to be promoted;

[0037] A copywriting generation module, which inputs the game description text and the user portrait into a large language model to obtain the advertising copy of the game to be promoted;

[0038] A video matching module extracts key frames from each game video clip, and determines a game video clip that matches the advertisement copy as a target video clip based on the key frames in each game video clip;

[0039] The video generation module synthesizes the target video clip and the advertisement text to obtain a promotional video of the game to be promoted. This specification provides a computer-readable storage medium, the storage medium stores a computer program, and the computer program implements the above-mentioned video generation method when executed by a processor.

[0040] This specification provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned video generation method when executing the program.

[0041] At least one of the above technical solutions adopted in this specification can achieve the following beneficial effects:

[0042] In the video generation method provided in this specification, several game video clips, game description texts, and user portraits of target users of the game to be promoted are obtained, and the game description texts and user portraits are input into a large language model to obtain the advertising copy of the game to be promoted. The key frames in each game video clip are extracted respectively, and the game video clips matching the advertising copy are determined according to the key frames in each game video clip as the target video clip, and the target video clip and the advertising copy are synthesized to obtain the game advertisement. In this method, based on the game video clips, game description texts, and user portraits of target users input by the user, the promotional video of the game to be promoted can be automatically generated, which reduces the manual operation in the game advertising and improves the creation efficiency of the game promotion video. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings described herein are used to provide a further understanding of this specification and constitute a part of this specification. The illustrative embodiments and descriptions of this specification are used to explain this specification and do not constitute an improper limitation on this specification. In the drawings:

[0044] Figure 1 A schematic diagram of a video generation method in this specification;

[0045] Figure 2 A schematic diagram of a video generation method and device provided in this specification;

[0046] Figure 3 The corresponding Figure 1 Schematic diagram of electronic equipment. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of this specification more clear, the technical solutions of this specification will be clearly and completely described below in combination with the specific embodiments of this specification and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0048] The technical solutions provided by the embodiments of this specification are described in detail below in conjunction with the accompanying drawings.

[0049] Figure 1 The following is a flow chart of a video generation method in this specification, which specifically includes the following steps:

[0050] S100: Acquire several game video clips, game description texts, and user portraits of target users of the game to be promoted.

[0051] All steps in the video generation method provided in this specification can be implemented by any electronic device with computing functions, such as a terminal, a server, etc. For ease of description, the video generation method provided in this specification is described below with the server as the execution subject.

[0052] When promoting a game, the game is generally in a state of development completion, and at this time the server can obtain game video clips such as game screen recordings and game replays during the game process of players or testers. Because the game video clips are the real game scenes of the players, the game video clips are used as a material for generating game promotion videos in this manual, which can provide visual information such as game screen style and special effect style for video generation, as well as operation information such as game play and player operation, so that the generated game advertisement corresponds to the game content of the game to be promoted.

[0053] In this specification, the server needs to first obtain the generated material of the promotional video of the game to be promoted input by the user. The material includes several game video clips, game description text and user portraits of target users.

[0054] Each game video clip can provide video material for the generation of a promotional video for the game to be promoted, and the game description text and the user portrait of the target user can provide text material for the promotional video for the game to be promoted, so that the server can obtain the promotional video for the game to be promoted based on the video material and the text material.

[0055] A game video clip can show a game scene of a player's game session. Multiple game video clips can reflect more diverse game scenes and provide more comprehensive visual information for the generation of game promotion videos.

[0056] S102: Input the game description text and the user portrait into a large language model to obtain the advertising copy of the game to be promoted.

[0057] The game description text is the original text related to the game that can be directly obtained by the user and determined during the game development stage. The user does not need to work for secondary creation, thus reducing the user's workload. This manual does not limit the specific content of the game description text, which can be a game background introduction, game play, game scene settings, etc.

[0058] Because the original purpose of the game description text in this manual is to guide the development of the game, the content of the game description text is often detailed and specific, and there is no secondary creation based on the user portrait of the target user, so it is not suitable to be used directly as advertising copy.

[0059] Therefore, in order to obtain advertising copy that matches the user portrait of the target user, the server inputs the game description text and the user portrait of the target user into the big language model, and generates advertising copy for the game to be promoted through the big language model.

[0060] The purpose of advertising is to gain more game users. In the video generation method provided in this manual, the game description text and the user portrait of the target user are used as the input of the large language model, so that the advertising copy can meet the user needs of the target user, so that after the game advertisement is released, a higher user conversion rate can be obtained and a better delivery effect can be achieved.

[0061] In addition, because the user does not need to perform secondary creation when determining the game description text in this manual, even ordinary users who do not have video creation skills can generate promotional videos for the games to be promoted through the video generation method in this manual. Therefore, the video generation method in this manual improves the creation efficiency of game advertisements while also lowering the creation threshold of game advertisements.

[0062] S104: extracting key frames from each game video clip respectively, and determining a game video clip matching the advertisement copy as a target video clip according to the key frames in each game video clip.

[0063] In the process of determining the advertising copy, the visual information in the game video clips is not taken into consideration. In each game video, there may be game videos whose visual semantics do not match the text semantics of the advertising copy. If all game videos are directly synthesized with the advertising copy, the advertising copy of the advertising video may not correspond to the video scene, affecting the quality of the generated game ads.

[0064] Therefore, in order to obtain high-quality game advertisements, the server extracts key frames from each game video clip respectively, and determines the game video clip that matches the advertisement copy as the target video clip based on the key frames in each game video clip.

[0065] In one or more embodiments of the present specification, the server may determine the key frame in the following manner.

[0066] First, the server determines, for each game video clip, each frame of video images contained in the game video clip, and determines image features of each frame of video images.

[0067] This specification does not limit the specific method for determining image features. Image features can be obtained by extracting features from each frame of video images through a convolutional neural network, or the directional gradient histogram, color histogram, texture features, etc. of each frame of video images can be used as image features.

[0068] Then, for each cluster, the server determines the representative image of the cluster in each frame of video image contained in the cluster. Finally, the server uses the representative image of each cluster as the key frame corresponding to the game video clip.

[0069] There are many ways to determine the representative image. For example, the server can use the video image corresponding to the cluster center of each cluster as the representative image of the cluster. Or, because the video images included in a cluster are already video images with relatively similar image features, the server can randomly select a video image from the video images included in the cluster as the representative image of the cluster.

[0070] A key frame of a game video can represent the visual information of the game video, so the server screens the game videos by key frames. The server determines the game video segments that match the advertisement copy as the target video segments based on the key frames in each game video segment.

[0071] First, for each game video clip, the server determines the video clip features of the game video clip and the text features of the advertisement text according to the key frames corresponding to the game video clip. Then, the server calculates the similarity between the video clip features of the game video clip and the text features.

[0072] Specifically, the server inputs the key frame corresponding to the game video clip into the image encoder to obtain the video clip features of the game video clip. The advertisement copy is input into the text encoder to obtain the copy features. This specification does not limit the specific network structure of the image encoder and the text encoder, but the image encoder and the text encoder need to be trained by comparison to align the feature space.

[0073] Finally, the server determines the target video clip in each game video clip according to the similarity between the video clip features and the text features of each game video clip.

[0074] This specification does not limit the number of target video clips. The server can select the game videos corresponding to the video clip features with the highest similarity to the copywriting features as the target video clips. The game videos can also be sorted according to the similarity and a specified number of game videos can be selected as the target videos.

[0075] S106: Synthesize the target video clip and the advertisement text to obtain a game advertisement.

[0076] After determining the advertisement text and the target video, the server adds the advertisement text to the target video clip, synthesizes the target video clip with the advertisement text, and obtains the game advertisement.

[0077] This specification does not limit the specific method used for the synthesis operation. In one embodiment, the server can display the advertisement text in the target video segment according to a preset display position. Alternatively, the server can call any video editing tool to perform the synthesis operation.

[0078] In the video generation method provided in this specification, several game video clips, game description texts, and user portraits of target users of the game to be promoted are obtained, and the game description texts and user portraits are input into a large language model to obtain the advertising copy of the game to be promoted. The key frames in each game video clip are extracted respectively, and the game video clips matching the advertising copy are determined according to the key frames in each game video clip as the target video clip, and the target video clip and the advertising copy are synthesized to obtain the promotional video of the game to be promoted. In this method, based on the game video clips, game description texts, and user portraits of target users input by the user, the promotional video of the game to be promoted can be automatically generated, which reduces the manual operation in the game advertising and improves the creation efficiency of the game promotion video.

[0079] At the same time, since the user portrait of the target user is used as the input of the large language model in the process of determining the advertising copy, the determined advertising copy is more in line with the user needs of the user. Therefore, after the promotional video obtained by this method is released to the media platform, the user conversion rate is higher, more game users are attracted, and better delivery effect is achieved.

[0080] In the above step S104 , in order to determine a more accurate key frame in each cluster, in one or more embodiments of the present specification, the server may determine a representative image of each cluster according to image information entropy.

[0081] Specifically, the server calculates the image information entropy of each frame of video images contained in each cluster, and determines the video image corresponding to the maximum image information entropy in the cluster as the representative image of the cluster.

[0082] First, for each video image included in the cluster, the server converts the video image into a grayscale image and determines the pixel value of each pixel in the grayscale image.

[0083] Then, the server counts the number of times each pixel value appears in the grayscale image. The server determines the probability of occurrence of each type of pixel value based on the ratio of the number of times the pixel value appears to the total number of pixel values ​​in the grayscale image.

[0084] Finally, the server determines the image information entropy of the video image based on the occurrence probability of each type of pixel value.

[0085] The calculation formula of image information entropy is as follows:

[0086]

[0087] Among them, H represents the image information entropy of the video image, P i represents the probability of occurrence of the i-th type of pixel value in the video image, and N represents the total number of different types of pixel values ​​contained in the video image.

[0088] Image information entropy can reflect the richness of information contained in an image. The representative image determined by this embodiment is a video image containing the richest image information in a cluster, which can more completely express the visual information of a game video clip, thereby generating more accurate game advertisements in subsequent steps.

[0089] In one or more embodiments of the present specification, when the server determines in the above step S104 that there are multiple target video segments.

[0090] In the above step S106, the server can sort the target video clips according to the similarity between the video clip features and the copy features of each target video clip to obtain a video sequence. Then, the video sequence is synthesized with the advertisement copy to obtain the game advertisement.

[0091] Because after being sorted according to similarity, the differences between adjacent target video clips are smaller, and the transition between target video clips is more reasonable, so that a game advertisement with a smoother viewing experience can be obtained.

[0092] In the above step S106, in order to obtain a higher quality game advertisement, in one or more embodiments of this specification, the server may add background music to the game advertisement, and synthesize the background music, the target video clip and the advertisement text to obtain the game advertisement.

[0093] This specification does not limit the method of determining the background music, and the server may pre-set a fixed background music. Alternatively, the server may pre-store multiple candidate music and determine the background music according to the user's selection.

[0094] Alternatively, the server may perform sentiment analysis on each candidate music stored in advance, determine the sentiment characteristics of each candidate music, and store each candidate music in correspondence with the sentiment characteristics of each candidate music, so as to be able to screen according to the sentiment characteristics and determine the candidate music that matches the sentiment of the advertisement copy as background music, thereby obtaining a higher quality game advertisement.

[0095] The server can determine the numerical expression of each candidate music according to the music features such as rhythm, pitch, loudness, etc. of each candidate music, and obtain the feature vector of each candidate music. Then, the feature vector of each candidate music is input into the pre-trained music sentiment analysis model to determine the sentiment polarity and sentiment category of each candidate music, and the sentiment polarity and sentiment category of each candidate music are vectorized to obtain the sentiment feature of each candidate music.

[0096] This specification does not limit the specific classification of emotion polarity and emotion category. In one example, emotion polarity can be divided into positive, neutral and negative, and emotion category can be divided into joy, sadness, anger, surprise, etc. The music emotion analysis model in this embodiment can be trained by using sample music with emotion polarity and emotion category marked as training data.

[0097] Specifically, the server determines each candidate music and the music emotional features of each candidate music that are pre-stored. The advertisement copy is input into the text emotional analysis model to obtain the emotional polarity and emotional category of the advertisement copy. The emotional polarity and emotional category of the advertisement copy are vectorized to obtain the text emotional features of the advertisement copy.

[0098] The text sentiment analysis model can be trained using a large amount of sample text labeled with sentiment polarity and sentiment category as training data.

[0099] Then, the server determines the similarity between the music emotion characteristics of each candidate music and the text emotion characteristics of the advertisement copy, and selects the candidate music corresponding to the greatest similarity as the background music according to the similarities. The background music is the candidate music that matches the emotion of the advertisement copy.

[0100] The above is the video generation method provided in this specification. Based on the same idea, this specification also provides a corresponding video generation device, such as Figure 2 shown.

[0101] Figure 2 A schematic diagram of a video generation device provided in this specification specifically includes:

[0102] The acquisition module 200 is used to acquire several game video clips, game description texts, and user portraits of target users of the game to be promoted;

[0103] A copywriting generation module 204 is used to input the game description text and the user portrait into a large language model to obtain an advertising copy for the game to be promoted;

[0104] The video matching module 206 is used to extract key frames from each game video clip, and determine the game video clip that matches the advertisement copy as the target video clip according to the key frames in each game video clip;

[0105] The video generation module 208 is used to synthesize the target video clip and the advertisement text to obtain a promotional video of the game to be promoted.

[0106] Optionally, the video matching module 206 is specifically used to determine, for each game video clip, each frame of video image contained in the game video clip, cluster the each frame of video image according to image features of the each frame of video image to obtain each cluster cluster, and for each cluster cluster, determine a representative image of the cluster among the each frame of video image contained in the cluster cluster, and use the representative image of each cluster cluster as a key frame corresponding to the game video clip.

[0107] Optionally, the video matching module 206 is specifically configured to calculate the image information entropy of each frame of video image included in the cluster, and determine the video image corresponding to the maximum image information entropy in the cluster as the representative image of the cluster.

[0108] Optionally, the video matching module 206 is specifically used to determine, for each game video clip, the video clip features of the game video clip according to the key frames corresponding to the game video clip; determine the text features of the advertising copy, determine the similarity between the video clip features of the game video clip and the text features, and determine the target video clip in each game video clip according to the similarity between the video clip features of each game video clip and the text features.

[0109] Optionally, there are multiple target video clips, and the video generation module 208 is specifically used to sort the target video clips according to the similarity between the video clip features of each target video clip and the copy features to obtain a video sequence, and synthesize the video sequence with the advertising copy to obtain a promotional video for the game to be promoted.

[0110] Optionally, the video generation module 208 is specifically used to determine background music, and synthesize the background music, the target video clip and the advertising copy to obtain a promotional video of the game to be promoted.

[0111] Optionally, determine each pre-stored candidate music and the music emotion characteristics of each candidate music, input the advertising copy into a text emotion analysis model to obtain the emotion polarity and emotion category of the advertising copy, determine the text emotion characteristics based on the emotion polarity and emotion category of the advertising copy, and determine the candidate music that matches the emotion of the advertising copy as background music based on the similarity between the music emotion characteristics of each candidate music and the text emotion characteristics.

[0112] This specification also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 1 Provided video generation method.

[0113] This manual also provides Figure 3 The schematic structure diagram of the electronic device shown in FIG. Figure 3 As mentioned above, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 1 Of course, in addition to the software implementation, this specification does not exclude other implementations, such as logic devices or a combination of software and hardware, etc., that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0114] For the improvement of a technology, it can be clearly distinguished whether it is a hardware improvement (for example, improvement of the circuit structure of diodes, transistors, switches, etc.) or a software improvement (improvement of the method flow). However, with the development of technology, many improvements of the method flow today can be regarded as direct improvements of the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that the improvement of a method flow cannot be implemented with a hardware entity module. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming themselves, without having to ask chip manufacturers to design and make dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.

[0115] The controller can be implemented in any appropriate manner, for example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) that can be executed by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art also know that in addition to implementing the controller in a purely computer-readable program code manner, the controller can be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, this controller can be considered as a hardware component, and the devices included therein for implementing various functions can also be regarded as structures within the hardware component. Or even, the devices for implementing various functions can be regarded as both software modules for implementing the method and structures within the hardware component.

[0116] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0117] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0118] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0119] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0120] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0121] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0122] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0123] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0124] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0125] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0126] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems or computer program products. Therefore, this specification may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0127] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0128] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0129] The above description is only an embodiment of this specification and is not intended to limit this specification. For those skilled in the art, this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification should be included in the scope of the claims of this application.

Claims

1. A video generation method, characterized in that: include: Obtaining several game video clips, game description texts, and user portraits of target users of the game to be promoted; Input the game description text and the user portrait into a large language model to obtain the advertising copy of the game to be promoted; Extracting key frames from each game video clip respectively, and determining a game video clip matching the advertisement copy as a target video clip according to the key frames from each game video clip; The target video clip and the advertising copy are synthesized to obtain a promotional video of the game to be promoted.

2. The method according to claim 1, characterized in that Extract the key frames from each game video clip separately, including: For each game video clip, determining each frame of video image contained in the game video clip; Clustering the video images according to the image features of the video images to obtain clusters; For each cluster, a representative image of the cluster is determined in each frame of video image included in the cluster; The representative images of the clusters are used as key frames corresponding to the game video clip.

3. The method according to claim 2, characterized in that Determining a representative image of the cluster in each frame of video images included in the cluster specifically includes: Calculate the image information entropy of each frame of video image contained in the cluster; In the cluster, the video image corresponding to the maximum image information entropy is determined as the representative image of the cluster.

4. The method according to claim 1, characterized in that Determining, according to the key frames in each of the game video clips, a game video clip matching the advertisement copy as a target video clip, specifically includes: For each game video clip, determine the video clip features of the game video clip according to the key frames corresponding to the game video clip; determine the copy features of the advertisement copy; Determining the similarity between the video clip features of the game video clip and the text features; According to the similarity between the video clip features of the game video clips and the text features, a target video clip is determined in the game video clips.

5. The method according to claim 4, characterized in that The number of the target video clips is multiple; The target video clip and the advertising copy are synthesized to obtain a promotional video of the game to be promoted, specifically including: According to the similarity between the video clip features of each target video clip and the copywriting features, the target video clips are sorted to obtain a video sequence; The video sequence is synthesized with the advertising copy to obtain a promotional video of the game to be promoted.

6. The method according to claim 1, characterized in that The target video clip and the advertising copy are synthesized to obtain a promotional video of the game to be promoted, specifically including: Determine the background music; The background music, the target video clip and the advertising copy are synthesized to obtain a promotional video of the game to be promoted.

7. The method according to claim 6, characterized in that Determine the background music, including: Determine each pre-stored candidate music and the music emotion characteristics of each candidate music; Inputting the advertisement copy into a text sentiment analysis model to obtain the sentiment polarity and sentiment category of the advertisement copy; Determining text sentiment characteristics according to the sentiment polarity and sentiment category of the advertisement copy; According to the similarity between the music emotion feature of each candidate music and the text emotion feature, the candidate music that matches the emotion of the advertisement copy is determined as the background music.

8. A video generating device, characterized in that: include: An acquisition module, which acquires several game video clips, game description texts, and user portraits of target users of the game to be promoted; A copywriting generation module, which inputs the game description text and the user portrait into a large language model to obtain the advertising copy of the game to be promoted; A video matching module extracts key frames from each game video clip, and determines a game video clip that matches the advertisement copy as a target video clip based on the key frames in each game video clip; The video generation module synthesizes the target video clip and the advertising copy to obtain a promotional video of the game to be promoted.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

Images