Video editing method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses a video editing method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the attribute analysis of a scene in a sample video frame, and obtaining the attribute information of the scene; wherein the scene attribute information at least comprises spatial information and semantic information corresponding to a scene in the sample video frame; determining a target editing effect; wherein the target editing effect comprises at least one of the following: editing a scene mirror and editing a scene shape; generating a target video frame based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame. And diversified video editing can be realized.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of computer technology, and in particular to a video editing method, device, electronic device, and storage medium. Background Art

[0002] Existing video editing technologies can usually achieve video cutting, splicing, tuning, speed change, and adding special effects (such as subtitles, stickers, filters, transitions, etc.). In related technologies, the video editing method is relatively limited. Summary of the invention

[0003] The embodiments of the present disclosure provide a video editing method, device, electronic device and storage medium, which can realize diversified video editing.

[0004] In a first aspect, an embodiment of the present disclosure provides a video editing method, comprising:

[0005] Performing attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame;

[0006] Determine a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape;

[0007] A target video frame is generated based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

[0008] In a second aspect, the present disclosure also provides a video editing device, including:

[0009] An analysis module, used to perform attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame;

[0010] An effect determination module, used to determine a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape;

[0011] An editing module is used to generate a target video frame based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

[0012] In a third aspect, an embodiment of the present disclosure further provides an electronic device, the electronic device comprising:

[0013] one or more processors;

[0014] a storage device for storing one or more programs,

[0015] When the one or more programs are executed by the one or more processors, the one or more processors implement the video editing method as described in any of the embodiments of the present disclosure.

[0016] In a fourth aspect, the embodiments of the present disclosure further provide a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to execute the video editing method as described in any one of the embodiments of the present disclosure.

[0017] The technical solution of the embodiment of the present disclosure performs attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame; determines a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape; generates a target video frame based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

[0018] Since the neural radiation field model is constructed based on sample video frames, it can enable the neural radiation field model to have the ability to present the scenes in the sample video frames at any angle and distance. By performing attribute analysis on the scenes in the sample video frames, at least the spatial, semantic and other scene attribute information corresponding to the scenes can be obtained, so that scene understanding can be achieved. Based on the scene attribute information, the neural radiation field model can perform camera movement and / or shape editing on the scenes in the sample video frames according to the target editing effect to generate target video frames, thereby achieving diversified video editing. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0020] Figure 1 A flowchart of a video editing method provided by an embodiment of the present disclosure;

[0021] Figure 2 A schematic diagram of a process of scene coordinate reconstruction in a video editing method provided by an embodiment of the present disclosure;

[0022] Figure 3 A flowchart of a video editing method provided by an embodiment of the present disclosure;

[0023] Figure 4 A schematic diagram of changing scene shooting and camera movement in a video editing method provided by an embodiment of the present disclosure;

[0024] Figure 5 A schematic diagram of changing a scene shape in a video editing method provided by an embodiment of the present disclosure;

[0025] Figure 6 A flowchart of a video editing method provided by an embodiment of the present disclosure;

[0026] Figure 7 A flowchart of a video editing method provided by an embodiment of the present disclosure;

[0027] Figure 8 A schematic diagram of a new target video frame of a video editing method provided by an embodiment of the present disclosure;

[0028] Fig. 9 A schematic diagram of the structure of a video editing device provided by an embodiment of the present disclosure;

[0029] Fig.10 A schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0030] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0031] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0032] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0033] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0034] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0035] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0036] It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

[0037] Figure 1 This is a flow chart of a video editing method provided by an embodiment of the present disclosure. The embodiment of the present disclosure is applicable to video editing situations, such as changing the camera movement when shooting a scene in a video, and / or changing the three-dimensional shape of a scene, etc. The method can be executed by a video editing device, which can be implemented in the form of software and / or hardware, and can be configured in an electronic device, such as a computer.

[0038] like Figure 1 As shown, the video editing method provided in this embodiment may include:

[0039] S110: Analyze the attributes of the scenes in the sample video frames to obtain scene attribute information.

[0040] In the disclosed embodiment, video frames of the video to be edited can be acquired through existing technologies, and sample video frames can be determined based on these video frames. For example, a frame can be extracted from the video frames as a sample video frame every predetermined number of frames or predetermined time.

[0041] In some optional implementations, before performing attribute analysis on the scene in the sample video frame, it may also include: selecting the sample video frame based on at least one of the following methods: selecting the sample video frame based on the co-viewing relationship between the video frames; selecting the sample video frame based on the content of each video frame.

[0042] In these optional implementations, a pose solving method, such as Structure from Motion (SfM), can be used to determine the camera position of each video frame. When the change in camera position is less than a preset distance, it can be considered that there is a common view relationship. Then, one frame can be selected from each video frame with a common view relationship as a sample video frame. Thereby, there is sufficient camera motion, that is, a change in viewing angle, between the selected sample video frames. By selecting sample video frames based on the common view relationship between the video frames, it is beneficial for each sample video frame to carry more complete and comprehensive scene attribute information.

[0043] The selection of sample video frames according to the content of the video frames may include selecting the sample video frames according to at least one indicator such as the clarity, brightness, noise, etc. of the content of the video frames. For example, Fourier transform may be used to analyze whether the video frames are clear, and blurry video frames may be screened out, and the remaining video frames may be used as sample video frames. By selecting sample video frames according to the content of the video frames, it is beneficial to improve the quality of the sample video frames, so that each sample video frame carries more accurate scene attribute information.

[0044] The scene attribute information at least includes the spatial information and semantic information corresponding to the scene in the sample video frame. The spatial information may include the spatial position of each object in the scene, the spatial relative position relationship and other information; the semantic information may include the underlying semantic information of each object in the scene (such as contour, edge, color, texture, shape and other information), the intermediate semantic information (such as object status, etc.) and the high-level semantic information (such as object meaning, such as cat, person, sky, building, etc.). The sample video frame can be processed by the existing image analysis method to obtain the scene attribute information including at least the spatial information and semantic information corresponding to the scene in the sample video frame. By performing attribute analysis on the scene in the sample video frame, scene understanding can be achieved, laying the foundation for the subsequent rendering of the target video frame according to the determined target editing effect.

[0045] S120: Determine the target editing effect.

[0046] The target editing effect includes at least one of the following: editing scene movement and editing scene shape. Editing scene movement can be understood as editing the movement of the scene in the sample video frame when shooting; editing scene shape can be understood as editing the three-dimensional shape of the scene in the sample video frame.

[0047] When the target editing effect includes camera movement of the editing scene, the target editing effect may include at least one of the following: first-person main perspective camera movement, camera movement around the target object, and Hitchcock camera movement (also known as push-track zoom or sliding zoom), etc. Among them, when the target editing effect is first-person main visual camera movement, the sense of immersion in the edited video can be enhanced; when the target editing effect is camera movement around the target object, the target object in the edited video can be emphasized; when the target editing effect is Hitchcock camera movement, that is, the size of the main object in the edited video remains unchanged, the background perspective changes dramatically, and the visual effect of the background moving away from / approaching the main object is presented, which can create a suspenseful and tense atmosphere for the edited video or highlight the strong emotional changes in the characters.

[0048] When the target editing effect includes editing the shape of the scene, the target editing effect may include at least one of the following: at least partial enlargement / reduction of the scene, at least partial stretching / compression of the scene in a predetermined direction, at least partial shaking and bending of the scene around a predetermined coordinate axis, etc. When the target editing effect is at least partial enlargement / reduction of the scene, at least partial enlargement operation can be performed on the scene after mapping to the three-dimensional space, and / or reduction operation can be performed on at least partial enlargement operation, so as to create a "Big People's Country / Lilliputian Country" scene different from the surrounding environment for the edited video. When the target editing effect is at least partial stretching / compression of the scene in a predetermined direction, at least partial stretching operation and / or compression operation can be performed on the scene after mapping to the three-dimensional space. When at least partial stretching and compression of the scene mapped to the three-dimensional space are continuously performed, the effect of scene fluctuation can be created for the edited video; when the scene is stretched, the effect of crossing a wormhole or very fast speed can be created for the edited video. Among them, when the target editing effect is the shaking or bending of at least part of the scene around a predetermined coordinate axis, the shaking and / or bending of at least part of the object after the scene is mapped to the three-dimensional space can create a surreal scene for the edited video.

[0049] In the disclosed embodiment, the video editing device may provide at least one user interface to interact with the user during the video editing process. The user interface may include, for example, a target editing effect selection interface, in which a selection control for at least one editing effect may be presented, and in response to a user triggering operation on the selection control, the target editing effect may be determined from the editing effects.

[0050] There is no strict time sequence relationship between step S110 and step S120. For example, the video editing device may first push at least one editing effect for the user to select in the target editing effect selection interface based on the scene attribute information obtained in step S110; then, in response to the user's triggering operation on the selection control, the target editing effect may be determined from the pushed editing effects. For another example, after the video editing device determines the target editing effect from the editing effects in response to the user's triggering operation on the selection control, the video editing device may analyze the relevant scene attribute information according to the target editing effect.

[0051] S130. Generate a target video frame through a neural radiation field model based on scene attribute information and target editing effects.

[0052] Among them, the neural radiation field model is constructed based on sample video frames. Neural Radiance Fields (NeRF) is a computer vision technology that uses deep learning technology to extract the collective shape and texture information of objects from images from multiple perspectives, and then uses this information to generate a continuous three-dimensional radiation field, so that highly realistic three-dimensional models / scenes can be presented at any angle and distance. In this embodiment, the construction of the neural radiation field model can be realized by using only sample video frames in a video shot once, and the scene in the video can be automatically reconstructed implicitly, that is, the three-dimensional information of the scene is learned.

[0053] Scene understanding can be achieved through scene attribute information, and on this basis, the camera movement trajectory and / or deformation processing operation of the scene can be generated according to the target editing effect. For example, when the target editing effect is to move the camera around the target object, the target object in the scene can be determined through the scene attribute information, and a new camera movement trajectory can be generated based on this.

[0054] On the premise of determining the camera movement trajectory and / or deformation processing operation of the scene, the neural radiation field model can draw the deformed or non-deformed scenes at various viewing angles in the trajectory to obtain the target video frame. Based on the existing methods, new videos can be generated according to each target video frame, so that various video editing effects can be easily achieved. There is no need to repeatedly shoot videos to obtain the desired camera movement, nor is there a need for professionals to use professional editing tools to edit videos. Videos with film-level effects can be produced with low thresholds and low time consumption.

[0055] In addition, implicit reconstruction of the scenes in the sample video frames through NeRF can not only ensure highly realistic rendering effects, but also take less time than traditional techniques for building 3D models based on depth point clouds, and can achieve reconstruction in hours.

[0056] The technical solution of the embodiment of the present disclosure performs attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame; determines a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape; generates a target video frame based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

[0057] Since the neural radiation field model is constructed based on sample video frames, it can enable the neural radiation field model to have the ability to present the scenes in the sample video frames at any angle and distance. By performing attribute analysis on the scenes in the sample video frames, at least the spatial, semantic and other scene attribute information corresponding to the scenes can be obtained, so that scene understanding can be achieved. Based on the scene attribute information, the neural radiation field model can perform camera movement and / or shape editing on the scenes in the sample video frames according to the target editing effect to generate target video frames, thereby achieving diversified video editing.

[0058] The embodiments of the present disclosure can be combined with the various optional solutions in the video editing method provided in the above embodiments. The video editing method provided in this embodiment describes in detail the process of performing attribute analysis on the scene in the sample video frame. By performing at least one analysis operation of three-dimensional coordinate reconstruction, symmetry analysis and semantic analysis, a foundation can be laid for scene understanding.

[0059] In the video editing method provided in this embodiment, the attribute analysis of the scene in the sample video frame may include at least one of the following: reconstructing the three-dimensional coordinates corresponding to the scene in the sample video frame; analyzing the symmetry of the scene in the sample video frame; and performing semantic analysis on the scene in the sample video frame.

[0060] Although the constructed neural radiation field model can reconstruct the scene in three dimensions, it lacks a unified reference coordinate system when rendering the scene from each perspective. Therefore, when performing attribute analysis on the scene in the sample video frame, the three-dimensional coordinates corresponding to the scene in the sample video frame can be reconstructed to ensure the uniform position of the scene rendering from each perspective. The three-dimensional coordinates corresponding to the scene in the sample video frame can be reconstructed. For example, the world coordinates corresponding to the scene can be determined based on the camera position and camera parameters of each sample video frame; for another example, the constructed neural radiation field model can be used to map the scene to three-dimensional space, and the three-dimensional coordinate system can be determined based on the straight lines in the scene in the three-dimensional space.

[0061] The symmetry of the scene in the sample video frame is analyzed, for example, existing image recognition methods (such as pattern matching method, optimization search method, etc.) can be used to identify whether the sample video frame has a symmetry axis. The symmetry axis can be the symmetry axis of the entire scene or the symmetry axis of a local scene. For example, see Figure 5 a, Figure 5 a can be considered as the center line of the road as the symmetry axis, and the entire scene is symmetrical. By analyzing the symmetry of the scene in the sample video frame, a reference can be provided for the subsequent editing of the scene movement or the editing of the scene shape. For example, when editing the scene movement, at least part of the edited movement path can be parallel to the symmetry axis; when editing the scene shape, it can be stretched or compressed along the symmetry axis.

[0062] Among them, semantic analysis is performed on the scene in the sample video frame. For example, semantic segmentation of the scene in the sample video frame can be performed using existing image segmentation methods. For example, in the case of indoor scenes, wall segmentation, ground segmentation, ceiling segmentation, and furniture segmentation can be performed; in the case of outdoor scenes, building segmentation, road segmentation, etc. can be performed. By performing semantic analysis on the scene in the sample video frame, a foundation can be laid for the subsequent editing of scene movement or the selection of objects in the editing scene shape. For example, when editing scene movement, the target object that needs to be surrounded during the movement can be determined based on the semantic segmentation results; when editing the scene shape, the target object of the shape to be edited can be determined.

[0063] In these optional implementations, the scene attribute analysis process may include at least one of three-dimensional coordinate reconstruction, symmetry analysis, and semantic analysis, laying the foundation for scene understanding. In addition, other scene attribute analysis methods may also be applied here, which are not exhaustive here.

[0064] For example, Figure 2 The following is a flow chart of scene coordinate reconstruction in a video editing method provided by an embodiment of the present disclosure. Figure 2 As shown, in some optional implementations, reconstructing the three-dimensional coordinates corresponding to the scene in the sample video frame may include:

[0065] S210, detecting straight lines in the scene in each sample video frame;

[0066] S220, determining the spatial position of each straight line through the neural radiation field model;

[0067] S230, clustering the straight lines based on their spatial positions to obtain a target straight line;

[0068] S240: Identify a target plane in the scene, and determine a three-dimensional coordinate system of the scene according to the target plane and the target straight line.

[0069] In this embodiment, a detection method such as Hough transform may be used to detect the straight lines of the scene in each sample video frame, that is, to obtain the two-dimensional straight lines in the sample video frame.

[0070] The neural radiation field model can be used to render the depth information of a two-dimensional straight line. On this basis, since the camera position is known, the two-dimensional straight line can be mapped to a three-dimensional straight line in space to obtain the spatial position of the three-dimensional straight line.

[0071] After the two-dimensional straight lines of the scene in each sample video frame are mapped to spatial straight lines, due to the existence of errors, multiple straight lines are included near the true value straight line. The straight lines can be clustered according to their spatial positions to obtain target straight lines with higher confidence in the space.

[0072] Among them, the result of semantic analysis of the scene in the sample video frame can be used to identify the target plane corresponding to the scene, such as the ground, desktop, etc. Among them, the target plane can be used to represent the horizontal plane. On this basis, the target straight line can be analyzed. For example, mutually perpendicular straight lines are extracted, and two of the mutually perpendicular lines can be parallel to the target plane. At this time, the two straight lines parallel to the target plane can represent the two axes on the horizontal plane, and the straight line perpendicular to the target plane can represent the axis perpendicular to the horizontal plane, thereby ensuring that the established three-dimensional coordinate system is consistent with the semantic information of the scene itself.

[0073] In these optional implementations, the constructed neural radiation field model can be used to map the scene into three-dimensional space, and the straight lines in the scene can be combined with semantic analysis to determine the three-dimensional coordinate system to ensure that the established three-dimensional coordinate system is consistent with the semantic information of the scene itself.

[0074] In some optional implementations, analyzing the symmetry of the scene in the sample video frame may include: using the constructed neural radiation field model to map each symmetry axis of the scene in the sample video frame to the three-dimensional space; determining the target symmetry axis based on the symmetry axis of the three-dimensional space parallel to any axis in the three-dimensional coordinate system.

[0075] In these optional implementations, the symmetry axis can be determined based on the created three-dimensional coordinate system. By preferentially selecting the symmetry axis parallel to the coordinate axis in the reconstructed three-dimensional coordinate system as the basis for subsequently determining the editing camera effect, the editing effect can be enriched.

[0076] The technical solution of the embodiment of the present disclosure describes in detail the process of performing attribute analysis on the scene in the sample video frame. By performing at least one analysis operation of three-dimensional coordinate reconstruction, symmetry analysis and semantic analysis, a foundation can be laid for scene understanding. The video editing method provided in the embodiment of the present disclosure and the video editing method provided in the above embodiment belong to the same public concept. The technical details not described in detail in this embodiment can be found in the above embodiment, and the same technical features have the same beneficial effects in this embodiment and in the above embodiment.

[0077] The various optional solutions in the video editing method provided in the embodiment of the present disclosure and the above embodiment can be combined. The video editing method provided in this embodiment limits the determination of the target editing effect. The editing scene camera movement and / or the editing scene shape can be determined as the target editing effect based on the scene attribute information, thereby providing a more automated video editing capability.

[0078] For example, Figure 3 Schematic diagram of a video editing method provided by an embodiment of the present disclosure. Figure 3 As shown, the video editing method provided in this embodiment may include:

[0079] S310: Analyze the attributes of the scene in the sample video frame to obtain scene attribute information.

[0080] The scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame.

[0081] S320: Determine a target editing effect according to the scene attribute information.

[0082] The target editing effect includes at least one of the following: editing scene movement and editing scene shape.

[0083] Among them, the camera movement trajectory can be automatically generated according to the semantic information in the scene attribute information, and the target editing effect can include editing the scene camera movement.

[0084] For example, the camera may surround the target object in the scene. The process of determining the target object may include: segmenting different objects in each sample video frame through semantic analysis; selecting the target object in response to the user's target object selection operation;

[0085] Alternatively, it can be automatically determined based on the results of semantic analysis; for example, a target object category is predefined, and objects belonging to this category in the semantic segmentation results are taken as target objects; for example, objects whose appearance frequency in each sample video frame is higher than a preset frequency are taken as target objects.

[0086] For example, the camera orbit trajectory can be automatically generated according to the distribution of different objects in the scene. Thus, with only simple user interaction or no interaction, film-level camera movement effects can be automatically generated.

[0087] The camera movement trajectory can also be automatically generated according to the symmetry plane in the scene attribute information, and the target editing effect can include editing the scene camera movement. For example, Hitchcock camera movement can be performed on the symmetry plane to create a suspenseful and tense atmosphere or highlight the strong emotional changes of the characters.

[0088] The target object may also be deformed according to the semantic information in the scene attribute information. In this case, the target editing effect may include editing the scene shape. For example, the target object may be enlarged, shaken, etc. The process of determining the target object may refer to the above, and will not be described in detail here.

[0089] For example, Figure 4 A schematic diagram of changing the scene shooting camera movement in a video editing method provided by an embodiment of the present disclosure. Figure 4 , Figure 4 The arrowed curve in a is the original camera trajectory. Figure 4 The arrow curve in b is the camera trajectory after editing the scene. By moving the camera around the building, a grand atmosphere can be created.

[0090] For example, Figure 5 A schematic diagram of changing the shape of a scene in a video editing method provided by an embodiment of the present disclosure. Figure 5 , Figure 5 a is the original street shape, Figure 5 b shows the shape of the street after editing the scene shape. By bending the street upward, a surreal atmosphere can be created.

[0091] S330. Generate a target video frame through a neural radiation field model based on scene attribute information and target editing effects.

[0092] Among them, the neural radiation field model is constructed based on sample video frames.

[0093] The technical solution of the embodiment of the present disclosure defines the determination of the target editing effect. The editing scene camera movement and / or the editing scene shape can be determined as the target editing effect based on the scene attribute information, thereby providing a more automated video editing capability. The video editing method provided by the embodiment of the present disclosure and the video editing method provided by the above embodiment belong to the same public concept, and the technical details not fully described in this embodiment can be found in the above embodiment, and the same technical features have the same beneficial effects in this embodiment and the above embodiment.

[0094] The various optional solutions in the video editing method provided in the embodiment of the present disclosure and the above embodiment can be combined. The video editing method provided in this embodiment can also be combined with the existing image generation model and the traditional rendering link to present a richer video editing effect.

[0095] For example, Figure 6 FIG. 1 is a flowchart of a video editing method provided by an embodiment of the present disclosure. Figure 6 As shown, in some implementations, the video editing process may first be processed in the video dimension based on an existing image generation model, and then a target video frame may be generated through a NeRF model, which may include:

[0096] S610: Analyze the attributes of the scenes in the sample video frames to obtain scene attribute information.

[0097] The scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame.

[0098] S620: Determine the target editing effect.

[0099] The target editing effect includes at least one of the following: editing scene movement and editing scene shape.

[0100] S630. Generate a target video frame through a neural radiation field model based on scene attribute information and target editing effects.

[0101] Among them, the neural radiation field model is constructed based on sample video frames.

[0102] S640: Edit the key frame of the target video frame to obtain an updated key frame.

[0103] After rendering the target video frames of each angle of the scene using the NeRF model, at least part of the target video frames can be used as key frames. The key frames can be edited based on the existing image generation model, such as modifying and enhancing the key frames. Editing the key frames can include stylizing the key frames.

[0104] This embodiment will take the stylized processing of key frames by the image generation model as an example for description. The same is true for other editing methods, which will not be described exhaustively. After the key frames are stylized, video frames with different styles from the original sample video frames can be obtained. For example, a realistic style sample video frame can be edited into a new Lego style video frame.

[0105] S650. Determine the optical flow information between target video frames through a neural radiance field model.

[0106] There is no strict time sequence relationship between step S640 and step S650.

[0107] In this embodiment, since each pixel in the scene rendered by the NeRF model has a unique spatial coordinate, the pixel in the generated target video frame and the pixel in the previous and next target video frames have a spatial correspondence, that is, they have a correct optical flow. The optical flow information between the target video frames can be determined by the position change of the pixel in the current target video frame and the corresponding pixel in the adjacent video frame. Among them, the optical flow information can be understood as information related to the movement of the pixel, such as the direction of movement, displacement, etc.

[0108] S660: Interpolate the updated key frame based on the optical flow information to obtain an intermediate video frame.

[0109] After determining the optical flow information between each target video frame, the optical flow information of the video frame between each adjacent key frame can be used as the optical flow information of the video frame between each adjacent updated key frame. Furthermore, based on the optical flow information of the video frame between each updated key frame, interpolation can be performed between each adjacent updated key frame to obtain an intermediate video frame.

[0110] For example, the optical flow information of the video frames between adjacent key frames before stylization can be used as the optical flow information of the intermediate video frames between adjacent key frames after stylization. Adjacent stylized key frames can be interpolated according to the optical flow information of the intermediate frames, so as to obtain stylized intermediate video frames.

[0111] S670: Use the updated key frame and the intermediate video frame as new target video frames.

[0112] A new target video frame can be generated from each updated key frame and the intermediate video frames between each adjacent updated key frame. For example, a stylized target video frame can be generated from a stylized key frame and the stylized intermediate video frames between each stylized key frame.

[0113] Traditional video stylization can be considered as the image generation model processing each frame. The same pixel in the scene may have different positions in each frame after stylization, which will cause the processed video to shake and flicker.

[0114] In this optional embodiment, since each pixel in the scene rendered by the NeRF model has a unique spatial coordinate, the pixels of the generated image have a corresponding relationship with the pixels in the previous and next images, that is, they have a correct optical flow. The NeRF model can use the correct optical flow between the target video frames to interpolate between the target video frames to obtain intermediate frames, thereby reducing the jitter and flicker of the edited video.

[0115] For example, Figure 7 FIG. 1 is a flowchart of a video editing method provided by an embodiment of the present disclosure. Figure 7 As shown, in some implementations, the video editing process may first generate a target video frame through a NeRF model, then process the target video frame based on an existing image generation model, and then feed back to the NeRF model to achieve video processing in the spatial dimension, which may include:

[0116] S710: Analyze the attributes of the scene in the sample video frame to obtain scene attribute information.

[0117] The scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame.

[0118] S720: Determine the target editing effect.

[0119] The target editing effect includes at least one of the following: editing scene movement and editing scene shape.

[0120] S730. Generate a target video frame through a neural radiation field model based on scene attribute information and target editing effects.

[0121] Among them, the neural radiation field model is constructed based on sample video frames.

[0122] S740: Edit the key frame of the target video frame to obtain an updated key frame.

[0123] After the target video frames of each angle of the scene are rendered using the NeRF model, at least part of the target video frames can be used as key frames. The key frames can be edited based on the existing image generation model, such as modifying and enhancing the key frames. Editing the key frames may include stylizing the key frames. This embodiment will be described by taking the stylization of the key frames by the image generation model as an example. Other editing methods are similar and will not be described exhaustively.

[0124] S750. Using the updated key frame, adjust the parameters in the neural radiation field model so that the neural radiation field model with adjusted parameters generates a new target video frame.

[0125] After obtaining the stylized keyframes, the stylized keyframes can be reversely mapped back to the NeRF model to modify the color information of each pixel after the scene is mapped to the three-dimensional space (i.e., modify the texture of the scene in the spatial dimension). Afterwards, the updated NeRF model can be used to continue subsequent rendering to achieve stylization of the original video.

[0126] In these optional embodiments, by changing the texture of the target video frame and projecting it back to the NeRF model, the NeRF model can render a target video frame with consistent texture, thereby ensuring that the original video has consistency after style processing.

[0127] In some optional implementations, after the target video frame is generated, the method may further include: generating a new target video frame including the target 3D model according to the depth information of the scene in the target video frame and the depth information of the target 3D model.

[0128] The three-dimensional model can be understood as a three-dimensional special effect material that can be added to the scene in the sample video frame. Before generating a new target video frame containing the target three-dimensional model, the three-dimensional model to be added can be obtained first. For example, the three-dimensional model to be added can be obtained from a preset three-dimensional model library based on the scene attribute information of the scene in the sample video frame. Specifically, the correspondence between the three-dimensional model and the scene semantics can be pre-set, and based on the correspondence, a three-dimensional model that matches the semantic information of the scene in the sample video frame can be determined from the preset three-dimensional model library as the three-dimensional model to be added.

[0129] Furthermore, a new target video frame including the target 3D model may be generated according to the depth information of the scene in the target video frame and the depth information of the target 3D model.

[0130] In the traditional rendering chain, rendering can be performed after constructing a three-dimensional model (such as a triangular patch model). The spatial position of the three-dimensional model in the scene can be predefined, or it can be determined by receiving user settings through a predefined user interface. The NeRF model can be used to obtain the depth information corresponding to the scene of the target video frame, as well as the depth information of the three-dimensional model at the spatial position. On this basis, because the geometric information of the three-dimensional model is known, in the new target video frame obtained after rendering the three-dimensional model into the three-dimensional scene, the three-dimensional model and the scene can have a reasonable occlusion relationship. The depth information of the three-dimensional model can be compared with the depth information corresponding to the scene in the target video frame, and pixels with a smaller depth can be determined from the three-dimensional model and the scene for rendering, so that the three-dimensional model can be added to the scene of the target video frame, and it can have a reasonable occlusion relationship with the scene.

[0131] For example, Figure 8 A schematic diagram of a new target video frame of a video editing method provided by an embodiment of the present disclosure. Figure 8 When generating a target video frame, the depth information of the scene (soccer goal) and the depth information of the soccer model in the target video frame can be used to add the soccer model to the scene, thereby enriching the scene content.

[0132] In these optional implementations, the target video frame can also be combined with a traditional rendering link to generate a new target video frame, which can present richer video editing effects.

[0133] The technical solution of the disclosed embodiment, the video editing process can also be combined with the generative artificial intelligence model and the traditional rendering link, which can present richer video editing effects. The video editing method provided by the disclosed embodiment and the video editing method provided by the above embodiment belong to the same public concept, and the technical details not described in detail in this embodiment can be referred to the above embodiment, and the same technical features have the same beneficial effects in this embodiment and the above embodiment.

[0134] Fig. 9 The video editing device provided in this embodiment is applicable to video editing, such as changing the camera movement when shooting a scene in a video, and / or changing the three-dimensional shape of a scene.

[0135] like Fig. 9 As shown, the video editing device provided by the embodiment of the present disclosure may include:

[0136] The analysis module 910 is used to perform attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame;

[0137] The effect determination module 920 is used to determine a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape;

[0138] The editing module 930 is used to generate a target video frame based on scene attribute information and target editing effects through a neural radiation field model; wherein the neural radiation field model is constructed based on sample video frames.

[0139] In some optional implementations, the analysis module may be used to perform at least one of the following attribute analyses on the scene in the sample video frame:

[0140] Reconstructing the three-dimensional coordinates corresponding to the scene in the sample video frame;

[0141] Analyzing the symmetry of scenes in sample video frames;

[0142] Perform semantic analysis on scenes in sample video frames.

[0143] In some optional implementations, the analysis module may be used to:

[0144] Detecting straight lines in the scene in each sample video frame;

[0145] The spatial position of each straight line is determined through the neural radiation field model;

[0146] Cluster the straight lines based on their spatial positions to obtain the target straight line;

[0147] Identify the target plane in the scene, and determine the three-dimensional coordinate system of the scene based on the target plane and the target line.

[0148] In some optional implementations, the effect determination module may be used to:

[0149] Determine the target editing effect based on the scene attribute information.

[0150] In some optional implementations, the video editing device may further include:

[0151] The sample video frame selection module is used to select the sample video frame based on at least one of the following methods before performing attribute analysis on the scene in the sample video frame:

[0152] Selecting sample video frames according to the co-viewing relationship between the video frames;

[0153] Sample video frames are selected according to the content of each video frame.

[0154] In some optional implementations, the video editing device may further include:

[0155] A post-editing module, used for editing the key frame of the target video frame after generating the target video frame to obtain an updated key frame;

[0156] Determine the optical flow information between target video frames through the neural radiance field model;

[0157] Interpolate the updated key frame based on the optical flow information to obtain an intermediate video frame;

[0158] The updated key frame and the intermediate video frame are used as new target video frames.

[0159] In some optional implementations, the post-editing module may also be used to: after generating the target video frame, edit the key frame of the target video frame to obtain an updated key frame;

[0160] Correspondingly, the model updating module can be used to adjust the parameters in the neural radiation field model using the updated key frames, so that the neural radiation field model with adjusted parameters generates a new target video frame.

[0161] In some optional implementations, the post-editing module may also be used to:

[0162] After the target video frame is generated, a new target video frame including the target three-dimensional model is generated according to the depth information of the scene in the target video frame and the depth information of the target three-dimensional model.

[0163] The video editing device provided in the embodiments of the present disclosure can execute the video editing method provided in any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0164] It is worth noting that the various units and modules included in the above-mentioned device are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the embodiments of the present disclosure.

[0165] Reference below Fig.10 , which shows an electronic device (eg, Fig.10 The terminal device in the embodiment of the present disclosure may include but is not limited to mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Fig.10 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0166] like Fig.10 As shown, the electronic device 1000 may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 1001, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1008 to a random access memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the electronic device 1000 are also stored. The processing device 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0167] Typically, the following devices may be connected to the I / O interface 1005: an input device 1006 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 1007 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 1008 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 may allow the electronic device 1000 to communicate with other devices wirelessly or by wire to exchange data. Although Fig.10 The electronic device 1000 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0168] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device 1009, or installed from a storage device 1008, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the video editing method of the embodiment of the present disclosure are executed.

[0169] The electronic device provided by the embodiment of the present disclosure and the video editing method provided by the above embodiment belong to the same public concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0170] The embodiment of the present disclosure provides a computer storage medium on which a computer program is stored. When the program is executed by a processor, the video editing method provided by the above embodiment is implemented.

[0171] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM) or a flash memory (FLASH), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0172] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (Hyper Text Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0173] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0174] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device:

[0175] Perform attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame; determine the target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape; generate a target video frame based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

[0176] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0177] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0178] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, wherein the names of the units and modules do not, in some cases, limit the units and modules themselves.

[0179] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: Field Programmable Gate Array (FPGA), Application Specific Integrated Circuit (ASIC), Application Specific Standard Parts (ASSP), System on Chip (SOC), Complex Programmable Logic Device (CPLD), etc.

[0180] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0181] According to one or more embodiments of the present disclosure, a video editing method is provided, the method comprising:

[0182] Performing attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame;

[0183] Determine a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape;

[0184] A target video frame is generated based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

[0185] According to one or more embodiments of the present disclosure, a video editing method is provided, further comprising:

[0186] In some optional implementations, the performing attribute analysis on the scene in the sample video frame includes at least one of the following:

[0187] Reconstructing the three-dimensional coordinates corresponding to the scene in the sample video frame;

[0188] Analyzing the symmetry of the scene in the sample video frame;

[0189] Perform semantic analysis on the scene in the sample video frame.

[0190] According to one or more embodiments of the present disclosure, a video editing method is provided, further comprising:

[0191] In some optional implementations, reconstructing the three-dimensional coordinates corresponding to the scene in the sample video frame includes:

[0192] Detecting straight lines in the scene in each of the sample video frames;

[0193] Determining the spatial position of each of the straight lines by using the neural radiation field model;

[0194] Clustering the straight lines based on their spatial positions to obtain a target straight line;

[0195] A target plane corresponding to a scene in the sample video frame is identified, and a three-dimensional coordinate corresponding to the scene is determined according to the target plane and the target straight line.

[0196] Analyzing the symmetry of the scene in the sample video frame According to one or more embodiments of the present disclosure, a video editing method is provided, further comprising:

[0197] In some optional implementations, determining the target editing effect includes:

[0198] The target editing effect is determined according to the scene attribute information.

[0199] According to one or more embodiments of the present disclosure, a video editing method is provided, further comprising:

[0200] In some optional implementations, after generating the target video frame, the method further includes:

[0201] The sample video frame is selected based on at least one of the following methods:

[0202] Selecting the sample video frames according to the co-viewing relationship between the video frames;

[0203] The sample video frames are selected according to the content of each video frame.

[0204] According to one or more embodiments of the present disclosure, a video editing method is provided, further comprising:

[0205] In some optional implementations, after generating the target video frame, the method further includes:

[0206] Editing the key frame of the target video frame to obtain an updated key frame;

[0207] Determining optical flow information between the target video frames through the neural radiance field model;

[0208] Interpolating the updated key frame based on the optical flow information to obtain an intermediate video frame;

[0209] The updated key frame and the intermediate video frame are used as new target video frames.

[0210] According to one or more embodiments of the present disclosure, a video editing method is provided, further comprising:

[0211] In some optional implementations, after generating the target video frame, the method further includes:

[0212] Editing the key frame of the target video frame to obtain an updated key frame;

[0213] The updated key frame is used to adjust the parameters in the neural radiation field model so that the neural radiation field model with the adjusted parameters generates a new target video frame.

[0214] According to one or more embodiments of the present disclosure, a video editing method is provided, further comprising:

[0215] In some optional implementations, after generating the target video frame, the method further includes:

[0216] A new target video frame including the target three-dimensional model is generated according to the depth information of the scene in the target video frame and the depth information of the target three-dimensional model.

[0217] According to one or more embodiments of the present disclosure, a video editing device is provided, the device comprising:

[0218] An analysis module, used to perform attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame;

[0219] An effect determination module, used to determine a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape;

[0220] An editing module is used to generate a target video frame based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

[0221] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0222] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0223] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. A video editing method, characterized in that: include: Performing attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame; Determine a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape; A target video frame is generated based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

2. The method according to claim 1, characterized in that The performing attribute analysis on the scene in the sample video frame includes at least one of the following: Reconstructing the three-dimensional coordinates corresponding to the scene in the sample video frame; Analyzing the symmetry of the scene in the sample video frame; Perform semantic analysis on the scene in the sample video frame.

3. The method according to claim 2, characterized in that The reconstructing the three-dimensional coordinates corresponding to the scene in the sample video frame includes: Detecting straight lines in the scene in each of the sample video frames; Determining the spatial position of each of the straight lines by using the neural radiation field model; Clustering the straight lines based on their spatial positions to obtain a target straight line; A target plane corresponding to a scene in the sample video frame is identified, and a three-dimensional coordinate corresponding to the scene is determined according to the target plane and the target straight line.

4. The method according to claim 1, characterized in that: The determining of the target editing effect includes: The target editing effect is determined according to the scene attribute information.

5. The method according to claim 1, characterized in that: Before performing attribute analysis on the scene in the sample video frame, the method further includes: The sample video frame is selected based on at least one of the following methods: Selecting the sample video frames according to the co-viewing relationship between the video frames; The sample video frames are selected according to the content of each video frame.

6. The method according to claim 1, characterized in that After the target video frame is generated, the method further includes: Editing the key frame of the target video frame to obtain an updated key frame; Determining optical flow information between the target video frames through the neural radiance field model; Interpolating the updated key frame based on the optical flow information to obtain an intermediate video frame; The updated key frame and the intermediate video frame are used as new target video frames.

7. The method according to claim 1, characterized in that After the target video frame is generated, the method further includes: Editing the key frame of the target video frame to obtain an updated key frame; The updated key frame is used to adjust the parameters in the neural radiation field model so that the neural radiation field model with the adjusted parameters generates a new target video frame.

8. The method according to claim 1, characterized in that After the target video frame is generated, the method further includes: A new target video frame including the target three-dimensional model is generated according to the depth information of the scene in the target video frame and the depth information of the target three-dimensional model.

9. A video editing device, characterized in that: include: An analysis module, used to perform attribute analysis on the scene in the sample video frame to obtain scene attribute information; wherein the scene attribute information at least includes spatial information and semantic information corresponding to the scene in the sample video frame; An effect determination module, used to determine a target editing effect; wherein the target editing effect includes at least one of the following: editing scene movement and editing scene shape; An editing module is used to generate a target video frame based on the scene attribute information and the target editing effect through a neural radiation field model; wherein the neural radiation field model is constructed based on the sample video frame.

10. An electronic device, characterized in that: The electronic device comprises: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the video editing method as described in any one of claims 1-8.

11. A storage medium comprising computer executable instructions, wherein the computer executable instructions are used to perform the video editing method according to any one of claims 1 to 8 when executed by a computer processor.

Images