A video storage, video extraction and display method, device and computer equipment

Abstract

This specification relates to the technical field of short video processing, and particularly to a video storage, video extraction and display method, device and computer equipment. The video storage method includes segmenting a video with accents to be removed to obtain at least one segmented audio and at least one image segment; extracting features for each segmented audio and each image segment to determine whether to perform duplicate removal processing on the segmented audio and the image segment; for the video with accents to be removed after duplicate removal processing, determining a first index identifier corresponding to the segmented audio that has been de-duplicated and a second index identifier corresponding to the image segment that has been de-duplicated to determine a target index identifier corresponding to the video with accents to be removed after duplicate removal processing. By using the embodiments of this specification, duplicate removal of audio and image segments in a video with accents to be removed is achieved, improving the utilization rate of storage space.

Description

Technical Field

[0001] This specification relates to the technical field of short video processing, and particularly to a method, apparatus, and computer device for video storage, video extraction, and display. Background Art

[0002] Currently, in the technical field of short video processing, in order to save storage space, duplicate removal processing is often performed on videos to be stored. However, when performing duplicate removal identification and processing on videos to be stored, the entire video to be stored is matched with the source video for consistency to determine whether duplicate removal processing should be performed on the video to be stored. However, when creating videos, users often capture multiple videos and then splice these multiple videos to obtain a target video. If this target video is the video to be stored, based on the current duplicate removal identification and processing method, no duplicate removal processing needs to be performed on this type of video to be stored. As a result, the degree of duplicate removal is relatively shallow, the storage space is still occupied in large amounts, and the utilization rate of the storage space is low.

[0003] How to perform in-depth duplicate removal on videos, release the storage space to a greater extent, and improve the utilization rate of the storage space is an urgent problem to be solved in the prior art. Summary of the Invention

[0004] To solve the problems in the prior art, embodiments of this specification provide a method, apparatus, and computer device for video storage, video extraction, and display, which realizes segmenting the audio and image segments in the video to be de-duplicated, and respectively performing duplicate removal identification and processing on the segmented audio and image segments, increasing the depth of the duplicate removal processing, further releasing the storage space, and improving the utilization rate of the storage space.

[0005] To solve the above technical problems, the specific technical solutions of this specification are as follows:

[0006] On the one hand, embodiments of this specification provide a video storage method, including:

[0007] Performing segmentation processing on the video to be de-duplicated to obtain at least one segmented audio and at least one image segment;

[0008] Performing feature extraction on each of the segmented audio and each of the image segments to obtain at least one segmented audio feature vector and at least one image segment feature vector;

[0009] For each of the segmented audio feature vectors, when the first similarity between the source audio feature vector and the segmented audio feature vector meets the first preset condition, performing duplicate removal processing on the segmented audio corresponding to the segmented audio feature vector, where the source audio feature vector is obtained by performing feature extraction on the source audio;

[0010] For each of the image segment feature vectors, when the second similarity between the source image segment feature vector and the image segment feature vector satisfies a second preset condition, perform duplicate removal processing on the image segment corresponding to the image segment feature vector, where the source image segment feature vector is obtained by performing feature extraction on a source image segment;

[0011] For the video with duplicate accents to be processed after duplicate removal processing, determine a first index identifier corresponding to the segmented audio that has been removed duplicates and a second index identifier corresponding to the image segment that has been removed duplicates; and

[0012] Based on the first index identifier, the second index identifier, and the remaining audio - video information that has not been processed for duplicate removal, determine a target index identifier corresponding to the video with duplicate accents to be processed after duplicate removal processing.

[0013] Further, before performing feature extraction on each of the segmented audio and each of the image segments to obtain at least one segmented audio feature vector and at least one image segment feature vector, it further includes,

[0014] Perform segmentation processing on the video with duplicate accents to be processed to obtain at least one characterization information;

[0015] For each of the characterization information, determine whether the storage space data of the characterization information satisfies a third preset condition;

[0016] When it is determined that the storage space data of the target characterization information satisfies the third preset condition, perform feature extraction on the target characterization information to obtain a target characterization information feature vector;

[0017] When the third similarity between the source characterization information feature vector and the target characterization information feature vector satisfies a fourth preset condition, perform duplicate removal processing on the target characterization information, where the source characterization information feature vector is obtained by performing feature extraction on source characterization information,

[0018] and;

[0019] For the video with duplicate accents to be processed after duplicate removal processing, determine a third index identifier corresponding to the target characterization information that has been removed duplicates; and

[0020] Based on the first index identifier, the second index identifier, the third index identifier, and the remaining audio - video information that has not been processed for duplicate removal, determine a target index identifier corresponding to the video with duplicate accents to be processed after duplicate removal processing.

[0021] Further, the determination of the first similarity between the source audio feature vector and the segmented audio feature vector further includes,

[0022] Performing segmentation processing on a preset audio according to the duration information of the segmented audio corresponding to the segmented audio feature vector to obtain a plurality of source audios;

[0023] Performing feature extraction on each of the source audios to obtain a plurality of source audio feature vectors;

[0024] Respectively determining sub-first segmented audio feature data and sub-first source audio feature data corresponding to each first dimension from the segmented audio feature vector and each source audio feature vector;

[0025] Determining a sub-first similarity between the sub-first segmented audio feature data and the sub-first source audio feature data under each first dimension; and

[0026] Performing weighted calculation on the sub-first similarity corresponding to each first dimension by using first weight data corresponding to each first dimension to obtain the first similarity.

[0027] Furthermore, the determination of the second similarity between the source image segment feature vector and the image segment feature vector further includes

[0028] Performing segmentation processing on a preset video according to the duration information of the image segment corresponding to the image segment feature vector to obtain a plurality of source image segments;

[0029] Performing feature extraction on each of the source image segments to obtain a plurality of source image segment feature vectors;

[0030] Respectively determining sub-second image segment feature data and sub-second source image segment feature data corresponding to each second dimension from the image segment feature vector and each source image segment feature vector;

[0031] Determining a sub-second similarity between the sub-second image segment feature data and the sub-second source image segment feature data under each second dimension; and

[0032] Performing weighted calculation on the sub-second similarity corresponding to each second dimension by using second weight data corresponding to each second dimension to obtain the second similarity.

[0033] Furthermore, the determination of the first index identifier further includes

[0034] Determining the first index identifier of the target source audio corresponding to the segmented audio for the segmented audio after duplicate removal processing, where the target source audio is the source audio corresponding to the source audio feature vector whose first similarity determined with the segmented audio feature vector satisfies the first preset condition, and the first index identifier points to the target source audio.

[0035] Further, the determination of the second index identifier further includes

[0036] For the image segment after deduplication processing, determining the second index identifier of the target source image segment corresponding to the image segment, where the target source image segment is the source image segment corresponding to the source image segment feature vector whose second similarity with the image segment feature vector is satisfied with the second preset condition, and the second index identifier points to the target source image segment.

[0037] On the other hand, an embodiment of this specification also provides a video extraction and display method, including

[0038] According to the received target index identifier, determining a plurality of sub-index identifiers and sub-audio / video index information;

[0039] For each of the sub-index identifiers, respectively determining the corresponding source image segment or source audio or source characterization information;

[0040] Based on the source image segment, source audio, source characterization information, and the sub-audio / video corresponding to the sub-audio / video index information, determining a target video for display.

[0041] On the other hand, an embodiment of this specification also provides a video storage device, including

[0042] A segmentation unit, configured to perform segmentation processing on the audio / video to be deduplicated to obtain at least one segmented audio and at least one image segment;

[0043] An extraction unit, configured to perform feature extraction on each of the segmented audio and each of the image segments to obtain at least one segmented audio feature vector and at least one image segment feature vector;

[0044] A first deduplication unit, configured to perform deduplication processing on the segmented audio corresponding to the segmented audio feature vector when the first similarity between the source audio feature vector and the segmented audio feature vector satisfies the first preset condition, where the source audio feature vector is obtained by performing feature extraction on the source audio;

[0045] A second deduplication unit, configured to perform deduplication processing on the image segment corresponding to the image segment feature vector when the second similarity between the source image segment feature vector and the image segment feature vector satisfies the second preset condition, where the source image segment feature vector is obtained by performing feature extraction on the source image segment;

[0046] A first determination unit, configured to determine, for the video to be de-accented after de-duplication processing, a first index identifier corresponding to the segmented audio to be de-duplicated and a second index identifier corresponding to the image segment to be de-duplicated; and

[0047] A second determination unit, configured to determine a target index identifier corresponding to the video to be de-accented after the de-duplication processing based on the first index identifier, the second index identifier, and the remaining audio-visual information that has not been de-duplicated.

[0048] On the other hand, an embodiment of this specification further provides a video extraction and display device, including

[0049] A third determination unit, configured to determine a plurality of sub-index identifiers and sub-audio-visual index information according to the received target index identifier;

[0050] A fourth determination unit, configured to respectively determine corresponding source image segments, source audio, or source characterization information for each of the sub-index identifiers;

[0051] A fifth determination unit, configured to determine a target video for display based on the source image segments, source audio, source characterization information, and sub-audio-visual corresponding to the sub-audio-visual index information.

[0052] On the other hand, an embodiment of this specification further provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the above method when executing the computer program.

[0053] On the other hand, an embodiment of this specification further provides a computer-readable storage medium, on which computer instructions are stored, and when the computer instructions are executed by a processor, the above method is implemented.

[0054] Using the embodiments of this specification, the video with accent to be removed is separated and segmented to obtain at least one segmented audio and at least one image segment; feature extraction is respectively performed on each segmented audio and each image segment to obtain at least one segmented audio feature vector and at least one image segment feature vector; for each segmented audio feature vector, the first similarity with the source audio feature vector is respectively determined, and when the first similarity meets the first preset condition, duplicate removal processing is performed on the segmented audio corresponding to the segmented audio feature vector. Similarly, for the image segment feature vector whose second similarity with the source image segment feature vector meets the second preset condition, duplicate removal processing is performed on the corresponding image segment. Furthermore, for the video with accent after duplicate removal processing, a first index identifier and a second index identifier are determined for constructing a target index identifier together with the remaining audio and video information that has not been subjected to duplicate removal processing. Thus, the audio and image segments in the video with accent are segmented, and duplicate removal recognition and processing are respectively performed on the segmented audio and image segments, increasing the depth of duplicate removal processing, further releasing the storage space, and improving the utilization rate of the storage space. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of this specification. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0056] Figure 1 The figure shows a schematic diagram of an implementation system for a video storage and video extraction and display method according to an embodiment of this specification;

[0057] Figure 2 The figure shows a flowchart of a video storage method according to an embodiment of this specification;

[0058] Figure 3A The figure shows a flowchart of a video duplicate removal processing method according to an embodiment of this specification;

[0059] Figure 3B The figure shows a flowchart of a method for determining a target index identifier according to an embodiment of this specification;

[0060] Figure 3C The figure shows a schematic diagram of a segmentation processing method according to an embodiment of this specification;

[0061] Figure 4A The figure shows a flowchart of a method for determining the first similarity according to an embodiment of this specification;

[0062] Figure 4BThe figure shows a flowchart of a second similarity determination method according to an embodiment of this specification;

[0063] Figure 5A The figure shows a schematic diagram of a principle of a video extraction and display method according to an embodiment of this specification;

[0064] Figure 5B The figure shows a schematic diagram of a principle of a video storage and video extraction and display method according to an embodiment of this specification;

[0065] Figure 6A The figure shows a schematic structural diagram of a video storage device according to an embodiment of this specification;

[0066] Figure 6B The figure shows a schematic structural diagram of a video storage device according to another embodiment of this specification;

[0067] Figure 6C The figure shows a schematic structural diagram of a video extraction and display device according to an embodiment of this specification;

[0068] Figure 7 The figure is a schematic structural diagram of a computer device according to an embodiment of this specification.

[0069]

Explanation of Reference Numerals

[0070] 101, user terminal;

[0071] 102, server;

[0072] 501, video to be de-accented;

[0073] 502, multiple segmented audio;

[0074] 503, multiple image segments;

[0075] 504, multiple characterization information;

[0076] 511, de-duplicated processed segmented audio;

[0077] 512, de-duplicated processed image segments;

[0078] 513, de-duplicated processed target characterization information;

[0079] 521, remaining audio;

[0080] 522, remaining image segments;

[0081] 523, remaining characterization information;

[0082] 531, first index identifier;

[0083] 532, second index identifier;

[0084] 533, third index identifier;

[0085] 540, target index identifier;

[0086] 6010, segmented unit;

[0087] 6020, extraction unit;

[0088] 6030, first deduplication unit;

[0089] 6040, second deduplication unit;

[0090] 6050, first determination unit;

[0091] 6060, second determination unit;

[0092] 6071, characterizes the segmented unit;

[0093] 6072, judgment unit;

[0094] 6073, characterization feature extraction unit;

[0095] 6074, characterizes the deduplication unit;

[0096] 6075, identification determination unit;

[0097] 6076, storage unit;

[0098] 6080, third determination unit;

[0099] 6090, fourth determination unit;

[0100] 6110, fifth determination unit;

[0101] 702. Computer equipment;

[0102] 704. Processing equipment;

[0103] 706. Storage resources;

[0104] 708, driving mechanism;

[0105] 710, input / output module;

[0106] 712. Input devices;

[0107] 714. Output device;

[0108] 716. Presentation equipment;

[0109] 718. Graphical User Interface;

[0110] 720, network interface;

[0111] 722. Communication link;

[0112] 724. Communication bus. Detailed implementation manner

[0113] The following will clearly and completely describe the technical solutions in the embodiments of this specification with reference to the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this specification without creative efforts shall fall within the scope of protection of this specification.

[0114] It should be noted that the terms "first", "second", etc. in the specification, claims and above-mentioned drawings of this specification are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of this specification described here can be implemented in an order different from those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or equipment that includes a series of steps or units does not necessarily need to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or equipment.

[0115] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0116] Figure 1The following is a schematic diagram of an implementation system for a video storage and video extraction and display method according to an embodiment of this specification, which may include: a user terminal 101 and a server 102. The user terminal 101 and the server 102 communicate with each other through a network. The network may include a local area network (LAN for short), a wide area network (WAN for short), the Internet, or a combination thereof, and is connected to websites, user devices (such as computing devices), and backend systems. After receiving a video storage request sent by the user through the user terminal 101, the server 102 performs segmentation processing on the de-duplication-required video included in the video storage request to obtain at least one segmented audio and at least one image segment; performs feature extraction on each segmented audio and each image segment to obtain at least one segmented audio feature vector and at least one image segment feature vector; for each segmented audio feature vector, when the first similarity between the source audio feature vector and the segmented audio feature vector meets a first preset condition, performs de-duplication processing on the segmented audio corresponding to the segmented audio feature vector; for each image segment feature vector, when the second similarity between the source image segment feature vector and the image segment feature vector meets a second preset condition, performs de-duplication processing on the image segment corresponding to the image segment feature vector; for the de-duplication-required video after de-duplication processing, determines a first index identifier corresponding to the de-duplicated segmented audio and a second index identifier corresponding to the de-duplicated image segment; and then, based on the first index identifier, the second index identifier, and the remaining audio-visual information that has not been de-duplicated, determines a target index identifier corresponding to the de-duplication-required video after de-duplication processing, and stores and sends the target index identifier to the user terminal 101. In addition, when the server 102 receives an extraction and display request for the de-duplication-required audio-visual video sent by the user terminal 101, for the target index identifier included in the extraction and display request, determines a plurality of sub-index identifiers and sub-audio-visual index information; for each sub-index identifier, respectively determines the corresponding source image segment, source audio, or source characterization information; and based on the source image segment, source audio, source characterization information, and the sub-audio-visual corresponding to the sub-audio-visual index information, determines and obtains the target video and sends it to the user terminal 101 so that the user terminal performs audio-visual display.

[0117] Optionally, the server 102 may be a node of a cloud computing system (not shown in the figure), or each server 102 may be a separate cloud computing system, including multiple computers interconnected by a network and working as a distributed processing system.

[0118] In an optional embodiment, the user terminal 101 may include electronic devices, including but not limited to smartphones, acquisition devices, desktop computers, tablet computers, laptop computers, smart speakers, digital assistants, augmented reality (AR) / virtual reality (VR) devices, smart wearable devices, and other types of electronic devices. Optionally, the operating systems running on the electronic devices may include but are not limited to Android, IOS, Linux, Windows, etc.

[0119] In addition, it should be noted that Figure 1 The application environment shown is only one provided in this specification. In actual applications, there may also be multiple user terminals 101, which are not limited in this specification.

[0120] Figure 2 The figure shows a flowchart of a video storage method according to an embodiment of this specification. The video storage process is described in this figure, but based on routine or non-creative labor, it may include more or fewer operation steps. The order of steps listed in the embodiment is only one way among the execution orders of numerous steps and does not represent the only execution order. When the actual system or device product executes, it can be executed in the order shown in the embodiment or the figure or in parallel. Specifically, as Figure 2 shown, the method may include:

[0121] S210, perform segmentation processing on the video to be de-accented to obtain at least one segmented audio and at least one image segment;

[0122] S220, perform feature extraction on each segmented audio and each image segment to obtain at least one segmented audio feature vector and at least one image segment feature vector;

[0123] S230, for each segmented audio feature vector, when the first similarity between the source audio feature vector and the segmented audio feature vector meets the first preset condition, perform de-duplication processing on the segmented audio corresponding to the segmented audio feature vector;

[0124] S240, for each image segment feature vector, when the second similarity between the source image segment feature vector and the image segment feature vector meets the second preset condition, perform de-duplication processing on the image segment corresponding to the image segment feature vector;

[0125] S250, for the video to be de-accented after de-duplication processing, determine the first index identifier corresponding to the de-duplicated segmented audio and the second index identifier corresponding to the de-duplicated image segment;

[0126] S260. Based on the first index identifier, the second index identifier, and the remaining audio-visual information that has not been deduplicated, determine the target index identifier corresponding to the audio-video to be deduplicated after the deduplication process.

[0127] Using the embodiments of this specification, perform separation and segmentation processing on the audio-video to be deduplicated to obtain at least one segmented audio and at least one image segment; perform feature extraction on each segmented audio and each image segment respectively to obtain at least one segmented audio feature vector and at least one image segment feature vector; for each segmented audio feature vector, determine the first similarity with the source audio feature vector respectively, and perform deduplication processing on the segmented audio corresponding to the segmented audio feature vector when the first similarity meets the first preset condition. Similarly, perform deduplication processing on the image segment corresponding to the image segment feature vector whose second similarity with the source image segment feature vector meets the second preset condition. Furthermore, for the audio-video to be deduplicated after the deduplication process, determine the first index identifier and the second index identifier for constructing the target index identifier together with the remaining audio-visual information that has not been deduplicated. Thus, it realizes segmenting the audio and image segments in the audio-video to be deduplicated, and respectively performing deduplication recognition and processing on the segmented audio and image segments, increasing the depth of the deduplication process, further releasing the storage space, and improving the utilization rate of the storage space.

[0128] According to an embodiment of this specification, the audio-video to be deduplicated is a short video published on a software platform after a user creates it through a user terminal. It should be noted that the audio-video to be deduplicated can also be, for example, a video in a film and television drama work. The segmented audio is a sub-audio obtained by segmenting the entire audio in the sound track included in the audio-video to be deduplicated. The image segment is a sub-pure video obtained by segmenting the entire pure video (without sound) in the pure video track included in the audio-video to be deduplicated.

[0129] Track separation is performed on the video to be de-accented to obtain the audio of the audio track and the pure video of the pure video track. The audio and the pure video are segmented to obtain at least one segmented audio obtained by segmenting the audio and at least one image segment obtained by segmenting the pure video. The track separation is implemented based on any existing method that can separate audio and pure video from audio and video. Segmenting the audio may include, for example, identifying the audio correlation between the audio corresponding to each moment and the audio of the previous moment. If the audio correlation between the audio of the current moment and the audio of the previous moment is greater than or equal to the audio correlation threshold, the audio of the current moment and the audio of the previous moment are assigned to the same segmented audio. If the audio correlation between the audio of the current moment and the audio of the previous moment is less than the audio correlation threshold, the audio of the current moment and the audio of the previous moment are segmented to obtain another segmented audio. That is, after audio segmentation, each segmented audio in at least one segmented audio is an audio that can express complete information. For example, when the entire audio includes "I ate today and the vegetables germinated", two segmented audios can be obtained, specifically, "I ate today" and "the vegetables germinated".

[0130] Segmenting the pure video may include, for example, identifying the correlation between the image corresponding to each moment and the image of the previous moment. If the image correlation between the image of the current moment and the image of the previous moment is greater than or equal to the image correlation threshold, the image of the current moment and the image of the previous moment are assigned to the same image segment. If the correlation between the image of the current moment and the image of the previous moment is less than the image correlation threshold, the image of the current moment and the image of the previous moment are segmented to obtain another image segment. That is, after segmenting the pure video, each image segment in at least one image segment is a continuous sub-pure video. For example, in a 60-second pure video, from 1 to 9 seconds is "advertisement" and from 10 to 60 seconds is "TV drama or movie", two image segments can be obtained, specifically, "advertisement" and "TV drama or movie". It should be noted that during segmentation, it can be segmented based on the category of the video or based on whether the images are continuous. For example, the advertisement includes "advertisement for selling product B by A" and "advertisement for selling product D by C", that is, the advertisement can be segmented into "advertisement for selling product B by A" and "advertisement for selling product D by C".

[0131] Using a feature tracking extractor and a feature vector generator, each segmented audio and each image segment after segmented processing are processed respectively to obtain a segmented audio feature vector corresponding to each segmented audio and an image segment feature vector corresponding to each image segment. The feature tracking extractor and the feature vector generator process the segmented audio, and the obtained segmented audio feature vector may include, for example, feature data indicating audio length, audio type, audio start and end times, audio encoding method, and audio address, etc. The feature tracking extractor and the feature vector generator process the image segment, and the obtained image segment feature vector may include, for example, feature data indicating pure video length, type, start and end times, encoding method, and pure video address, etc.

[0132] The source audio feature vector is obtained by performing feature extraction on the source audio. This feature extraction can be implemented, for example, based on the feature tracking extractor and the feature vector generator. The source image segment feature vector is obtained by performing feature extraction on the source image segment. This feature extraction can be implemented, for example, based on the feature tracking extractor and the feature vector generator. The source audio is a preset audio belonging to the same audio category as the segmented audio. The source image segment is a preset video belonging to the same video category as the image segment. The same video category can be, for example, a movie and TV drama identifier, a protagonist identifier, or a performance category identifier, etc.

[0133] For each audio feature vector, using a similarity calculation formula, the audio feature and each source audio feature vector are processed to obtain a plurality of first similarities. Each first similarity is compared with a first threshold. When it is determined that at least one of the plurality of first similarities is greater than or equal to the first threshold, it is determined that the first similarity greater than or equal to the first threshold satisfies a first preset condition. Otherwise, it does not satisfy the first preset condition. Furthermore, duplicate removal processing is performed on the segmented audio corresponding to the first similarity that satisfies the first preset condition, and a first index identifier corresponding to the segmented audio is determined. The segmented audio corresponding to the first similarity that does not satisfy the first preset condition is used as the remaining audio that has not been subjected to duplicate removal processing.

[0134] The duplicate removal processing can be, for example, an operation of deleting the segmented audio or image segment in the storage space.

[0135] Similarly, for each image segment feature vector, using the similarity calculation formula, the image segment feature and each source image segment feature vector are processed to obtain multiple second similarities. Each second similarity is compared with a second threshold. When it is determined that at least one of the multiple second similarities is greater than or equal to the second threshold, it is determined that the second similarity greater than or equal to the second threshold meets the second preset condition. Otherwise, it does not meet the second preset condition. Furthermore, duplicate removal processing is performed on the image segment corresponding to the second similarity that meets the second preset condition, and a second index identifier corresponding to the image segment is determined. The image segments corresponding to the second similarities that do not meet the second preset condition are used as the remaining image segments that have not been subject to duplicate removal processing.

[0136] The remaining audio and video information that has not been subject to duplicate removal processing may include, for example, the remaining audio that has not been subject to duplicate removal processing and the remaining image segments that have not been subject to duplicate removal processing. The similarity calculation formula may be, for example, any formula for determining the distance between two vectors, such as, for example, the cosine similarity formula.

[0137] After determining the first index identifier, the second index identifier, the remaining audio that has not been subject to duplicate removal processing, and the remaining image segments that have not been subject to duplicate removal processing, according to the order of appearance times of the segmented audio and segmented video corresponding to the first index identifier, the second index identifier, the remaining audio that has not been subject to duplicate removal processing, and the remaining image segments that have not been subject to duplicate removal processing in the audio and video to be de-duplicated, a target index identifier indicating the first index identifier, the second index identifier, the remaining audio that has not been subject to duplicate removal processing, and the remaining image segments that have not been subject to duplicate removal processing is determined. Based on the target index identifier, the segmented audio and image segments corresponding to each moment or each duration can be indexed to determine the audio and video to be de-duplicated.

[0138] According to another embodiment of this specification, the determination of the first index identifier includes: for the segmented audio subject to duplicate removal processing, determining the first index identifier of the target source audio corresponding to the segmented audio, where the target source audio is the source audio corresponding to the source audio feature vector whose first similarity with the segmented audio feature vector meets the first preset condition, and the first index identifier points to the target source audio.

[0139] For example, when the first similarity between the source audio A and the segmented audio S is greater than the first threshold, and the first similarity between the source audio B and the segmented audio A is less than the first threshold, the source audio A is the target source audio. The first index identifier may be, for example, an identifier associated with the source audio A.

[0140] According to another embodiment of this specification, the determination of the second index identifier includes: for the image segment after duplicate removal processing, determining the second index identifier of the target source image segment corresponding to the image segment, where the target source image segment is the source image segment corresponding to the source image segment feature vector whose second similarity with the feature vector of the image segment meets the second preset condition, and the second index identifier points to the target source image segment.

[0141] For example, when the second similarity between the source image segment C and the image segment Y is greater than the second threshold, and the second similarity between the source image segment D and the image segment Y is less than the second threshold, the source image segment C is the target source image segment. The second index identifier can be, for example, an identifier associated with the source image segment C.

[0142] Figure 3A The figure shows a flowchart of a video duplicate removal processing method according to an embodiment of this specification. Figure 3C The figure shows a schematic diagram of a segmentation processing method according to an embodiment of this specification. A video duplicate removal process is described in this figure, but it may include more or fewer operation steps based on routine or non-creative labor. Specifically, as Figure 3A shown, the method may include:

[0143] S3110, performing segmentation processing on the video to be de-duplicated to obtain at least one piece of characterization information;

[0144] S3120, for each piece of characterization information, determining whether the storage space data of the characterization information meets the third preset condition;

[0145] S3130, when it is determined that the storage space data of the target characterization information meets the third preset condition, performing feature extraction on the target characterization information to obtain the target characterization information feature vector;

[0146] S3140, determining whether the third similarity between the source characterization information feature vector and the target characterization information feature vector meets the fourth preset condition

[0147] S3150, when the third similarity between the source characterization information feature vector and the target characterization information feature vector meets the fourth preset condition, performing duplicate removal processing on the target characterization information;

[0148] S3160, when it is determined that the storage space data of the target characterization information does not meet the third preset condition, or the third similarity between the source characterization information feature vector and the target characterization information feature vector does not meet the fourth preset condition, regarding the target characterization information as the remaining audio and video information.

[0149] Using the embodiments of this specification, in addition to audio and pure video, the de-accented video usually also includes characterization information. If the storage space occupied by the characterization information is large, it will also result in low utilization rate of the storage space.

[0150] According to another embodiment of this specification, the characterization information may include, for example, special effects and text.

[0151] For the de-accented video, after track separation, in addition to obtaining the audio of the audio track and the pure video of the pure video track, the entire characterization information of the characterization information track can also be obtained. For example, the special effects of the special effects track and the text of the text track. Then, the entire characterization information is segmented to obtain at least one piece of characterization information. The segmentation method is similar to the method of segmenting the entire audio or the entire pure video. As Figure 3C shown, for the de-accented video, separation processing is performed according to five tracks: track 1 video track, track 2 video track, track 3 video track, track 4 picture track, and track 5 audio track, to obtain the corresponding segmented audio (audio segment 1), image segments (video segments 1, 2, 3, 4, 5, and 6), and characterization information (continuum 1, 2, 3, 4, picture 1, and image 2).

[0152] For each piece of characterization information, determine whether the storage space data of the characterization information is greater than the storage space threshold. If it is determined that the storage space data of the characterization information is greater than the storage space threshold, determine that the storage space data meets the third preset condition. Otherwise, determine that the storage space data does not meet the third preset condition. The characterization information corresponding to the storage space data that meets the third preset condition is used as the target characterization information. The characterization information corresponding to the storage space data that does not meet the third preset condition is used as the remaining characterization information.

[0153] The process of extracting the target characterization information feature vector from the target characterization information is similar to the process of extracting features from the segmented audio or image segments above. The target characterization information feature vector may include, for example, feature data indicating the special effect length, special effect type, encoding method, etc.

[0154] The source characterization information feature vector is obtained by extracting features from the source characterization information. The source characterization information may be, for example, preset characterization information, such as preset special effects and preset text.

[0155] Using a similarity calculation formula, process the target representation information feature vector and each source representation information feature vector to obtain multiple third similarities. Compare each third similarity with each fourth threshold. When it is determined that at least one third similarity is greater than or equal to the fourth threshold, it is determined that the third similarity greater than or equal to the fourth threshold meets the fourth preset condition; otherwise, it is determined that the third similarity does not meet the fourth preset condition.

[0156] Perform deduplication processing on the target representation information corresponding to the third similarity that meets the fourth preset condition. This deduplication processing is an operation to delete the target representation information. Leave the remaining representation information of the target representation information corresponding to the third similarity that does not meet the fourth preset condition.

[0157] The remaining audio-visual information includes, for example, the remaining representation information in addition to the remaining audio and remaining image segments.

[0158] Figure 3B The following is a flowchart of a method for determining a target index identifier according to an embodiment of this specification. A process for determining a target index identifier is described in this figure, but based on routine or non-creative labor, it may include more or fewer operation steps. Specifically, as Figure 3B shown, the method may include:

[0159] S3250, for the audio-visual information to be deduplicated after the deduplication processing, determine the third index identifier corresponding to the target representation information to be deduplicated;

[0160] S3260, based on the first index identifier, the second index identifier, the third index identifier, and the remaining audio-visual information that has not been deduplicated, determine the target index identifier corresponding to the audio-visual information to be deduplicated after the deduplication processing.

[0161] According to another embodiment of this specification, determine the third index identifier for the target representation information to be deduplicated. The determination of the third index identifier may include, for example, for the target representation information subjected to the deduplication processing, determining the third index identifier of the target source representation information corresponding to the target representation information. The target source representation information is the source representation information corresponding to the source representation information feature vector whose third similarity determined with the target representation information feature vector meets the fourth preset condition, and the third index identifier points to the target source representation information.

[0162] Determine a target index identifier including the first index identifier, the second index identifier, the third index identifier, the remaining audio that has not been deduplicated, the remaining image segments that have not been deduplicated, and the remaining characterization information that has not been deduplicated according to the order of appearance of the segmented audio, segmented video, and characterization information corresponding to the first index identifier, the second index identifier, the third index identifier, the remaining audio that has not been deduplicated, the remaining image segments that have not been deduplicated, and the remaining characterization information that has not been deduplicated in the audio-video to be deduplicated.

[0163] Figure 4A The figure shows a flowchart of a first similarity determination method according to an embodiment of the present specification. A first similarity determination process is described in this figure, but based on routine or non-creative labor, it may include more or fewer operation steps. Specifically, as Figure 4A shown, the method may include:

[0164] S431, segment the preset audio according to the duration information of the segmented audio corresponding to the segmented audio feature vector to obtain a plurality of source audios;

[0165] S432, perform feature extraction on each source audio to obtain a plurality of source audio feature vectors;

[0166] S433, respectively determine sub-first segmented audio feature data and sub-first source audio feature data corresponding to each first dimension from the segmented audio feature vector and each source audio feature vector;

[0167] S434, determine the sub-first similarity between the sub-first segmented audio feature data and the sub-first source audio feature data under each first dimension;

[0168] S435, use the first weight data corresponding to each first dimension to perform weighted calculation on the sub-first similarity corresponding to each first dimension to obtain the first similarity.

[0169] According to another embodiment of the present specification, for each segmented audio, a preset audio is determined respectively based on the category information of the segmented audio. The category information is, for example, information that can index the preset audio from a preset audio library. For example, the category information is the World Cup event at E o'clock on A year B month C day.

[0170] For example, if the duration information of the segmented audio is F seconds, then based on the preset step size, a plurality of source audios of F seconds are segmented from the determined preset audio. The preset step size can be, for example, 1 second.

[0171] The process of performing feature extraction on each source audio to obtain a plurality of source audio feature vectors may be similar to the process of performing feature extraction on the segmented audio to obtain the segmented audio feature vector, for example.

[0172] The first dimension includes feature categories in the audio features. For example, in the segmented audio feature vector, there is feature data indicating the audio length, audio type, audio start and end times, audio encoding method, and audio address. Then the first dimension includes the audio length, audio type, audio start and end times, audio encoding method, and audio address.

[0173] For example, the source audio feature vector is (a, s, d, f, g, h, j), and the segmented audio feature vector is (z, x, c, v, b, n, m). Among them, a and z are feature data corresponding to the audio length, s, d, x, and c are feature data corresponding to the audio type, f, g, v, and b are feature data corresponding to the audio start and end times, h and n are feature data corresponding to the audio encoding method, and j and m are feature data corresponding to the audio address.

[0174] Determine the sub-first similarity between a and z, determine the sub-first similarity between s, d and x, c, determine the sub-first similarity between f, g and v, b, determine the sub-first similarity between h and n, and determine the sub-first similarity between j and m. And perform weighted calculation on multiple sub-first similarities based on the first weight data corresponding to the audio length, audio type, audio start and end times, audio encoding method, and audio address respectively to obtain the first similarity.

[0175] Figure 4B The figure shows a flowchart of a method for determining a second similarity according to an embodiment of the present specification. A process for determining a second similarity is described in this figure, but it may include more or fewer operation steps based on routine or non-creative labor. Specifically, as Figure 4B shown, the method may include:

[0176] S441, according to the duration information of the image segment corresponding to the image segment feature vector, perform segmentation processing on the preset video to obtain multiple source image segments;

[0177] S442, perform feature extraction on each source image segment to obtain multiple source image segment feature vectors;

[0178] S443, respectively determine the sub-second image segment feature data and sub-second source image segment feature data corresponding to each second dimension from the image segment feature vector and each source image segment feature vector;

[0179] S444, under each second dimension, determine the sub-second similarity between the sub-second image segment feature data and the sub-second source image segment feature data;

[0180] S445, use the second weight data corresponding to each second dimension to perform weighted calculation on the sub-second similarity corresponding to each second dimension to obtain the second similarity.

[0181] According to another embodiment of the present specification, based on Figure 4A performing processing on the image segment feature vector and the preset video in a manner similar to that included in S431 - S435 to obtain a second similarity.

[0182] It should be noted that for each image segment, a preset video is determined respectively based on the category information of the image segment.

[0183] Figure 5A The figure shows a schematic diagram of a video extraction and display method according to an embodiment of the present specification. Figure 5B The figure shows a schematic diagram of a video storage and video extraction and display method according to an embodiment of the present specification. In this figure, a video extraction and display process is described, but based on routine or non - creative labor, it may include more or fewer operation steps. Specifically, as Figure 5A shown, the method may include:

[0184] S510, determining a plurality of sub - index identifiers and sub - audio - video index information according to the received target index identifier;

[0185] S520, respectively determining the corresponding source image segment, source audio, or source characterization information for each sub - index identifier;

[0186] S530, determining a target video for display based on the source image segment, source audio, source characterization information, and the sub - audio - video corresponding to the sub - audio - video index information.

[0187] When using the embodiment of the present specification to extract the de - duplicated video currently, usually directly according to the index identifier corresponding to the video, the entire video is directly indexed. And based on this, the storage space utilization rate during the storage process of the de - duplicated video is relatively low, and the specific reasons have been described above.

[0188] According to another embodiment of the present specification, when a user wants to watch a certain stored video, click or select the identifier corresponding to the video on the user terminal to control the user terminal to send a target index identifier to the server.

[0189] When the server receives the target index identifier, it splits the target index identifier to obtain multiple sub-index identifiers and sub-audio / video index information. The sub-index identifier can be, for example, the first index identifier, the second index identifier, and the third index identifier mentioned above, which are respectively associated with the source audio corresponding to the segmented audio after deduplication processing, the source image segment corresponding to the image segment after deduplication processing, and the source representation information corresponding to the target representation information after deduplication processing. Specifically, for example, the first index identifier, the second index identifier, and the third index identifier are associated with the target source audio, the target source image segment, and the target source representation information mentioned above. The sub-audio / video index information is, for example, information associated with the remaining audio, the remaining image segments, and the remaining representation information that have not been deduplicated.

[0190] Based on the arrangement order of the first index identifier, the second index identifier, the third index identifier, and the sub-audio / video index information in the target index identifier, the source audio, the source image segment, the source representation information, and the sub-audio / video are spliced to obtain a target video, and the target video is sent to the user terminal for display. The sub-audio / video can include, for example, the remaining audio, the remaining image segments, and the remaining representation information that have not been deduplicated.

[0191] It should be noted that the target index identifier can be, for example, the target index identifier determined based on Figure 2 the steps S210 - S260. Specifically, as shown in Figure 5B .

[0192] As shown in Figure 5B , after the sub-audio / video 501 to be deduplicated is processed by splitting tracks and segments, multiple segmented audios 502, multiple image segments 503, and multiple representation information 504 are obtained.

[0193] For each segmented audio in the multiple segmented audios 502, the first similarity with the source segmented audio in the preset audio is determined respectively. The segmented audio corresponding to the first similarity that meets the first preset condition is used as the deduplicated segmented audio 511. Deduplication processing is performed on the deduplicated segmented audio 511, and the identifier corresponding to the source segmented audio corresponding to the first similarity that meets the first preset condition is determined as the first index identifier 531. In addition, the segmented audio corresponding to the first similarity that does not meet the first preset condition is used as the remaining audio 521.

[0194] Similarly, for each of the multiple image segments 503, the second similarity with the source image segment in the preset video is determined respectively. The image segment corresponding to the second similarity that meets the second preset condition is used as the duplicate-removal processed image segment 512. Duplicate removal processing is performed on this duplicate-removal processed image segment 512, and the identifier corresponding to the source image segment corresponding to the second similarity that meets the second preset condition is determined as the second index identifier 532. In addition, the image segments corresponding to the second similarities that do not meet the second preset condition are used as the remaining image segments 522.

[0195] Similarly, for each of the multiple characterization information 504, it is determined whether the storage space data of this characterization information meets the third preset condition. When it is determined that the third preset condition is met, the third similarity with each source characterization information is determined respectively. The target characterization information corresponding to the third similarity that meets the fourth preset condition is used as the duplicate-removal processed target characterization information 513. Duplicate removal processing is performed on this duplicate-removal processed target characterization information 513, and the identifier corresponding to the source characterization information corresponding to the third similarity that meets the fourth preset condition is determined as the third index identifier 533. In addition, the characterization information corresponding to the third similarities that do not meet the fourth preset condition is used as the remaining characterization information 523.

[0196] Thereafter, based on the occurrence time sequence of each segmented audio 502, each image segment 503, and each characterization information 504 in the audio-video to be duplicate-removed 501, sorting and processing are performed on the first index identifier 531, the remaining audio 521, the second index identifier 532, the remaining image segments 522, the third index identifier 533, and the remaining characterization information 523 to construct the target index identifier 540.

[0197] Figure 6A The following shows a schematic structural diagram of a video storage device according to an embodiment of this specification. As Figure 6A shown, it includes

[0198] A segmentation unit 6010, configured to perform segmentation processing on the audio-video to be duplicate-removed to obtain at least one segmented audio and at least one image segment;

[0199] An extraction unit 6020, configured to perform feature extraction on each segmented audio and each image segment to obtain at least one segmented audio feature vector and at least one image segment feature vector;

[0200] A first duplicate-removal unit 6030, configured to perform duplicate removal processing on the segmented audio corresponding to the segmented audio feature vector when the first similarity between the source audio feature vector and the segmented audio feature vector meets the first preset condition, where the source audio feature vector is obtained by performing feature extraction on the source audio;

[0201] A second duplicate removal unit 6040, configured to perform duplicate removal processing on an image segment corresponding to an image segment feature vector when a second similarity between a source image segment feature vector and the image segment feature vector meets a second preset condition, where the source image segment feature vector is obtained by performing feature extraction on a source image segment;

[0202] A first determination unit 6050, configured to determine a first index identifier corresponding to the segmented audio to be de-duplicated and a second index identifier corresponding to the image segment to be de-duplicated for the audio-video to be de-duplicated after the duplicate removal processing; and

[0203] A second determination unit 6060, configured to determine a target index identifier corresponding to the audio-video to be de-duplicated after the duplicate removal processing based on the first index identifier, the second index identifier, and remaining audio-video information that has not been subjected to duplicate removal processing.

[0204] Since the principle of the above device for solving problems is similar to that of the above method, the implementation of the above device can refer to the implementation of the above method, and the repeated parts will not be elaborated.

[0205] Figure 6B The following shows a schematic structural diagram of a video storage device according to another embodiment of this specification. As Figure 6B shown, it includes

[0206] A characterization segmentation unit 6071, configured to perform segmentation processing on the audio-video to be de-duplicated to obtain at least one piece of characterization information;

[0207] A judgment unit 6072, configured to judge whether the storage space data of each piece of characterization information meets a third preset condition;

[0208] A characterization feature extraction unit 6073, configured to perform feature extraction on target characterization information to obtain a target characterization information feature vector when it is determined that the storage space data of the target characterization information meets the third preset condition;

[0209] A characterization duplicate removal unit 6074, configured to perform duplicate removal processing on the target characterization information when a third similarity between a source characterization information feature vector and the target characterization information feature vector meets a fourth preset condition, where the source characterization information feature vector is obtained by performing feature extraction on source characterization information,

[0210] and;

[0211] An identifier determination unit 6075, configured to determine a third index identifier corresponding to the target characterization information to be de-duplicated for the audio-video to be de-duplicated after the duplicate removal processing; and

[0212] A storage unit 6076 is configured to determine a target index identifier corresponding to the duplicate-eliminated audio-video to be duplicate-eliminated based on a first index identifier, a second index identifier, a third index identifier, and the remaining audio-video information that has not been duplicate-eliminated.

[0213] Figure 6C The following is a schematic structural diagram of a video extraction and display device according to an embodiment of the present specification. As Figure 6C shown, it includes

[0214] A third determination unit 6080 is configured to determine a plurality of sub-index identifiers and sub-audio-video index information according to the received target index identifier;

[0215] A fourth determination unit 6090 is configured to respectively determine corresponding source image segments, source audio, or source characterization information for each sub-index identifier;

[0216] A fifth determination unit 6110 is configured to determine a target video for display based on the source image segments, source audio, source characterization information, and the sub-audio-video corresponding to the sub-audio-video index information.

[0217] Since the principle of the above device for solving problems is similar to the above method, the implementation of the above device can refer to the implementation of the above method, and the repeated parts will not be elaborated.

[0218] As Figure 7 shown, the following is a schematic structural diagram of a computer device according to an embodiment of the present specification. The device in the present specification may be the computer device in this embodiment and execute the method in the present specification. The computer device 702 may include one or more processing devices 704, such as one or more central processing units (CPUs), and each processing unit may implement one or more hardware threads. The computer device 702 may also include any storage resource 706, which is used to store any type of information such as code, settings, data, etc. Non-limitingly, for example, the storage resource 706 may include any one or more combinations of the following: any type of RAM, any type of ROM, flash memory devices, hard disks, optical discs, etc. More generally, any storage resource may use any technology to store information. Further, any storage resource may provide volatile or non-volatile retention of information. Further, any storage resource may represent a fixed or removable component of the computer device 702. In one case, when the processing device 704 executes the associated instructions stored in any storage resource or combination of storage resources, the computer device 702 may perform any operation of the associated instructions. The computer device 702 also includes one or more drive mechanisms 708 for interacting with any storage resource, such as a hard disk drive mechanism, an optical disc drive mechanism, etc.

[0219] The computer device 702 may further include an input / output module 710 (I / O) for receiving various inputs (via the input device 712) and for providing various outputs (via the output device 714). A specific output mechanism may include a presentation device 716 and an associated graphical user interface (GUI) 718. In other embodiments, the input / output module 710 (I / O), the input device 712, and the output device 714 may not be included, and it may only be a computer device in the network. The computer device 702 may further include one or more network interfaces 720 for exchanging data with other devices via one or more communication links 722. One or more communication buses 724 couple the components described above together.

[0220] The communication link 722 may be implemented in any manner, for example, through a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 722 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc. governed by any protocol or combination of protocols.

[0221] Embodiments of this specification also provide a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the above method is implemented.

[0222] Embodiments of this specification also provide a computer program product including a computer program, and when the computer program is executed by a processor, the above method is implemented.

[0223] Those skilled in the art should understand that the embodiments of this specification may be provided as a method, a system, or a computer program product. Therefore, this specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. 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.

[0224] This specification is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for implementing in the process Figure 1 one process or multiple processes and / or blocksFigure 1 means for the functions specified in one or more boxes.

[0225] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one Figure 1 process or more processes and / or boxes Figure 1 or more boxes.

[0226] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one Figure 1 process or more processes and / or boxes Figure 1 or more boxes.

[0227] In the above specific embodiments, the purpose, technical solutions and beneficial effects of this specification have been further described in detail. It should be understood that the above are only specific embodiments of this specification and are not used to limit the protection scope of this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of this specification shall be included in the protection scope of this specification.

Claims

1. A video storage method, characterized in that, Including: Performing segmentation processing on the video with accent to be removed, to obtain at least one segmented audio and at least one image segment; Performing feature extraction on each of the segmented audios and each of the image segments, to obtain at least one segmented audio feature vector and at least one image segment feature vector; For each of the segmented audio feature vectors, when the first similarity between the source audio feature vector and the segmented audio feature vector meets the first preset condition, performing duplicate removal processing on the segmented audio corresponding to the segmented audio feature vector, where the source audio feature vector is obtained by performing feature extraction on the source audio; For each of the image segment feature vectors, when the second similarity between the source image segment feature vector and the image segment feature vector meets the second preset condition, performing duplicate removal processing on the image segment corresponding to the image segment feature vector, where the source image segment feature vector is obtained by performing feature extraction on the source image segment; For the video with accent to be removed after the duplicate removal processing, determining a first index identifier corresponding to the segmented audio that has been removed from duplicates and a second index identifier corresponding to the image segment that has been removed from duplicates; And Based on the first index identifier, the second index identifier, and the remaining audio-visual information that has not been subjected to duplicate removal processing, determining a target index identifier corresponding to the video with accent to be removed after the duplicate removal processing.

2. The method according to claim 1, wherein Before performing feature extraction on each of the segmented audios and each of the image segments to obtain at least one segmented audio feature vector and at least one image segment feature vector, it further includes: Performing segmentation processing on the video with accent to be removed, to obtain at least one piece of characterization information; For each piece of characterization information, determining whether the storage space data of the characterization information meets the third preset condition; When it is determined that the storage space data of the target characterization information meets the third preset condition, performing feature extraction on the target characterization information to obtain a target characterization information feature vector; When the third similarity between the source characterization information feature vector and the target characterization information feature vector meets the fourth preset condition, performing duplicate removal processing on the target characterization information, where the source characterization information feature vector is obtained by performing feature extraction on the source characterization information; The determining of the target index identifier corresponding to the video with accent to be removed after the duplicate removal processing based on the first index identifier, the second index identifier, and the remaining audio-visual information that has not been subjected to duplicate removal processing further includes: For the video with accent to be removed after the duplicate removal processing, determining a third index identifier corresponding to the target characterization information that has been removed from duplicates; and Based on the first index identifier, the second index identifier, the third index identifier, and the remaining audio-visual information that has not been subjected to duplicate removal processing, determining a target index identifier corresponding to the video with accent to be removed after the duplicate removal processing.

3. The method according to claim 1, wherein The determination of the first similarity between the source audio feature vector and the segmented audio feature vector includes: According to the duration information of the segmented audio corresponding to the segmented audio feature vector, performing segmentation processing on the preset audio to obtain a plurality of source audios; Performing feature extraction on each of the source audios to obtain a plurality of the source audio feature vectors; From the segmented audio feature vectors and each of the source audio feature vectors, respectively determine sub-first segmented audio feature data and sub-first source audio feature data corresponding to each first dimension; Under each of the first dimensions, determine a sub-first similarity between the sub-first segmented audio feature data and the sub-first source audio feature data; and Use first weight data corresponding to each of the first dimensions to perform a weighted calculation on the sub-first similarities corresponding to each of the first dimensions to obtain the first similarity.

4. The method according to claim 1, wherein The determination of the second similarity between the source image segment feature vector and the image segment feature vector includes: According to the duration information of the image segment corresponding to the image segment feature vector, perform segmentation processing on a preset video to obtain a plurality of source image segments; Perform feature extraction on each of the source image segments to obtain a plurality of the source image segment feature vectors; From the image segment feature vector and each of the source image segment feature vectors, respectively determine sub-second image segment feature data and sub-second source image segment feature data corresponding to each second dimension; Under each of the second dimensions, determine a sub-second similarity between the sub-second image segment feature data and the sub-second source image segment feature data; and Use second weight data corresponding to each of the second dimensions to perform a weighted calculation on the sub-second similarities corresponding to each of the second dimensions to obtain the second similarity.

5. The method according to claim 1, characterized in that The determination of the first index identifier includes: For the segmented audio after duplicate removal processing, determine the first index identifier of the target source audio corresponding to the segmented audio, where the target source audio is the source audio corresponding to the source audio feature vector whose first similarity determined with the segmented audio feature vector satisfies the first preset condition, and the first index identifier points to the target source audio.

6. The method according to claim 1, wherein The determination of the second index identifier includes: For the image segment after duplicate removal processing, determine the second index identifier of the target source image segment corresponding to the image segment, where the target source image segment is the source image segment corresponding to the source image segment feature vector whose second similarity determined with the image segment feature vector satisfies the second preset condition, and the second index identifier points to the target source image segment.

7. A video extraction and display method, characterized in that, For a video stored by the method according to any one of claims 1-6, the video extraction and display method includes: According to the received target index identifier, determine a plurality of sub-index identifiers and sub-audio-video index information; For each of the sub-index identifiers, respectively determine the corresponding source image segment or source audio or source characterization information; Based on the source image segment, source audio, source characterization information, and the sub-audio-video corresponding to the sub-audio-video index information, determine a target video for display.

8. A video storage device, characterized in that, Includes: A segmentation unit for performing segmentation processing on a to-be-duplicate-removed audio-video to obtain at least one segmented audio and at least one image segment; An extraction unit for performing feature extraction on each of the segmented audio and each of the image segments to obtain at least one segmented audio feature vector and at least one image segment feature vector; A first deduplication unit, configured to perform deduplication processing on each of the segmented audio feature vectors for the segmented audio corresponding to the segmented audio feature vector when a first similarity between the source audio feature vector and the segmented audio feature vector meets a first preset condition, where the source audio feature vector is obtained by performing feature extraction on a source audio; A second deduplication unit, configured to perform deduplication processing on each of the image segment feature vectors for the image segment corresponding to the image segment feature vector when a second similarity between the source image segment feature vector and the image segment feature vector meets a second preset condition, where the source image segment feature vector is obtained by performing feature extraction on a source image segment; A first determination unit, configured to determine a first index identifier corresponding to the segmented audio that has been deduplicated and a second index identifier corresponding to the image segment that has been deduplicated for the audio-visual content to be deduplicated after the deduplication processing; and A second determination unit, configured to determine a target index identifier corresponding to the audio-visual content to be deduplicated after the deduplication processing based on the first index identifier, the second index identifier, and the remaining audio-visual information that has not been subjected to the deduplication processing.

9. A video extraction and display device, characterized in that, For a video stored according to the method described in any one of claims 1-6, the video extraction and display device includes: A third determination unit, configured to determine a plurality of sub-index identifiers and sub-audio-visual index information according to the received target index identifier; A fourth determination unit, configured to respectively determine a corresponding source image segment, source audio, or source characterization information for each of the sub-index identifiers; A fifth determination unit, configured to determine a target video for display based on the source image segment, source audio, source characterization information, and the sub-audio-visual content corresponding to the sub-audio-visual index information.

10. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method described in any one of claims

Images