A method for recommending multimedia resources and related equipment

Abstract

The present application discloses a method for recommending multimedia resources and related equipment, the recommendation method comprising obtaining the playback progress of a target video, and determining the hot clips corresponding to the target video according to the playback progress, determining the recommendation dimension corresponding to the target video based on the hot clips and a trained video understanding model, determining the recommended resources corresponding to the target video according to the recommended dimensions, and recommending the determined recommended resources. The embodiment of the present application determines the hot clips in the target video based on the playback progress of the target video being played, identifies and understands the hot clips through a trained video understanding model, and finally determines the recommended video based on the identified recommendation dimensions, which can improve the relevance between the video content of the recommended video and the video content of the target video, thereby improving the accuracy of the recommended video, thereby improving the video recommendation effect.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method for recommending multimedia resources and related equipment. Background Art

[0002] Smart TVs are already common terminal devices in households, and video on demand is the core function and application scenario of TVs. When users use the video on demand function, manufacturers will actively recommend videos to users, such as video recommendations based on user behavior habits and video recommendations based on the source association of film details. However, the existing video recommendation methods do not highly correlate the video content of the video that the user is watching, so that the recommended videos cannot match the user's current needs, resulting in poor recommendation results. Summary of the invention

[0003] The technical problem to be solved by the present application is to provide a multimedia resource recommendation method and related equipment in view of the deficiencies of the prior art, so as to solve the problem that the video content of the existing recommended video is not highly correlated with the target video content.

[0004] In order to solve the above technical problems, the first aspect of the embodiments of the present application provides a method for recommending multimedia resources, the method comprising:

[0005] Obtain the playback progress of the target video, and determine the hot clips corresponding to the target video according to the playback progress;

[0006] Determine the recommended dimension corresponding to the target video based on the popular clips and the trained video understanding model;

[0007] Determine the recommended resource corresponding to the target video according to the recommendation dimension, and recommend the determined recommended resource.

[0008] In one embodiment, the hot segment is included in the target video, and a reference playback progress of the hot segment is located before the playback progress.

[0009] In one embodiment, the video understanding model is configured with a plurality of default dimensions; and determining the recommended dimensions corresponding to the target video based on the popular clips and the trained video understanding model comprises:

[0010] Inputting the hot clips into the video understanding model, and outputting the dimensional content of the hot clips in each default dimension through the video understanding model;

[0011] The recommended dimensions corresponding to the target video are generated based on the dimensional contents of the popular clips in each default dimension.

[0012] In one embodiment, the default dimensions include at least information about the film source, actor roles, architecture, shooting techniques, and film type.

[0013] In one embodiment, determining the recommended resource corresponding to the target video according to the recommendation dimension and recommending the determined recommended resource specifically includes:

[0014] Sending the recommended dimensions to the media library server, and controlling the media library server to obtain the user preference of the user corresponding to the target video;

[0015] The media resource library server is controlled to determine the recommended resources corresponding to the target video based on the user preference and the recommendation dimension, and recommend the determined recommended resources.

[0016] In one embodiment, the controlling the media resource library server to determine the recommended resource corresponding to the target video based on the user preference and the recommendation dimension, and recommending the determined recommended resource specifically includes:

[0017] Controlling the media asset library server to determine a first candidate recommendation resource based on the user preference and the recommendation dimension;

[0018] Controlling the media resource library server to obtain a second candidate recommendation resource formed based on the hot segment;

[0019] The first candidate recommended resource and the second candidate recommended resource are used as recommended resources corresponding to the target video, and the determined recommended resource is recommended.

[0020] In one embodiment, after determining the recommended resource corresponding to the target video according to the recommendation dimension and recommending the determined recommended resource, the method further includes:

[0021] Monitoring the recommended duration of the recommended resource;

[0022] When the recommended duration reaches the preset duration, the step of obtaining the playback progress of the target video and determining the hot clips corresponding to the target video according to the playback progress is performed again to update the recommended resource.

[0023] A second aspect of an embodiment of the present application provides a multimedia resource recommendation device, the recommendation device comprising:

[0024] An acquisition module is used to acquire the playback progress of the target video and determine the hot clips corresponding to the target video according to the playback progress;

[0025] A determination module, used to determine the recommendation dimension corresponding to the target video based on the popular clips and the trained video understanding model;

[0026] The recommendation module is used to determine the recommended resources corresponding to the target video according to the recommendation dimension, and recommend the determined recommended resources.

[0027] A computer-readable storage medium stores one or more programs, wherein the one or more programs can be executed by one or more processors to implement the steps in any of the above-mentioned methods for recommending multimedia resources.

[0028] A terminal device comprises: a processor, a memory and a communication bus; the memory stores a computer-readable program executable by the processor;

[0029] The communication bus realizes the connection and communication between the processor and the memory;

[0030] When the processor executes the computer-readable program, the processor implements the steps in any of the above-mentioned methods for recommending multimedia resources.

[0031] Beneficial effects: The present application determines the popular segments in the target video based on the playback progress of the target video being played, identifies and understands the popular segments through a trained video understanding model, and finally determines the recommended video based on the recommended dimensions obtained by the identification. This can improve the relevance of the video content of the recommended video with the video content of the target video, thereby improving the accuracy of the recommended video and thus improving the video recommendation effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without inventive work.

[0033] Figure 1 A flowchart of a method for recommending multimedia resources provided in this application.

[0034] Figure 2 A specific flow chart of a multimedia resource recommendation method provided for this application

[0035] Figure 3 This is a structural principle diagram of a multimedia resource recommendation device provided in this application.

[0036] Figure 4 This is a schematic diagram of the structure of the terminal device provided in this application. DETAILED DESCRIPTION

[0037] The present application provides a multimedia resource recommendation method and related equipment. To make the purpose, technical solution and effect of the present application clearer and more specific, the present application is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0038] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0039] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as here.

[0040] It should be understood that the sequence numbers and sizes of the steps in this embodiment do not mean the order of execution. The execution order of each process is determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0041] After research, the inventor found that smart TVs are already common terminal devices in households, and video on demand is the core function and application scenario of TVs. When users use the video on demand function, manufacturers will actively recommend videos to users, such as video recommendations based on user behavior habits and video recommendations based on the source association of film details. However, the existing video recommendation method does not highly correlate the video content of the video that the user is watching, so that the recommended video cannot match the user's current needs, resulting in poor recommendation results.

[0042] In order to solve the above problems, in an embodiment of the present application, the playback progress of the target video is obtained, and the hot clips corresponding to the target video are determined according to the playback progress. Based on the hot clips and a trained video understanding model, the recommendation dimension corresponding to the target video is determined. Based on the recommendation dimension, the recommended resource corresponding to the target video is determined, and the determined recommended resource is recommended. The embodiment of the present application determines the hot clips in the target video based on the playback progress of the target video being played, identifies and understands the hot clips through a trained video understanding model, and finally determines the recommended video based on the identified recommendation dimension. This can improve the relevance between the video content of the recommended video and the video content of the target video, thereby improving the accuracy of the recommended video, thereby improving the video recommendation effect. 。

[0043] The application content is further explained below through the description of embodiments in conjunction with the accompanying drawings.

[0044] This embodiment provides a method for recommending multimedia resources. Figure 1 and Figure 2 As shown, the method includes:

[0045] S100: Obtain the playback progress of the target video, and determine the hot clips corresponding to the target video according to the playback progress.

[0046] Specifically, the target video is a video resource in an online video media library, such as Tencent Video APP, Youku APP, iQiyi APP, etc. The playback progress is the current playback position of the target video, that is, the playback progress is the playback progress of the target video when the multimedia resource recommendation method provided in this embodiment is executed. For example, when the multimedia resource recommendation method provided in this embodiment is executed, the target video is played to the 25th minute, then the playback progress is the 25th minute of the target video. The popular clip is included in the target video, and the reference playback progress of the popular clip is located before the playback progress, that is, the popular clip is located on the reference playback progress, and the reference playback progress is located before the current playback position.

[0047] In one implementation of this embodiment, the hot clip may be a clip with a number of viewers greater than a preset number threshold, or a clip selected by the user, or a clip with a preset building or scene, or a clip with the most viewers in the video where the hot clip is located, etc. For example, if the video duration of the target video is 120 minutes, and the 20th to 25th minutes of the target video are the clips with the most viewers in the target video, then the 20th to 25th minutes of the target video may be the hot clip of the target video. In addition, when determining the hot clip based on the playback progress, one hot clip may be obtained, or multiple hot clips may be obtained; when multiple hot clips are obtained, the hot clip closest to the playback progress may be selected as the hot clip corresponding to the target video under the playback progress; or multiple hot clips may be displayed to the user, and the hot clip corresponding to the target video under the playback progress may be determined based on the user's selection.

[0048] In an implementation of this embodiment, the target video may be determined and played based on a user's on-demand operation, or may be determined and played through a control instruction sent by an external device, etc. In a typical implementation, the target video may be determined and played based on a user's on-demand operation, that is, before obtaining the playback progress of the target video and determining the hot clips corresponding to the target video according to the playback progress, the method may also include a target video on-demand process, and the target video on-demand process may specifically include opening an online video playback software, obtaining online video media library data, wherein the online video media library data is the film details content, such as the film title, actors, plot summary, number of episodes, etc., and the online video media library data is displayed on the online video playback software interface; select a target video on the online video playback software interface to enter the target video details; click on the video on-demand on the target video details interface, and then play the target video.

[0049] S200: Determine the recommendation dimension corresponding to the target video based on the popular clips and the trained video understanding model.

[0050] Specifically, the video understanding model is a trained neural network model, the input item of the video understanding model is a video clip, and the output item is a recommended dimension. The video understanding model is configured with several default dimensions, and the default dimensions may include information about the source, actor roles, architecture, classical architecture, special architecture, shooting techniques, and film types, so that the video understanding model can identify the information about the source, actor roles, architecture, shooting techniques, and film types in the popular clips, wherein the information about the source may include the title, director, actor, and creative team, etc., the actor roles include peasants, elites, and immortals, etc., the plot may include ancient costumes, modern, and palace fighting, etc., the architecture may be mountains, seas, and high-rise buildings, etc., the classical architecture may include royal buildings, palaces, temples, mansions, and ancient cities, etc., the special architecture may be the Great Wall and monuments, etc., and the film type may be action films, horror films, comedies, science fiction films, and romance films, etc. Of course, in practical applications, the default dimensions configured by the video understanding model may also include other dimensions, such as historical eras, geographical locations, etc. In addition, during the use of the video understanding model, the default dimensions in the video understanding model can be expanded. When the default dimensions configured in the video understanding model are expanded, the video understanding model before the expansion can be used as the initial model, and then the video understanding model can be trained so that the trained video understanding module can recognize all the expanded default dimensions.

[0051] In an implementation of this embodiment, the video understanding model is configured with a plurality of default dimensions; and determining the recommended dimensions corresponding to the target video based on the popular clips and the trained video understanding model includes:

[0052] S201, inputting the hot clips into the video understanding model, and outputting the dimensional content of the hot clips in each default dimension through the video understanding model;

[0053] S202: Generate a recommended dimension corresponding to the target video based on the dimensional content of the popular clips in each default dimension.

[0054] Specifically, the dimensional content is a specific recognition result on each default dimension. For example, the default dimension is architecture, and the recommended content may be a high-rise building, or the default dimension is film source information, and the recommended content may be an actor. In addition, when identifying popular clips, the dimensional content on each default dimension may be identified, and the dimensional content on some default dimensions may also be identified. In other words, a popular clip may carry content on all default dimensions, or may carry content on some default dimensions. For example, the default dimensions may include film source information, actor roles, architecture, classical architecture, special architecture, shooting techniques, and film types. Through the video understanding model, the recommended content for the popular clip b in video a based on the film source information is actor A, and the recommended content on classical architecture is royal architecture.

[0055] The recommended dimension can be used as a query condition formed by the dimensional content of the hot clips on each default dimension, and the corresponding recommended resources are found in the video resources based on the recommended dimension. In one implementation, the recommended dimension is a search statement formed based on the dimensional content of the hot clips on each default dimension, wherein the search statement includes the dimensional content on each default dimension. For example, the hot clip b in video a has recommended content including actor A and royal buildings through the video understanding model, so the recommended dimension can be the royal building scene of the ancient costume character played by actor A.

[0056] S300: Determine the recommended resource corresponding to the target video according to the recommendation dimension, and recommend the determined recommended resource.

[0057] Specifically, the recommended resource is a recommended video clip retrieved by using the recommendation dimension as a search statement. There may be one, multiple, or no retrieval result for the recommended video clip. It is understandable that the reason for the lack of recommended video retrieval results is that the recommended video source cannot be searched from the media resource library server based on the recommendation dimension of the popular clip. In order to facilitate users to view and broadcast recommended videos in a timely manner when watching the target video, the display area of ​​the recommended video is set on the same interface as the target video playback window and is adjacent to the target video playback window. The display area of ​​the recommended video can be located on the left, right, or top of the video playback window. When there are multiple recommended resources, they can be displayed in columns or rows.

[0058] In a typical implementation, the recommended resources are set on the right side of the target video playback window, as shown in the figure: the recommended resource display area is parallel to the target video playback window and is located on the right side of the target video playback window, and multiple recommended resources are arranged in columns in the recommended resource display area. A target video playback progress bar is displayed below the target video playback window. The target video playback progress bar marks the popular segments with colors to facilitate users to view the location of popular segments and to facilitate users to view or mark popular segments according to their preferences.

[0059] In an implementation of this embodiment, determining the recommended resource corresponding to the target video according to the recommendation dimension and recommending the determined recommended resource specifically includes:

[0060] S301, sending the recommended dimension to the media library server, and controlling the media library server to obtain the user preference of the user corresponding to the target video;

[0061] S302: Control the media resource library server to determine the recommended resources corresponding to the target video based on the user preference and the recommendation dimension, and recommend the determined recommended resources.

[0062] Specifically, the user preference is the user's preference for a certain type of video resource or scene in the online video media library, or a video resource shot by a certain actor or director in the video resource or scene, which is understood through the user's previous or current viewing habits when the user uses the online video media library. The user preference may be one, multiple, or none. It is understandable that when there is no user preference in the online video media library, it is likely that the user is watching a video in the online video media library for the first time. When there are multiple user preferences, the user preferences can be displayed in the online video media library, such as setting a user preference column in the online video media library. When using the online video media library, the user can modify it in the user preference column by himself, wherein the modification includes adding and deleting, etc., so that the user can accurately locate his own preferences.

[0063] In an implementation of this embodiment, the controlling the media resource library server to determine the recommended resource corresponding to the target video based on the user preference and the recommendation dimension, and recommending the determined recommended resource specifically includes:

[0064] S3021, controlling the media asset library server to determine a first candidate recommendation resource based on the user preference and the recommendation dimension;

[0065] S3022, controlling the media resource library server to obtain a second candidate recommendation resource formed based on the hot segment;

[0066] S3023: Use the first candidate recommended resource and the second candidate recommended resource as recommended resources corresponding to the target video, and recommend the determined recommended resource.

[0067] Specifically, the first candidate recommendation resource is a recommendation resource for further creation of the hot segment of the target video based on the user preference and the recommendation dimension. In an implementation of this embodiment, the first candidate recommendation resource can be a film review of the hot segment, or a quick movie showing a summary of the content of the hot segment, or a clip of the hot segment, or a video with non-hot segments added to the hot segment. The first candidate recommendation resource is displayed in the display area of ​​the recommended video. In an implementation of this embodiment, for example, the hot segment b in video a obtains recommended content including actor A and royal buildings through the video understanding model, then the recommendation dimension The dimension may be the royal architectural scene of the ancient character played by actor A, and the user prefers to watch film reviews, then the first candidate recommendation resource displays the film reviews of the popular segment b in video a; the second candidate recommendation resource is other videos that are non-target videos based on the user preference and the recommendation dimension. In one implementation of this embodiment, for example, the popular segment b in video a obtains recommended content including actor A and royal architecture through the video understanding model, then the recommendation dimension may be the royal architectural scene of the ancient character played by actor A, and the user prefers to watch film C, then the user is given priority to watch film C or similar films C that have royal architectural scenes with ancient characters played by actor A. Both the first candidate recommended video and the second candidate recommended video can be displayed in the display area of ​​the recommended video. It can be understood that the order in which the first candidate recommended video and the second candidate recommended video are displayed in the recommended video display area can be arranged according to user preferences. For example, the user can set priorities for multiple user preferences; or the user can choose to set the order by himself. For example, the first candidate recommended resource is arranged before the second candidate recommended resource. The first candidate recommended resource and the second candidate recommended resource in the recommended video display area can be collected, played, sequentially moved or deleted.

[0068] In an implementation of this embodiment, after determining the recommended resource corresponding to the target video according to the recommendation dimension and recommending the determined recommended resource, the method further includes:

[0069] S401, monitoring the recommended duration of the recommended resource;

[0070] S402. When the recommended duration reaches a preset duration, re-execute the step of obtaining the playback progress of the target video and determining the hot clips corresponding to the target video according to the playback progress to update the recommended resources.

[0071] Specifically, the recommended duration is equal to the preset duration, which can be a reference playback progress duration of the popular segment, or the playback duration between the playback progress of the front end point of the previous popular segment and the playback progress of the front end point of the next popular segment. When the preset duration is the reference playback progress duration of the popular segment, as soon as the playback progress reaches the front end of the current popular segment of the target video, the recommended resources will be displayed in the recommended video playback area. When the playback progress reaches the last end of the current popular segment, the recommended video playback area will stop displaying the recommended resources of the current popular segment, re-execute the acquisition of the playback progress of the target video, and then determine the popular segment corresponding to the target video according to the playback progress, and then re-execute the above operation to update the recommended resources of the next popular segment; when the preset duration is the playback duration between the playback progress of the front end of the previous popular segment and the playback progress of the front end of the next popular segment, as soon as the playback progress reaches the front end of the current popular segment of the target video, the recommended resources will be displayed in the recommended video playback area. Only when the playback progress reaches the front end of the next popular segment, the recommended video playback area will stop displaying the recommended resources of the current popular segment, re-execute the acquisition of the playback progress of the target video, and update the display of the recommended resources of the next popular segment.

[0072] The present invention also provides a multimedia resource recommendation device, such as Figure 3 As shown, it includes:

[0073] The acquisition module 100 is used to acquire the playback progress of the target video and determine the hot clips corresponding to the target video according to the playback progress;

[0074] A determination module 200, configured to determine a recommendation dimension corresponding to the target video based on the popular clips and the trained video understanding model;

[0075] The recommendation module 300 is used to determine the recommended resources corresponding to the target video according to the recommendation dimension, and recommend the determined recommended resources.

[0076] Based on the above-mentioned method for recommending multimedia resources, this embodiment provides a computer-readable storage medium, which stores one or more programs. The one or more programs can be executed by one or more processors to implement the steps in the method for recommending multimedia resources as described in the above-mentioned embodiment.

[0077] Based on the above-mentioned multimedia resource recommendation method, the present application also provides a terminal device, such as Figure 4As shown, it includes at least one processor (processor) 20; display screen 21; and memory (memory) 22, and may also include a communications interface (Communications Interface) 23 and a bus 24. Among them, the processor 20, the display screen 21, the memory 22 and the communication interface 23 can communicate with each other through the bus 24. The display screen 21 is configured to display a preset user guide interface in the initial setting mode. The communication interface 23 can transmit information. The processor 20 can call the logic instructions in the memory 22 to execute the method in the above embodiment.

[0078] In addition, the logic instructions in the memory 22 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.

[0079] The memory 22 is a computer-readable storage medium that can be configured to store software programs, computer executable programs, such as program instructions or modules corresponding to the methods in the embodiments of the present disclosure. The processor 20 executes functional applications and data processing by running the software programs, instructions or modules stored in the memory 22, that is, implementing the methods in the above embodiments.

[0080] The memory 22 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory 22 may include a high-speed random access memory and may also include a non-volatile memory. For example, a variety of media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, may also be a transient storage medium.

[0081] In addition, the specific process of loading and executing the plurality of instructions in the storage medium and the processor in the mobile terminal has been described in detail in the method above, and will not be described one by one here.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for recommending multimedia resources, characterized in that: The method comprises: Obtain the playback progress of the target video, and determine the hot clips corresponding to the target video according to the playback progress; Determine the recommended dimension corresponding to the target video based on the popular clips and the trained video understanding model; Determine the recommended resource corresponding to the target video according to the recommendation dimension, and recommend the determined recommended resource; The video understanding model is configured with a number of default dimensions; and the determining of the recommended dimensions corresponding to the target video based on the popular clips and the trained video understanding model includes: Inputting the hot clips into the video understanding model, and outputting the dimensional content of the hot clips in each default dimension through the video understanding model; Generating a recommended dimension corresponding to the target video based on the dimensional content of the popular clips on each default dimension, wherein the recommended dimension is a query condition formed based on the dimensional content of the popular clips on each default dimension; The hot segment is included in the target video, and the reference playback progress of the hot segment is located before the playback progress; The playback progress is the current playback position of the target video; The hot clip is a clip with a number of viewers greater than a preset number threshold, or a clip selected by the user, or a clip with a preset building or scene, or a clip with the largest number of viewers in the video in which the hot clip is located; Determining the recommended resource corresponding to the target video according to the recommendation dimension and recommending the determined recommended resource specifically includes: Sending the recommended dimensions to the media library server, and controlling the media library server to obtain the user preference of the user corresponding to the target video; Controlling the media resource library server to determine the recommended resource corresponding to the target video based on the user preference and the recommendation dimension, and recommending the determined recommended resource; The controlling the media resource library server to determine the recommended resource corresponding to the target video based on the user preference and the recommendation dimension, and recommending the determined recommended resource specifically includes: Controlling the media asset library server to determine a first candidate recommendation resource based on the user preference and the recommendation dimension; Controlling the media resource library server to obtain a second candidate recommendation resource formed based on the hot segment; Using the first candidate recommended resource and the second candidate recommended resource as recommended resources corresponding to the target video, and recommending the determined recommended resource; The first candidate recommendation resource is a recommendation resource for further creating a hot segment of the target video based on the user preference and the recommendation dimension; The second candidate recommendation resource is other videos that are non-target videos and are derived based on the user preference and the recommendation dimension.

2. The method for recommending multimedia resources according to claim 1, characterized in that: The default dimensions include at least information about the film source, actor roles, architecture, shooting techniques, and film types.

3. The method for recommending multimedia resources according to claim 1, characterized in that: After determining the recommended resource corresponding to the target video according to the recommendation dimension and recommending the determined recommended resource, the method further includes: Monitoring the recommended duration of the recommended resource; When the recommended duration reaches a preset duration, the step of obtaining the playback progress of the target video and determining the hot clips corresponding to the target video according to the playback progress is performed again to update the recommended resources.

4. A multimedia resource recommendation device, characterized in that: The device comprises: An acquisition module is used to acquire the playback progress of the target video and determine the hot clips corresponding to the target video according to the playback progress; A determination module, used to determine the recommendation dimension corresponding to the target video based on the popular clips and the trained video understanding model; A recommendation module, used to determine the recommended resources corresponding to the target video according to the recommendation dimension, and recommend the determined recommended resources; The video understanding model is configured with a number of default dimensions; and the determining of the recommended dimensions corresponding to the target video based on the popular clips and the trained video understanding model includes: Inputting the hot clips into the video understanding model, and outputting the dimensional content of the hot clips in each default dimension through the video understanding model; Generating a recommended dimension corresponding to the target video based on the dimensional content of the popular clips on each default dimension, wherein the recommended dimension is a query condition formed based on the dimensional content of the popular clips on each default dimension; The hot segment is included in the target video, and the reference playback progress of the hot segment is located before the playback progress; The playback progress is the current playback position of the target video; The hot clip is a clip with a number of viewers greater than a preset number threshold, or a clip selected by the user, or a clip with a preset building or scene, or a clip with the largest number of viewers in the video in which the hot clip is located; The recommendation module specifically includes: Sending the recommended dimensions to the media library server, and controlling the media library server to obtain the user preference of the user corresponding to the target video; Controlling the media resource library server to determine the recommended resource corresponding to the target video based on the user preference and the recommendation dimension, and recommending the determined recommended resource; The controlling the media resource library server to determine the recommended resource corresponding to the target video based on the user preference and the recommendation dimension, and recommending the determined recommended resource specifically includes: Controlling the media asset library server to determine a first candidate recommendation resource based on the user preference and the recommendation dimension; Controlling the media resource library server to obtain a second candidate recommendation resource formed based on the hot segment; Using the first candidate recommended resource and the second candidate recommended resource as recommended resources corresponding to the target video, and recommending the determined recommended resource; The first candidate recommendation resource is a recommendation resource for further creating a hot segment of the target video based on the user preference and the recommendation dimension; The second candidate recommendation resource is other videos that are non-target videos and are derived based on the user preference and the recommendation dimension.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the multimedia resource recommendation method as described in any one of claims 1-3.

6. A terminal device, characterized in that: include: Processor, memory and communication bus; The memory stores a computer-readable program executable by the processor; The communication bus realizes the connection and communication between the processor and the memory; When the processor executes the computer-readable program, the processor implements the steps in the method for recommending multimedia resources as described in any one of claims 1 to 3.

Images