CN113422988B - Live broadcast cover generation method and device, electronic equipment and storage medium - Google Patents

Abstract

The application discloses a live broadcast cover generation method and device, electronic equipment and a storage medium, and relates to the technical field of internet live broadcast. The method comprises the following steps: after obtaining the interactive data distribution of a target live broadcast room in a preset time period before the current live broadcast time, firstly grading the interactive data distribution in the preset time period in a segmented manner according to a preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, then determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period, finally determining a target video frame from a plurality of video frames of the target live broadcast room in the target sub-time period, and updating a live broadcast cover of the target live broadcast room by using the target video frame. The method and the device can improve the quality of the live cover and enable the live cover to be closer to real-time live content.

Description

Live broadcast cover generation method and device, electronic equipment and storage medium

Technical Field

The present application relates to the field of internet live broadcast technologies, and in particular, to a live broadcast cover generation method and apparatus, an electronic device, and a storage medium.

Background

With the development of internet technology, the live broadcast industry is rapidly developing, and people begin to choose to watch live webcast as an entertainment mode. For a user watching a live webcast, it is the most straightforward way to know the anchor by browsing the live cover page of the anchor's live room. However, the existing live cover generation method causes the generated live cover to have low quality or not to be consistent with live content.

Disclosure of Invention

In view of this, embodiments of the present application provide a live cover generation method and apparatus, an electronic device, and a storage medium to solve the above problem.

In a first aspect, an embodiment of the present application provides a live cover generation method, where the method includes: acquiring interactive data distribution of a target live broadcast room in a preset time period before the current live broadcast time, wherein the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension; grading the interactive data distribution in a preset time period in a segmented manner according to a preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interactive data indicates the interactive degree of audiences in a target live broadcast room; determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period; and determining a target video frame from a plurality of video frames in the target sub-time period in the target live broadcasting room, and updating a live broadcasting cover of the target live broadcasting room by using the target video frame, wherein the live broadcasting cover is used for displaying the cover in the target live broadcasting room.

In a second aspect, an embodiment of the present application provides a live cover generating device, where the device includes: the system comprises an acquisition module, a segmentation grading module, a target sub-time period determination module and a live cover updating module. The system comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for acquiring interactive data distribution of a target live broadcast room in a preset time period before the current live broadcast time, and the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension; the segmentation grading module is used for performing segmentation grading on the interaction data distribution in the preset time period according to the preset time length to obtain a comprehensive score corresponding to the interaction data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interaction data indicates the interaction degree of audiences in the target live broadcast room; the target sub-time period determining module is used for determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period; and the live broadcast cover updating module is used for determining a target video frame from a plurality of video frames in the target sub-time period in the target live broadcast room, updating the live broadcast cover of the target live broadcast room by using the target video frame, and displaying the live broadcast cover in the target live broadcast room.

In a third aspect, an embodiment of the present application provides an electronic device, including a processor and a memory; one or more programs are stored in the memory and configured to be executed by the processor to implement the methods described above.

In a fourth aspect, the present application provides a computer-readable storage medium having program code stored therein, where the program code executes the method described above when executed by a processor.

In a fifth aspect, embodiments of the present application provide a computer program product or a computer program comprising computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the above-described method.

According to the live broadcast cover generation method, the live broadcast cover generation device, the electronic equipment and the storage medium, after interactive data distribution of a target live broadcast room in a preset time period before a current live broadcast time is obtained, the interactive data distribution in the preset time period is graded in a segmented mode according to a preset time length to obtain comprehensive scores corresponding to the interactive data of all sub time periods included in the preset time period, then the target sub time period is determined from all the sub time periods based on the comprehensive scores corresponding to the interactive data of all the sub time periods, finally a target video frame is determined from a plurality of video frames of the target live broadcast room in the target sub time period, and the live broadcast cover of the target live broadcast room is updated through the target video frame. Therefore, through the mode, on the one hand, because the comprehensive score obtained based on audience interaction data calculation can reflect the interaction degree of audiences in the live broadcast room, the interaction degree is high, the possibility of occurrence of wonderful covers is higher, the live broadcast covers determined from the situation can attract the interests of audiences, and therefore the quality of the finally determined live broadcast covers is higher.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 shows a schematic view of a display interface of a live client according to an embodiment of the present application;

fig. 2 is a schematic diagram of an application environment proposed in the embodiment of the present application;

fig. 3 is a flowchart illustrating a live cover generation method according to an embodiment of the present application;

fig. 4 is a flowchart illustrating another live cover generation method according to an embodiment of the present application;

fig. 5 is a flowchart of an embodiment of S230 in a live cover generation method according to the embodiment shown in fig. 4;

fig. 6 is a flowchart illustrating another live cover generation method according to an embodiment of the present application;

fig. 7 is a flowchart illustrating another live cover generation method according to an embodiment of the present application;

fig. 8 is a flowchart illustrating another live cover generation method according to an embodiment of the present application;

fig. 9 is a flowchart illustrating another live cover generation method according to an embodiment of the present application;

fig. 10 is a block diagram of a live cover generation apparatus according to an embodiment of the present application;

fig. 11 is a block diagram illustrating an electronic device for executing a live cover generation method according to an embodiment of the present application;

fig. 12 illustrates a storage unit for storing or carrying program codes for implementing a live cover generation method according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

With the development of computer networks, live broadcasts are becoming more and more popular. As shown in fig. 1, a live room list is typically provided to the viewer in the interface of the live client, and in order to facilitate the viewer to quickly understand the summary of the live content in the live room, live covers are typically added to each live room in the live room list.

However, the existing live cover has certain limitation in generation.

For example, in some related technologies, the random screenshot may be generated by an operation configuration, a main broadcast upload or a live random screenshot, where the operation configuration and the main broadcast upload may cause a situation inconsistent with live content, and the live random screenshot may cause poor quality of live cover. Therefore, the audience user can find the content which the audience wants to watch only by frequently switching the live broadcast rooms, the user time is wasted to a certain extent, the efficiency of selecting live broadcast videos by the user is reduced, and the effect of the live broadcast cover is poor.

Therefore, in view of the above problems, the inventor has made long-term research and proposed a live cover generation method, an apparatus, an electronic device and a storage medium provided by the present application, in which method, a time period in which a highlight cover may appear is obtained by processing audience interaction data distribution before a current live broadcast time, and then a live cover is selected from the time period in which the highlight cover may appear, so as to determine the live cover, on one hand, since audience interaction data can reflect the degree of audience interaction in a live broadcast, the degree of interaction is high, the probability of occurrence of the highlight cover is higher, and the live cover determined therefrom can attract the interests of the audience better, so that the cover quality is higher, on the other hand, since the obtained live broadcast interaction data distribution in a preset time period before the current live broadcast time avoids obtaining interaction data distribution in all time periods since the broadcast is started, the finally determined content of the live cover can be closer to real-time live content.

For the convenience of describing the scheme of the present application in detail, an application environment in the embodiment of the present application is described below with reference to the accompanying drawings.

Referring to fig. 2, an application environment schematic diagram of a live cover generating method according to an embodiment of the present application is shown in fig. 2, where the application environment can be understood as a live system 10 according to an embodiment of the present application, and the live system 10 includes: server 11,anchor client 12, andviewer client 13.

The server 11 may be a server (network access server), a server cluster (cloud server) composed of a plurality of servers, or a cloud computing center (database server). Theanchor client 12 and theviewer client 13 may be any device with communication and storage capabilities, including but not limited to a PC (Personal Computer), a PDA (tablet), a smart tv, a smart phone, a smart wearable device, or other smart communication device with network connection capabilities. The server 11 may communicate with theanchor client 12 and theviewer client 13 by connecting wirelessly or by wire.

In addition, it should be noted that theanchor client 12 and theaudience client 13 may be the same client, that is, theanchor client 12 and theaudience client 13 are relative concepts, so that, in some specific scenes, theanchor client 12 may become theaudience client 13, and theaudience client 13 may also become theanchor client 12.

Optionally, theanchor client 12 and theviewer client 13 in the live broadcast system 10 may be application clients (such as a live application in a mobile APP), or may be web page clients (such as a live webcast platform), which is not limited herein.

Embodiments of the present application will be described in detail below with reference to the accompanying drawings.

Referring to fig. 3, fig. 3 is a flowchart illustrating a live broadcast cover generation method according to an embodiment of the present application, where the embodiment of the present application takes the server in the live broadcast system shown in fig. 2 as an example to explain the method, and the method may include the following steps:

s110, according to a preset time interval, periodically obtaining interactive data distribution of a target live broadcast room in a preset time period before the current live broadcast time, wherein the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension.

And the current live broadcast moment is the moment when a preset time interval arrives. It can be understood that, in the live broadcast process of the anchor user, the live broadcast content of the anchor user is changed at any time along with the live broadcast, therefore, in order to enable the content of the live broadcast cover to be closer to the real-time live broadcast content, the server can periodically acquire the interactive data distribution of the target live broadcast room in the preset time period before the current live broadcast time according to a certain time interval, and accordingly the live broadcast cover of the target live broadcast room in the live broadcast process is periodically updated.

The target live broadcast room is a live broadcast room which needs to generate a live broadcast cover subsequently. It can be understood that, in practice, in a certain live system, for example, in a certain live platform, an anchor user may register a live account with a server through an anchor client, so as to obtain a live room corresponding to the live account, a large number of anchor users may exist in the live system at every moment to use the live room registered by the anchor client to perform live broadcast, at this time, a large number of anchor clients exist to access the server, and therefore, in some embodiments, the live rooms corresponding to the live clients that access the server to perform live broadcast can all become target live rooms.

In addition, in other embodiments, the target live broadcast room may also be a live broadcast room corresponding to a client that has access to the server for live broadcast and meets certain conditions. For example, in some cases, a anchor user may freely set whether a live broadcast room of the anchor user needs to enable the method of the embodiment of the present application, so as to generate a live broadcast room cover. Specifically, the anchor user may send an instruction carrying the user identifier or the live broadcast room identifier and representing that the method of the embodiment of the present application is enabled to the server, and at this time, after receiving the instruction, the server may record the user identifier or the live broadcast room identifier.

Optionally, the instruction for characterizing that the method of the embodiment of the present application is enabled may be generated and sent by clicking a preset control in a system setting interface of the anchor client, for example, clicking a control such as "open cover generation" or "automatically generate cover".

The preset time period can be set according to needs. For example, any time period such as 1 minute, 2 minutes, 10 minutes, and half an hour may be set. The preset time period is not specifically limited in the embodiment of the present application. It should be understood that the smaller the preset time period setting, the closer to the current live time, and thus the greater the likelihood that the acquired content is similar to the content currently live by the anchor user.

In addition, it is understood that in a live broadcast system, a server usually records various types of data in a live broadcast process so as to be used for data statistics and data analysis. For example, the number of people watching, the number of viewer likes, the number of viewer comments, etc. in real time during the live broadcast of each live broadcast room may be recorded. Therefore, when the server executes the live broadcast cover generation method of the embodiment of the application, the distribution condition of the interactive data generated by the audience in the target live broadcast room in the time dimension in the preset time period before the current live broadcast time can be obtained.

When the server executes the live cover page generation method of the embodiment of the application, the used interactive data may include one or a combination of multiple types of the number of watching people, the number of audience praise, the number of audience comments, and the like.

Optionally, the interactive data may include only the number of viewers, only the number of viewer likes, only the number of viewer comments, only the number of viewers and only the number of viewer likes, only the number of viewer comments, only the number of viewer likes and only the number of viewer comments, and optionally, the interactive data may also include both the number of viewers, the number of viewer likes, and the number of viewer comments. Accordingly, the interactive data distribution acquired by the server corresponds to the content included in the interactive data. For example, when the interactive data simultaneously includes the number of viewers, the number of the viewers agreeing with the viewer and the number of the viewers comments, the server acquires the number distribution of the viewers, the number distribution of the viewers agreeing with the viewer and the number distribution of the viewers comments in a preset time period before the current live broadcast time of the target live broadcast room, and simultaneously determines the live broadcast cover of the target live broadcast room based on the number distribution of the viewers, the number distribution of the viewers agreeing with the viewer and the number distribution of the viewers comments.

The above description only describes some types of interactive data by way of example, and in some cases, the interactive data may further include data such as the amount of gifts given by the audience, the amount of attention paid by the audience, and the like.

And S120, grading the interactive data distribution in the preset time period in a segmented manner according to the preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interactive data indicates the interaction degree of the audience in the target live broadcast.

After the interactive data distribution of the target live broadcast room in the preset time period before the current live broadcast time is obtained, the preset time period can be further divided by taking the preset time length as a basic time unit, so that a plurality of sub-time periods included in the preset time period are obtained, the interactive data corresponding to the sub-time periods are further obtained, then the interactive data corresponding to the sub-time periods are scored, the comprehensive scores corresponding to the interactive data of the sub-time periods are obtained, and the interactive degree of audiences in the target live broadcast room in the sub-time periods is obtained. Therefore, the segmentation scoring may include two main processes of segmenting the interactive data distribution in the preset time period according to the preset time length to obtain the interactive data of each sub-time period included in the preset time period, and scoring the interactive data of each sub-time period respectively to obtain the comprehensive score corresponding to the interactive data of each sub-time period.

It will be appreciated that the higher the composite score corresponding to the interaction data for a sub-period, the higher the degree of viewer interaction in the target live broadcast room, and the greater the likelihood of highlight occurrence in that sub-period.

It should be noted that the time length of the preset time period and the preset time length may be set according to experience, or may be set freely by the anchor user according to the preference or habit of the anchor user.

It can be understood that, in order to segment the interaction data distribution in the preset time period accurately according to the preset time length, the preset time length may be smaller than the time length corresponding to the preset time period. When the set preset time length is less than the time length corresponding to the preset time period, the interactive data distribution in the preset time period is segmented according to the preset time length, and the interactive data of each sub-time period included in the preset time period can be obtained in various modes.

In some possible scenes, when the time length of the preset time period and the preset time length are values set by research and development personnel according to experience, the time length of the preset time period is most likely to be exactly integral multiple of the preset time length, and at this time, the interactive data distribution in the preset time period can be segmented in time according to the preset time length directly. For example, when the time length of the preset time period is 1 minute and the preset time length is 10 seconds, the time length of the preset time period is exactly an integral multiple of the preset time length, and at this time, the preset time period may be segmented once every 10 seconds according to the preset time length to obtain 6 sub-time periods in total, and obtain the interaction data in the 6 sub-time periods.

In other possible scenarios, when the time length of the preset time period and the preset time length are values freely set by the anchor user according to the preference or habit of the anchor user, the time length of the preset time period is probably not an integral multiple of the preset time length, and at this time, the interactive data distribution in the preset time period is segmented in time according to the preset time length, so that the time length of one sub-time period may be less than the preset time length. In this case, it may be considered that the interactive data of the sub-period having the time length shorter than the preset time length is directly ignored.

The time sequence segmentation of the interactive data distribution in the preset time period can also include a plurality of segmentation modes according to the preset time length.

As an implementation manner, the segments may be sequentially segmented according to a preset time length and a time sequence, so that the sub-time period with the time length less than the preset time length is a time period closest to the current live broadcast time.

For example, when the time length of the preset time period is 1 minute, and it is assumed that the start time of the preset time period is 0 th second, the end time (current live broadcast time) of the preset time period is 60 th second, and the preset time length is 14 seconds, the time length of the preset time period is not an integral multiple of the preset time length, at this time, the preset time period may be sequentially segmented according to the time sequence, that is, the preset time period is segmented every 14 seconds from the 0 th second time, so as to obtain 4 sub-time periods, that is, 0 second to 14 seconds, 14 second to 28 seconds, 28 second to 42 seconds, and 42 second to 56 seconds, and obtain the interactive data in the 4 sub-time periods, thereby ignoring the interactive data in the 56 seconds to 60 seconds nearest to the current live broadcast time.

As another embodiment, the current live broadcast time may be sequentially segmented to a time period before the current live broadcast time, so that a sub-time period with a time length less than a preset time length is a time period farthest from the current live broadcast time, and the interactive data corresponding to the sub-time period is discarded. In this time, because the interaction data of the sub-time period farthest from the current live time is ignored, compared with the live contents of other sub-time periods, the live contents of the sub-time period have a larger possibility that the current live contents of the live broadcasting room cannot be truly reflected, and therefore, the interaction data of the sub-time period farthest from the current live time is considered, and the contents of the live broadcasting cover can be further close to the real-time live contents.

For example, when the time length of the preset time period is 1 minute, and it is assumed that the start time of the preset time period is 0 th second, the end time of the preset time period (current live broadcast time) is 60 th second, and the preset time length is 14 seconds, the time length of the preset time period is not an integral multiple of the preset time length, at this time, the preset time period may be sequentially segmented according to the preset time length from the current live broadcast time to the time period before the current live broadcast time, that is, the preset time period is segmented every 14 seconds from the 60 th second time to the front, so as to obtain 4 sub-time periods, that is, 60 seconds to 46 seconds to 32 seconds, 32 seconds to 18 seconds, and 18 seconds to 4 seconds, and obtain the interactive data in the 4 sub-time periods, thereby ignoring the interactive data in 4 seconds to 0 second farthest from the current live broadcast time.

In addition, in the embodiment of the present application, scoring is performed on the interaction data of each sub-time period, so as to obtain a comprehensive score corresponding to the interaction data of each sub-time period, which may also be achieved in various manners.

In one embodiment, it is considered that the interactive data itself is a numerical data, and therefore, the interactive data may not be processed, and the value of the interactive data in each sub-period may be directly used as the score value of the total score in each sub-period. Under the condition, the method can be divided into two specific conditions, in the first condition, the interaction influence degree of various types of interaction data is not considered, namely, the weight of various types of interaction data is set to be 1, and therefore the sum of various types of interaction data of each sub-time period is directly used as the final comprehensive score of each sub-time period. In the second case, considering that different interactive behaviors have different degrees of influence, for example, compared with watching a deeper interactive behavior, the interactive behavior of the deeper interactive behavior has a larger degree of influence, so that a relatively larger weight can be set for the number of prawns or comments, and a relatively smaller weight can be set for the number of people watching, and the weighted sum of various types of interactive data of each sub-period is used as the final comprehensive score of each sub-period. As can be seen from the foregoing, the type of the interactive data may include one or more types, and when the type of the interactive data includes one type, the second type case may be understood as the weight being 1.

As another embodiment, in order to improve the accuracy of calculating the composite score, that is, to more accurately obtain the interaction degree of the audience in the target live broadcast room, it may be considered to perform normalization processing on the interaction data of each sub-time period to obtain a normalized score of the interaction data of each sub-time period, and further calculate the composite score of the interaction data of each sub-time period based on the normalized score of the interaction data of each sub-time period. In this case, the two cases can be also divided, in the first case, the weight of each type of interaction data in each sub-period is set to 1 regardless of the interaction influence degree of each type of interaction data, so that the normalized scores of each type of interaction data in each sub-period are directly added to serve as the final comprehensive score of each sub-period. In the second case, the weighted sum of various types of interaction data of each sub-period is used as the final comprehensive score of each sub-period in consideration of different interaction behavior influence degrees. Likewise, when the type included in the interactive data is one type, the second type of case may be understood as the weight being 1.

It should be noted that, in order to further improve the quality of the generated live broadcast cover and attract viewers, the set weight may be continuously compared and optimized according to the actual use process or the previous testing process, for example, in the comparison and optimization process, the interactive data distribution may be divided into at least two parts, then different weights are set respectively, then at least two target video frames corresponding to the target live broadcast room are finally obtained based on different interactive data distributions and corresponding weights, then the at least two target video frames are respectively used as the live broadcast cover of the target live broadcast room and randomly sent to different viewer clients, when sent to different viewer clients, the same release proportion of each live broadcast cover is ensured as much as possible, after a period of time, the number of live broadcast covers entering the target live broadcast room by clicking each cover is counted, and then the weight is updated to the weight used by the live broadcast cover with the largest number when being generated, and then continuously carrying out the contrast optimization process.

And S130, determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interaction data of each sub-time period.

The target sub-period is one or more sub-periods determined from the respective sub-periods that satisfy the composite score condition. The composite score condition may be various, for example, a score threshold condition or a score ordering condition, etc. Therefore, various modes can be provided for determining the target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period.

As an embodiment, a score threshold may be preset, in which case, each sub-period having a composite score greater than or equal to the score threshold corresponding to the interaction data may be determined as the target sub-period.

As another embodiment, after the comprehensive scores corresponding to the interactive data of each sub-period are obtained, the sub-periods may be sequentially sorted according to the high-low relationship of the comprehensive scores, so as to select a preset number of sub-periods corresponding to the comprehensive scores in the front of the comprehensive score sorting, and determine the sub-periods as the target sub-periods.

It can be understood that, in the foregoing embodiment, the comprehensive score only refers to the degree of interaction of the audience in the target live broadcast room, and takes into account real-time variability of the live broadcast content, that is, the probability that the live broadcast content at a time farther away from the current live broadcast time is different from the live broadcast content at the current live broadcast time is higher, so in some embodiments, in order to obtain a comprehensive score that takes a more comprehensive consideration, according to a preset time length, the interactive data distribution in a preset time period is graded in a segmented manner, so as to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, which may specifically include: the method comprises the steps of grading interactive data distribution in a preset time period in a segmented mode according to a preset time length to obtain interactive scores corresponding to interactive data of all sub-time periods included in the preset time period, and setting corresponding similarity coefficients for all the sub-time periods in advance based on the distance between each sub-time period and the current live broadcast time, wherein the similarity coefficient corresponding to the sub-time period closer to the current live broadcast time is larger, and the comprehensive score corresponding to each simple interactive data is obtained based on the interactive scores corresponding to the interactive data of all the sub-time periods and the corresponding similarity coefficients.

In this embodiment, a similarity coefficient representing the similarity between the live cover and the current live content can be further added to each sub-time period according to the distance between each sub-time period and the current live time, so that the quality of the generated live cover can be further improved based on the product of the interaction score of the interaction data of the same sub-time period and the similarity coefficient, that is, a more comprehensive score can be obtained.

And S140, determining a target video frame from a plurality of video frames in the target sub-time period in the target live broadcasting room, and updating a live broadcasting cover of the target live broadcasting room by using the target video frame, wherein the live broadcasting cover is used for displaying the cover in the target live broadcasting room.

It will be appreciated that only one live cover is usually required to be provided, and the obtained target sub-period may include a plurality of video frames, so that after the target video frame is obtained, the target video frame may be further determined from the plurality of video frames in the target sub-period in the target live room. It is understood that when the number of the target sub-periods is 1, one target video frame is determined from all the video frames of the one target sub-period, and when the number of the target sub-periods is multiple, for example, 2, 3, or 5, the one target video frame is determined from all the video frames included in the multiple target sub-periods.

There are various ways to determine the target video frame from the plurality of video frames in the target sub-period in the target live broadcast room.

Alternatively, the determination may be made by manual selection. For example, a plurality of video frames are sent to a main broadcast client corresponding to a target live broadcast room, so that a main broadcast user can select the video frames by himself, after the target video frames are selected, a target video frame instruction is sent to a server, the server can determine the target video frames according to the target video frame instruction, and accordingly the target video frames are used for updating the live broadcast covers of the target live broadcast room.

Alternatively, the determination may be made by automatic selection by the server. There may be many ways to determine the determination by the automatic selection of the server.

As an embodiment, the server may determine the target video frame from a plurality of video frames within the target sub-period by means of an AI (Artificial Intelligence) model. For example, after obtaining the plurality of video frames in the target sub-period, the server may input the plurality of video frames in the target sub-period into the video frame quality scoring model, respectively, obtain the quality score corresponding to each video frame output by the video frame scoring model, and then determine the video frame with the largest quality score as the target video frame.

The quality scoring model is obtained by training an initial model through a plurality of sample video frames carrying score labels.

Alternatively, the server may determine the target video frame from a plurality of video frames within the target sub-period by a preset selection policy. Alternatively, the preset selection policy may be a random selection policy, that is, the server randomly selects one video frame from a plurality of video frames in the target sub-period as the target video frame at a time. Optionally, the preset selection policy may also be a policy for selecting the target video frame according to a time point in the sub-period, and for example, assuming that the preset time length is 10 seconds, after obtaining a plurality of video frames in the target sub-period, a video frame at an intermediate 5 th second time within a 10 second time range corresponding to the sub-period may be set to be determined as the target video frame. If the number of the sub-time periods is multiple, the video frames at the middle 5 th second moment within the 10 second time range corresponding to each sub-time period can be acquired first, and then one video frame is randomly selected from the acquired video frames to be determined as the target video frame.

As still another embodiment, the server may determine the target video frame by means of an online click-through rate test.

In this embodiment of the application, after determining the target video frame, the server may use the target video frame as a live broadcast cover of the target live broadcast room, so that when the viewer client enters the interface shown in fig. 1 to load a live broadcast room cover, if a target live broadcast room exists in the live broadcast list, the live broadcast cover corresponding to the target live broadcast room sent to the viewer client by the server may be obtained, and thus the viewer client may load and display the live broadcast cover corresponding to the target live broadcast room in the interface shown in fig. 1.

After acquiring the interactive data distribution of a target live broadcast room in a preset time period before the current live broadcast time, grading the interactive data distribution in the preset time period in a segmented manner according to the preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, then determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period, finally determining a target video frame from a plurality of video frames of the target live broadcast room in the target sub-time period, and updating the live broadcast cover of the target live broadcast room by using the target video frame. Therefore, by the mode, on one hand, the comprehensive score calculated based on the audience interaction data can reflect the interaction degree of the audience in the live broadcast room, the interaction degree is high, the possibility of the occurrence of the wonderful cover is higher, the live broadcast cover determined from the comprehensive score can attract the interest of the audience better, so that the finally determined live broadcast cover has higher quality, on the other hand, because the interactive data distribution in the preset time period before the current live broadcast time is obtained, the interactive data distribution of all live broadcast time periods from the broadcast is avoided, the content of the finally determined live broadcast cover can be closer to the real-time live broadcast content, in addition, the periodical operation is realized by setting the preset time interval, along with the live broadcast, the live cover can be continuously acquired from the wonderful live content, the quality of the live cover in the live process is further improved, and the content of the live cover can be closer to real-time live content. Therefore, the audience user can find the content which the audience user wants to watch without frequently switching the live broadcast room, the time of the user is saved, the efficiency of selecting live broadcast videos of the user is improved, and the effect of live broadcast covers is improved.

In addition, as can be seen from the foregoing, the interactive data may include at least two types of interactive data, in this case, please refer to fig. 4, where fig. 4 is a flowchart illustrating a live cover generation method according to another embodiment of the present application, and the embodiment of the present application is described by taking an example of applying the method to a server in the live system shown in fig. 2, where the method may include the following steps:

s210, according to a preset time interval, periodically acquiring interactive data distribution of the target live broadcast room in a preset time period before the current live broadcast time, wherein the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension.

And S220, carrying out sectional statistics on the interactive data distribution corresponding to each type of interactive data in the preset time period according to the preset time length to obtain the data volume of each type of interactive data corresponding to each sub-time period included in the preset time period.

In the embodiment of the application, in order to calculate the comprehensive score according to the normalized score and improve the calculation accuracy of the comprehensive score under the condition that at least two types of interactive data exist, the interactive data distribution corresponding to each type of interactive data in the preset time period can be subjected to sectional statistics according to the preset time length, so that the data amount of each type of interactive data corresponding to each sub-time period included in the preset time period is obtained.

Exemplarily, assuming that the interactive data simultaneously includes three categories, namely, the number of audience comments, the number of audience praise and the number of watching people, in this case, the distribution of the number of audience comments, the distribution of the number of audience praise and the distribution of the number of watching people in the preset time period may be separated, and the distribution of the number of audience comments, the distribution of the number of audience praise and the distribution of the number of watching people in the preset time period are respectively subjected to segment statistics to obtain the number of audience comments, the number of audience praise and the number of watching people corresponding to each sub-time period included in the preset time period.

And S230, respectively carrying out normalization processing on the data quantity of each type of interaction data corresponding to each sub-time period to obtain the normalization score of each type of interaction data corresponding to each sub-time period.

After the data amount of each type of interaction data corresponding to each sub-time period included in the preset time period is obtained, normalization processing can be performed on each type of interaction data corresponding to each sub-time period respectively, so that a normalization score of each type of interaction data corresponding to each sub-time period is obtained.

As an embodiment, as shown in fig. 5, the normalizing the data amount of each type of interaction data corresponding to each sub-time period to obtain the normalized score of each type of interaction data corresponding to each sub-time period may specifically include the following steps:

s231, a first data volume and a second data volume corresponding to each type of interaction data are obtained, wherein the first data volume is the data volume corresponding to the sub-time period with the largest data volume, and the second data volume is the data volume corresponding to the sub-time period with the smallest data volume.

In this embodiment, after the data amount of each type of interactive data corresponding to each sub-time period included in the preset time period is obtained, the data amount corresponding to the sub-time period with the largest data amount and the data amount corresponding to the sub-time period with the smallest data amount can be obtained, so as to obtain the first data amount and the second data amount corresponding to each type of interactive data.

S232, for each type of interactive data, obtain a first difference between the data size corresponding to each sub-time period in the same type of interactive data and the second data size, and a second difference between the first data size and the second data size.

In order to obtain the normalization score corresponding to each sub-time period in each type of interaction data through subsequent calculation, a first difference between the data amount corresponding to each sub-time period in the same type of interaction data and the second data amount and a second difference between the first data amount and the second data amount may be obtained first.

In this embodiment, the first difference may be represented as: xi_j -Xi_min The second difference may be expressed as: xi_max -Xi_min . Wherein Xi_j Represents the data quantity corresponding to the jth sub-time period in the ith type of interaction data, Xi_max Indicating the data size corresponding to the sub-period with the largest data size in the ith type of interaction data, i.e. the first data size, Xi_min The representation represents the data volume corresponding to the sub-period with the minimum data volume in the ith type of interaction data, namely the second data volume.

S233, determining a ratio of the first difference to the second difference corresponding to each sub-time period in each type of interactive data as a normalization score corresponding to each sub-time period in each type of interactive data.

In this embodiment, when the normalized score corresponding to each sub-time period in each type of interactive data is determined, the normalized score may be obtained based on a ratio of a first difference value to a second difference value corresponding to each sub-time period in each type of interactive data.

As an embodiment, the normalized score corresponding to each sub-period in each type of interaction data may be determined by the following formula:

Xi_jnorm ＝(Xi_j -Xi_min )/(Xi_max -Xi_min )

wherein Xi_jnorm And expressing the normalized score corresponding to the jth sub-time period in the ith type of interaction data.

And S240, weighting the normalized scores of each type of interaction data corresponding to the same sub-time period aiming at each sub-time period to obtain the comprehensive scores corresponding to the interaction data of each sub-time period included in the preset time period.

In this embodiment, after obtaining the normalization score corresponding to each sub-time period in each type of interactive data, the normalization score of each type of interactive data in the same word time period may be weighted according to each sub-time period, so as to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period.

Exemplarily, it is assumed that the interactive data includes three categories of audience comment number, audience praise number and audience number, and the preset weights are W₁ 、W₂ And W₃ The preset time period comprises 4 sub-time periods of 60 seconds to 46 seconds of the 1 st sub-time period, 46 seconds to 32 seconds of the 2 nd sub-time period, 32 seconds to 18 seconds of the 3 rd sub-time period and 18 seconds to 4 seconds of the 4 th sub-time period, and the normalized score of the audience comment number in the first sub-time period is X1_1norm The normalized score of the viewer's approval number in the first sub-period is X2_1norm The normalized score of the number of viewers in the first sub-period is X3_1norm At this time, the comprehensive SCORE corresponding to the interactive data of the first sub-period is:

SCORE＝X1_1norm *W₁ +X2_1norm *W₂ +X3_1norm *W₃

similarly, the comprehensive scores corresponding to the interactive data of other sub-time periods can be obtained.

And S250, determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period.

S260, determining a target video frame from a plurality of video frames in the target sub-time period in the target live broadcasting room, and updating a live broadcasting cover of the target live broadcasting room by using the target video frame, wherein the live broadcasting cover is used for displaying the cover in the target live broadcasting room.

According to the live-broadcast cover generation method, under the condition that at least two types of interactive data exist, the comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period is obtained through the normalized score calculation corresponding to each sub-time period in each type of interactive data and the preset weight, the calculation accuracy of the comprehensive score corresponding to the interactive data of each sub-time period is improved, and the quality of a subsequently generated live-broadcast cover is further improved.

In some embodiments, the server may determine the target video frame by way of an online click-through rate test. In this case, please refer to fig. 6, fig. 6 is a flowchart illustrating a live cover generating method according to another embodiment of the present application, where the embodiment of the present application takes an example that the method is applied to a server in the live system shown in fig. 2 as an example, the method may include the following steps:

s300, according to a preset time interval, periodically obtaining interactive data distribution of a target live broadcast room in a preset time period before the current live broadcast time, wherein the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension.

And S310, grading the interactive data distribution in the preset time period in a segmented manner according to the preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interactive data indicates the interaction degree of the audience in the target live broadcast room.

And S320, determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period.

S330, based on a first preset rule, at least two video frames to be audited are determined from a plurality of video frames in the target sub-time period in the target live broadcast room.

The first preset rule is a rule which is preset in the server and is used for determining at least two video frames to be audited from a plurality of video frames in the target sub-time period. The first predetermined rule may be various.

Alternatively, the first preset rule may be a random selection rule. At this time, the server randomly selects at least two video frames from the plurality of video frames in the target sub-time period as video frames to be audited.

Optionally, the preset rule may also be a rule for selecting a video frame to be selected according to a time point in a sub-period, and specifically, the rule may refer to the foregoing policy for selecting a target video frame according to a time point in a sub-period, where a difference between the two rules is that the rule for selecting a video frame to be selected according to a time point in a sub-period in this embodiment determines a video frame to be selected that is subsequently sent to a host user for selection, and determines a target video frame from the video frame to be selected by selecting the video frame to be selected by the host user, and the policy for selecting a target video frame according to a time point in a sub-period in the foregoing embodiment determines that the policy is directly a cover broadcast of a target live broadcast.

Optionally, the preset rule may also be a quality scoring model in the foregoing embodiment, the multiple video frames in the target sub-period are respectively input into the video frame quality scoring model, quality scores corresponding to the video frames output by the video frame scoring model are obtained, then the quality is sorted from high to low, and then a preset number of video frames with the quality scores sorted in the front are determined as the video frames to be selected.

It should be noted that the above-mentioned various preset rules are also applicable to the case where there is one target sub-period or there are multiple target sub-periods.

It should be noted that the above-mentioned various preset rules are also applicable to the case where there is one target sub-period or there are multiple target sub-periods.

S340, performing qualification screening on the at least two video frames to be audited to obtain screened qualified video frames to be tested.

It can be understood that the cover is required to be displayed to the audience users, so that in consideration of the network specification, the qualified screening of the video frames to be checked can be performed first, and the screened qualified video frames to be tested are obtained.

The step of screening the qualification of the video frame to be checked refers to the step of screening whether the video content of the video frame to be checked meets the live broadcast content specified in the live broadcast specification, for example, the content of the video frame cannot relate to yellow, gambling or fraud and the like.

The qualification screening of the at least two video frames to be audited can be carried out in various screening modes. Optionally, the eligibility screening may be performed in a manual review manner, in this case, the server may send at least two video frames to be reviewed to the reviewer client, so that the reviewer performs review, and after the review by the reviewer, the review result is returned to the server, so that the server obtains the video frames to be tested that are eligible for screening according to the review result. Optionally, the eligibility screening may be performed on at least two video frames to be audited by means of an AI model. In this case, the AI model may be trained from a plurality of sample video frames carrying eligibility labels.

And S350, judging whether the number of the video frames to be tested is at least two.

After the video frames to be tested are acquired, the number of the video frames to be tested can be further judged, if the number of the video frames to be tested is at least two, the steps S370 to S380 are executed, and if the number of the video frames to be tested is one, the step S390 is executed.

S360, if the number of the video frames to be tested is at least two, the at least two video frames to be tested are respectively used as the test covers corresponding to the target live broadcast room, and the click rate test is carried out on each test cover to obtain the test click rate corresponding to each test cover.

In this embodiment, the on-line click rate test may be performed using partial traffic. Optionally, each video frame to be tested may be sent to the online for AB test as a test cover at a flow rate of 20%, and then the test click rate, i.e. the ratio of the click rate to the exposure, of each cover to be tested in the test process is recorded.

The AB test is to make two (A/B) or a plurality of (A/B/n) versions for a Web or App interface or process, make visitor groups (target population) with the same (similar) components randomly access the versions in the same time dimension, collect user experience data and service data of each group, analyze and evaluate the best version, and adopt the version formally.

And S370, determining a target video frame from at least two video frames to be tested based on the test click rate.

It can be understood that the test click rate may reflect the click effect of each test cover, and therefore, in this embodiment, the video frame corresponding to the test cover with the highest test click rate may be determined as the target video frame.

And S380, if the number of the video frames to be tested is one, determining the video frames to be tested as target video frames.

In some cases, the number of the video frames to be tested may be only one, and in this case, the on-line click rate test is not required, but the one video frame to be tested may be directly determined as the target video frame.

And S390, updating the live broadcast cover of the target live broadcast room by using the target video frame, wherein the live broadcast cover is used for displaying the cover of the target live broadcast room.

The live broadcast cover generation method includes the steps that at least two video frames to be audited are determined from a plurality of video frames in a target sub-time period in a target live broadcast room based on a first preset rule, the number of the video frames to be audited for qualification screening can be reduced, screening speed is increased, the video frames to be audited are screened for qualification, it can be guaranteed that a test cover displayed for audience users meets live broadcast standard requirements, and finally the live broadcast cover can be close to most of user preferences by determining the target video frames through an online click rate test mode, and live broadcast cover quality is further improved.

In addition, when determining a target video frame from a plurality of video frames in a target sub-period in a target live broadcast room in a manual selection manner, if all the video frames are directly given to a main broadcast user for selection, the main broadcast selection is time-consuming due to a large number of video frames, and a normal live broadcast behavior of the main broadcast may be affected, please refer to fig. 7, where fig. 7 is a flowchart of a live broadcast cover generation method provided in another embodiment of the present application, which is described in the embodiment of the present application by taking as an example a server in which the method is applied to the live broadcast system shown in fig. 2 as an example, the method may include the following steps:

s410, according to a preset time interval, periodically obtaining interactive data distribution of the target live broadcast room in a preset time period before the current live broadcast time, wherein the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension.

And S420, grading the interactive data distribution in the preset time period in a segmented manner according to the preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interactive data indicates the interaction degree of the audience in the target live broadcast room.

And S430, determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interaction data of each sub-time period.

S440, based on a second preset rule, at least two video frames to be selected are determined from a plurality of video frames of the target live broadcast room in the target sub-time period.

The second preset rule is a rule which is preset in the server and is used for determining at least two video frames to be selected from the plurality of video frames in the target sub-period. The second predetermined rule may be various.

Alternatively, the preset rule may be a random selection rule. At this time, the server randomly selects at least two video frames from a plurality of video frames in the target sub-time period as video frames to be selected.

Optionally, the preset rule may also be a rule for selecting a video frame to be selected according to a time point in a sub-period, and specifically, the rule may refer to the foregoing policy for selecting a target video frame according to a time point in a sub-period, where a difference between the two rules is that the rule for selecting a video frame to be selected according to a time point in a sub-period in this embodiment determines a video frame to be selected that is subsequently sent to a host user for selection, and determines a target video frame from the video frame to be selected by selecting the video frame to be selected by the host user, and the policy for selecting a target video frame according to a time point in a sub-period in the foregoing embodiment determines that the policy is directly a cover broadcast of a target live broadcast.

Optionally, the preset rule may also be a quality scoring model in the foregoing embodiment, the multiple video frames in the target sub-period are respectively input into the video frame quality scoring model to obtain quality scores corresponding to the video frames output by the video frame scoring model, then the quality is sorted from high to low, and then a preset number of video frames with quality scores sorted in the top are determined as the video frames to be selected.

It should be noted that the above-mentioned various preset rules are also applicable to the case where there is one target sub-period or there are multiple target sub-periods.

S450, at least two video frames to be selected are sent to the target anchor client corresponding to the target live broadcast room, so that the target anchor client generates a target video frame selection instruction.

After determining at least two video frames to be selected, the server can send the at least two video frames to be selected to a target anchor client corresponding to the target live broadcast room, so that the anchor client can render and display on a display interface for an anchor user to select.

After the anchor user selects a target video frame from at least two video frames to be selected, an instruction can be input at the anchor client, so that the anchor client can generate a corresponding target video frame selection instruction and send the target video frame selection instruction to the server.

And S460, receiving and responding to a target video frame selection instruction sent by the target anchor client, and determining a target video frame from at least two video frames to be selected.

As can be known from the foregoing, a large number of anchor users may exist at each time point in the server, that is, a large number of live broadcast rooms exist, so that, in order to enable the server to respond to a target video frame selection instruction sent by a target client, a target video frame is accurately determined from at least two video frames to be selected, optionally, the target video frame selection instruction may carry a corresponding user identifier or a live broadcast room identifier, and may also carry a corresponding video frame identifier.

And S470, updating the live broadcast cover of the target live broadcast room by using the target video frame, wherein the live broadcast cover is used for displaying the cover of the target live broadcast room.

After receiving the target video frame selection instruction, the server can update the live broadcast cover of the target live broadcast room by using the target video frame according to the user identification or the live broadcast room identification and the video frame identification.

The method for generating the live broadcast cover considers that even if the video frame to be selected is determined from the sub-time period with high audience interaction degree, the video frame with poor quality is still low and is likely to appear, therefore, in order to avoid determining the video frame with poor quality as the live broadcast cover, the video frame to be selected can be determined by primarily screening the video frame to be selected by a server, then the video frame to be selected is sent to a main broadcast client side for secondary screening of a main broadcast user, and finally the live broadcast cover is determined.

In addition, in combination with the above contents, as the live broadcast progresses, the live broadcast content of the anchor user changes at any time, therefore, in order to further enable the content of the live cover to be closer to the real-time live content, the server may circularly execute the steps of S410-S470 described above at certain time intervals, thereby periodically updating the live cover of the target live room, however, it is considered that if the time interval is too short, it is possible that the operation of the anchor user frequently selecting a target video frame from the video frames to be selected affects the normal live broadcast of the anchor user, and therefore, in some embodiments, please refer to fig. 8, fig. 8 is a flowchart illustrating a live cover generating method according to another embodiment of the present application, where the embodiment of the present application is described by taking an example of applying the method to a server in the live system shown in fig. 2, and the method may include the following steps:

and S510, receiving and responding to a live cover generation request sent by a main broadcast client corresponding to the target live broadcast room, wherein the live cover generation request comprises a preset time interval.

In some embodiments, the anchor user may set or change a preset time interval in a setting option of the live client before starting live broadcast through the live client, and after setting or changing the preset time interval, the anchor client may send a live broadcast cover generation request carrying the preset time interval to the server.

S520, according to a preset time interval, periodically acquiring the interactive data distribution of the target live broadcast room in a preset time period before the current live broadcast time, wherein the current live broadcast time is the time when the preset time interval arrives.

The server can acquire the preset time interval after receiving the live broadcast cover generation request carrying the preset time interval, so that the server can periodically acquire the interactive data distribution of the target live broadcast room in the preset time period before the current live broadcast time according to the preset time interval, wherein the current live broadcast time is the time when the preset time interval arrives.

Exemplarily, assuming a preset time interval of 10 seconds, the corresponding time instant is now 12: 00:00, the server obtains a target live broadcast room in 12: after the interactive data distribution in the preset time period before the time 00:00, the data distribution may be respectively performed at 12: the time 00:10 is that the target live broadcast room is acquired at 12: the distribution of the interaction data in the preset time period before the time 00:10 is carried out in a mode of 12: the time 00:20 for acquiring the target live broadcast is 12: the distribution of the interactive data in the preset time period before the time 00:20 is carried out in a mode of 12: the time 00:30 is obtained when the target live broadcast room is 12: and (5) distributing the interaction data in a preset time period before the time of 00:30, and so on. And wherein, at 12: at time 00:10, the corresponding current live broadcast time is 12: time 00:10, at 12: at time 00:20, the corresponding current live broadcast time is 12: time 00:20, at 12: at time 00:30, the corresponding current live broadcast time is 12: time 00: 30.

S530, grading the interactive data distribution in a preset time period in a segmented manner according to the preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interactive data indicates the interactive degree of the audience in the target live broadcast room.

And S540, determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interaction data of each sub-time period.

And S550, determining at least two video frames to be selected from the plurality of video frames in the target sub-time period in the target live broadcast room based on a second preset rule.

And S560, sending at least two video frames to be selected to the target anchor client corresponding to the target live broadcast room, so that the target anchor client generates a target video frame selection instruction.

And S570, receiving and responding to a target video frame selection instruction sent by the target anchor client, and determining a target video frame from at least two video frames to be selected.

And S580, updating a live broadcast cover of the target live broadcast room by using the target video frame, wherein the live broadcast cover is used for displaying the cover of the target live broadcast room.

According to the live broadcast cover generation method, the preset time interval is set by the anchor user, so that the updating frequency of the live broadcast cover can accord with the use habit of the anchor user, and the influence on the normal live broadcast of the anchor user is avoided.

Referring to fig. 9, fig. 9 is a flowchart illustrating a live cover page generating method according to another embodiment of the present application, where the method may be applied to a live system, where the live system includes a main broadcasting client, a server, and a viewer client, and the method may include the following steps:

s601, the server periodically obtains the interactive data distribution of the target live broadcast room in a preset time period before the current live broadcast time according to a preset time interval.

S602, the server carries out sectional statistics on the interactive data distribution corresponding to each type of interactive data in the preset time period according to the preset time length to obtain the data volume of each type of interactive data corresponding to each sub-time period included in the preset time period.

And S603, respectively carrying out normalization processing on the data volume of each type of interactive data corresponding to each sub-time period by the server to obtain a normalization score corresponding to each sub-time period in each type of interactive data.

S604, the service performs weighting processing on the normalized score of each type of interaction data corresponding to the same sub-time period aiming at each sub-time period to obtain a comprehensive score corresponding to the interaction data of each sub-time period included in the preset time period.

And S605, the server determines a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period.

S606, the server determines at least two video frames to be selected from the plurality of video frames in the target sub-time period based on a second preset rule.

S607, the server sends at least two video frames to be selected to the target anchor client corresponding to the target live broadcast room, so that the target anchor client generates a target video frame selection instruction.

S608, the anchor client responds to the selection instruction of the anchor user for at least two video frames to be selected, and generates a target video frame selection instruction.

And S609, the anchor client sends a target video frame selection instruction to the server.

In some embodiments, the above steps S606 to S609, that is, determining the target video frame from the plurality of video frames in the target sub-period in the target live broadcast room, may also be implemented by an online click rate test, and the specific implementation process is detailed in the foregoing steps S330 to S380, and is not described herein again.

S610, the server receives and responds to a target video frame selection instruction sent by the target anchor client, and determines a target video frame from at least two video frames to be selected.

S611, the server updates the live broadcast cover of the target live broadcast room by using the target video frame, and sends the live broadcast cover corresponding to the target live broadcast room to the audience client.

And S612, displaying the live broadcast cover corresponding to the target live broadcast room by the audience client.

The application provides a live broadcast cover generation method, on one hand, because the comprehensive score obtained by calculation based on audience interaction data can reflect the interaction degree of audiences in a live broadcast room, the interaction degree is high, the possibility of occurrence of wonderful covers is higher, the live broadcast cover determined from the comprehensive score can attract the interests of the audiences, so that the finally determined live broadcast cover quality is higher, on the other hand, because the interaction data distribution in the preset time period before the current live broadcast moment is obtained, the interaction data distribution of all live broadcast time periods after starting broadcast is avoided, the finally determined content of the live broadcast cover can be closer to the real-time live broadcast content, on the other hand, a server primarily screens and determines a video frame to be selected, then sends the video frame to be selected to a main broadcast client for secondary screening of a main broadcast user, finally determines the live broadcast cover, and combines a manual selection mode through a machine device, the cover quality is further improved, and the effect of the live broadcast cover is finally improved.

It should be noted that the present application provides some specific examples of the foregoing implementable embodiments, and on the premise of no conflict, the examples of the embodiments may be combined at will to form a new live cover generation method. It should be understood that a new live cover generation method formed by any combination of examples is within the scope of the present application.

Referring to fig. 10, fig. 10 is a block diagram illustrating a live cover generating apparatus 700 according to an embodiment of the present application, where the apparatus 700 may include: anacquisition module 710, asegment scoring module 720, a targetsub-period determination module 730, and a livecover update module 740.

The obtainingmodule 710 is configured to periodically obtain, according to a preset time interval, interactive data distribution in a preset time period before a current live broadcast time of a target live broadcast room, where the interactive data distribution indicates a distribution situation of interactive data generated by an audience in the target live broadcast room in a time dimension, and the current live broadcast time is a time when the preset time interval arrives.

And thesegmentation grading module 720 is configured to perform segmentation grading on the interaction data distribution in the preset time period according to the preset time length to obtain a comprehensive score corresponding to the interaction data of each sub-time period included in the preset time period, where the comprehensive score corresponding to the interaction data indicates the interaction degree of the audience in the target live broadcast room.

And a target sub-period determiningmodule 730, configured to determine a target sub-period from each sub-period based on the composite score corresponding to the interaction data of each sub-period.

And the live broadcastcover updating module 740 is configured to determine a target video frame from the plurality of video frames in the target sub-time period in the target live broadcast room, and update a live broadcast cover of the target live broadcast room by using the target video frame, where the live broadcast cover is used for displaying a cover in the target live broadcast room.

As an implementation manner, the livecover updating module 740 is further configured to determine, based on a first preset rule, at least two video frames to be audited from among the plurality of video frames in the target sub-time period in the target live broadcasting room; screening the qualification of the at least two video frames to be examined to obtain screened qualified video frames to be tested; if the number of the video frames to be tested is at least two, the at least two video frames to be tested are respectively used as test covers corresponding to the target live broadcast room, and the click rate test is carried out on each test cover to obtain the test click rate corresponding to each test cover; and determining the target video frame from the at least two video frames to be tested based on the test click rate.

As an embodiment, the livecover updating module 740 is further configured to determine at least two video frames to be selected from the plurality of video frames in the target sub-period based on a second preset rule; at least two video frames to be selected are sent to a target anchor client corresponding to a target live broadcast room, so that the target anchor client generates a target video frame selection instruction; and receiving and responding to a target video frame selection instruction sent by the target anchor client, and determining a target video frame from at least two video frames to be selected.

As an implementation manner, the apparatus 700 further includes a receiving module, where the receiving module is configured to receive and respond to a live cover generation request sent by a host client corresponding to the target live broadcast room, where the live cover generation request includes a preset time interval.

In one embodiment, the interaction data includes at least two types of interaction data, in which case thesegmentation scoring module 720 includes: the device comprises a segmentation statistical unit, a normalization processing unit and a weighting processing unit. The device comprises a segmentation statistical unit, a data processing unit and a data processing unit, wherein the segmentation statistical unit is used for carrying out segmentation statistics on interaction data distribution corresponding to each type of interaction data in a preset time period according to a preset time length to obtain the data volume of each type of interaction data corresponding to each sub-time period included in the preset time period; the normalization processing unit is used for respectively performing normalization processing on the data quantity of each type of interaction data corresponding to each sub-time period to obtain a normalization score corresponding to each sub-time period in each type of interaction data; and the weighting processing unit is used for weighting the normalized scores of each type of interaction data corresponding to the same sub-time period aiming at each sub-time period to obtain the comprehensive scores corresponding to the interaction data of each sub-time period included in the preset time period.

As an implementation manner, the normalization processing unit is further configured to obtain a first data volume and a second data volume corresponding to each type of interaction data, where the first data volume is a data volume corresponding to a sub-time period with the largest data volume, and the second data volume is a data volume corresponding to a sub-time period with the smallest data volume; for each type of interaction data, acquiring a first difference value between the data quantity corresponding to each sub-time period in the same type of interaction data and the second data quantity, and a second difference value between the first data quantity and the second data quantity; and determining the ratio of the first difference value and the second difference value corresponding to each sub-time period in each type of interaction data as the normalization score corresponding to each sub-time period in each type of interaction data.

The interactive data may include one or more of a number of viewer comments, a number of viewer likes, and a number of viewers.

The application provides a live broadcast cover generation device, on one hand, because the comprehensive score obtained based on audience interaction data calculation can reflect the interaction degree of audiences in a live broadcast room, the interaction degree is high, the possibility of occurrence of wonderful covers is higher, the live broadcast cover determined from the excellent results can attract the interests of audiences, so that the finally determined live broadcast cover quality is higher, on the other hand, because the interaction data distribution in the preset time period before the current live broadcast time is obtained, the interaction data distribution of all live broadcast time periods after starting broadcast is avoided, the finally determined content of the live broadcast cover can be closer to the real-time live broadcast content, on the other hand, a server preliminarily screens and determines a video frame to be selected, then sends the video frame to be selected to a main broadcast client for secondary screening of a main broadcast user, finally determines the live broadcast cover, and combines a manual selection mode through a machine device, the quality of the cover is further improved.

It should be noted that the device embodiment and the method embodiment in the present application correspond to each other, and specific principles in the device embodiment may refer to the contents in the method embodiment, which is not described herein again.

An electronic device provided by the present application will be described below with reference to fig. 11.

Referring to fig. 11, based on the above live cover generating method, anotherelectronic device 200 including aprocessor 104 capable of executing the live cover generating method is further provided in the embodiment of the present application, where theelectronic device 200 may be a smart phone, a tablet computer, a portable computer, or the like.Electronic device 200 also includesmemory 104,network module 106, andscreen 108. Thememory 104 stores programs that can execute the content of the foregoing embodiments, and theprocessor 102 can execute the programs stored in thememory 104.

Processor 102 may include, among other things, one or more cores for processing data and a message matrix unit. Theprocessor 102 interfaces with various components throughout theelectronic device 200 using various interfaces and circuitry to perform various functions of theelectronic device 200 and process data by executing or executing instructions, programs, code sets, or instruction sets stored in thememory 104 and invoking data stored in thememory 104. Alternatively, theprocessor 102 may be implemented in hardware using at least one of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). Theprocessor 102 may integrate one or a combination of a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), a modem, and the like. Wherein, the CPU mainly processes an operating system, a user interface, an application program and the like; the GPU is used for rendering and drawing display content; the modem is used to handle wireless communications. It is understood that the modem may not be integrated into theprocessor 102, but may be implemented by a communication chip.

TheMemory 104 may include a Random Access Memory (RAM) or a Read-Only Memory (Read-Only Memory). Thememory 104 may be used to store instructions, programs, code sets, or instruction sets. Thememory 104 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing various method embodiments described below, and the like. The storage data area may also store data created by the terminal 100 in use, such as a phonebook, audio-video data, chatting log data, and the like.

Thenetwork module 106 is configured to receive and transmit electromagnetic waves, and achieve interconversion between the electromagnetic waves and the electrical signals, so as to communicate with a communication network or other devices, for example, an audio playing device. Thenetwork module 106 may include various existing circuit elements for performing these functions, such as an antenna, a radio frequency transceiver, a digital signal processor, an encryption/decryption chip, a Subscriber Identity Module (SIM) card, memory, and so forth. Thenetwork module 106 may communicate with various networks such as the internet, an intranet, a wireless network, or with other devices via a wireless network. The wireless network may include a cellular telephone network, a wireless local area network, or a metropolitan area network. For example, thenetwork module 106 may interact with a base station.

Thescreen 108 may display interface content and may also be used to respond to touch gestures.

It should be noted that, in order to implement more functions, theelectronic device 200 may also protect more devices, for example, may also protect a structured light sensor for acquiring face information or may also protect a camera for acquiring an iris.

Referring to fig. 12, a block diagram of a computer-readable storage medium provided in an embodiment of the present application is shown. The computer readable medium 1100 has stored therein a program code that can be called by a processor to execute the method described in the above method embodiments.

The computer-readable storage medium 1100 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read only memory), an EPROM, a hard disk, or a ROM. Alternatively, the computer-readable storage medium 1100 includes a non-volatile computer-readable storage medium. The computer readable storage medium 1100 has storage space forprogram code 1110 for performing any of the method steps of the method described above. The program code can be read from or written to one or more computer program products. Theprogram code 1110 may be compressed, for example, in a suitable form.

Based on the above-mentioned live cover generation method, according to an aspect of an embodiment of the present application, there is provided a computer program product or a computer program, the computer program product or the computer program including computer instructions, the computer instructions being stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method provided in the various alternative implementations described above.

To sum up, in the live cover generation method, apparatus, electronic device, storage medium, and computer program product or computer program provided in the embodiments of the present application, on one hand, since the comprehensive score obtained by calculation based on the audience interaction data can reflect the degree of interaction of the audience in the live room, the degree of interaction is high, and the probability of occurrence of a highlight cover is high, the live cover determined therefrom can attract the interest of the audience better, so that the quality of the finally determined live cover is high, on the other hand, since the obtained interaction data distribution in the preset time period before the current live broadcast time is obtained, the obtaining of the interaction data distribution in all live broadcast time periods since the start of broadcast is avoided, so that the content of the finally determined live cover can be closer to the real-time content, on the other hand, the server primarily screens and determines the video frame to be selected, and then sends the video frame to be selected to the anchor client for secondary screening by the anchor user, finally, the live cover is determined, the cover quality is further improved through a mode of combining mechanical equipment with manual selection, and then the effect of the live cover is finally improved.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not necessarily depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.

Claims (10)

1. A live cover generation method is characterized by comprising the following steps:

according to a preset time interval, periodically acquiring interactive data distribution of a target live broadcast room in a preset time period before a current live broadcast time, wherein the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension, and the current live broadcast time is the time when the preset time interval arrives;

grading the interactive data distribution in the preset time period in a segmented manner according to the preset time length to obtain a comprehensive score corresponding to the interactive data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interactive data indicates the interaction degree of audiences in the target live broadcast room;

determining a target sub-time period from the sub-time periods based on the comprehensive scores corresponding to the interactive data of the sub-time periods;

determining a target video frame from a plurality of video frames of the target live broadcast room in the target sub-time period, and updating a live broadcast cover of the target live broadcast room by using the target video frame, wherein the live broadcast cover is used for displaying the cover in the target live broadcast room;

the method comprises the following steps of grading the interactive data distribution in the preset time period in a segmented manner according to the preset time length to obtain the comprehensive scores corresponding to the interactive data of each sub-time period included in the preset time period, wherein the comprehensive scores comprise:

according to the preset time length, carrying out segmentation grading on the interactive data distribution in the preset time period to obtain interactive scores corresponding to the interactive data of each sub-time period included in the preset time period;

obtaining a comprehensive score corresponding to each interactive data based on the interactive score corresponding to the interactive data of each sub-time period and the corresponding similarity coefficient; and setting the similarity coefficient corresponding to the sub-time period based on the distance between the sub-time period and the current live broadcast time, wherein the similarity coefficient corresponding to the sub-time period closer to the current live broadcast time is larger.

2. The method of claim 1, wherein determining a target video frame from a plurality of video frames in the target sub-period from the target live broadcast room comprises:

determining at least two video frames to be audited from a plurality of video frames in the target sub-time period in the target live broadcast room based on a first preset rule;

screening the qualification of the at least two video frames to be examined to obtain screened qualified video frames to be tested;

if the number of the video frames to be tested is at least two, the at least two video frames to be tested are respectively used as test covers corresponding to the target live broadcast room, and the click rate test is carried out on each test cover to obtain the test click rate corresponding to each test cover;

and determining the target video frame from the at least two video frames to be tested based on the test click rate.

3. The method of claim 1, wherein determining a target video frame from a plurality of video frames in the target sub-period from the target live broadcast room comprises:

determining at least two video frames to be selected from a plurality of video frames in the target sub-time period in the target live broadcast room based on a second preset rule;

sending the at least two video frames to be selected to a target anchor client corresponding to the target live broadcast room so that the target anchor client generates a target video frame selection instruction;

and receiving and responding to a target video frame selection instruction sent by the target anchor client, and determining a target video frame from the at least two video frames to be selected.

4. The method as claimed in claim 3, wherein before the periodically obtaining, at preset time intervals, the interactive data distribution of the target live broadcast room in a preset time period before the current live broadcast time, the method further comprises:

and receiving and responding to a live broadcast cover generation request sent by a main broadcast client corresponding to the target live broadcast room, wherein the live broadcast cover generation request comprises the preset time interval.

5. The method of claim 1, wherein the interaction data comprises one or more of a number of viewer comments, a number of viewer likes, and a number of viewers combined.

6. A live cover creation apparatus, comprising:

the system comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for acquiring interactive data distribution of a target live broadcast room in a preset time period before the current live broadcast time, and the interactive data distribution indicates the distribution condition of interactive data generated by audiences in the target live broadcast room in a time dimension;

the segmentation grading module is used for performing segmentation grading on the interaction data distribution in the preset time period according to a preset time length to obtain a comprehensive score corresponding to the interaction data of each sub-time period included in the preset time period, wherein the comprehensive score corresponding to the interaction data indicates the interaction degree of audiences in the target live broadcast room;

the segmentation grading module is specifically used for performing segmentation grading on the interaction data distribution in the preset time period according to the preset time length to obtain interaction scores corresponding to the interaction data of each sub-time period included in the preset time period; obtaining a comprehensive score corresponding to each interactive data based on the interactive score corresponding to the interactive data of each sub-time period and the corresponding similarity coefficient; the similarity coefficient corresponding to the sub-time period is set based on the distance between the sub-time period and the current live broadcast time, wherein the similarity coefficient corresponding to the sub-time period closer to the current live broadcast time is larger;

the target sub-time period determining module is used for determining a target sub-time period from each sub-time period based on the comprehensive score corresponding to the interactive data of each sub-time period;

and the live broadcast cover updating module is used for determining a target video frame from a plurality of video frames in the target sub-time period in the target live broadcast room, updating the live broadcast cover of the target live broadcast room by using the target video frame, and displaying the live broadcast cover of the target live broadcast room.

7. The apparatus of claim 6, wherein the live cover update module is further configured to determine at least two video frames to be reviewed from a plurality of video frames in the target sub-period in the target live broadcast room based on a first preset rule; screening the qualification of the at least two video frames to be examined to obtain screened qualified video frames to be tested; if the number of the video frames to be tested is at least two, the at least two video frames to be tested are respectively used as test covers corresponding to the target live broadcast room, and the click rate test is carried out on each test cover to obtain the test click rate corresponding to each test cover; and determining the target video frame from the at least two video frames to be tested based on the test click rate.

8. The apparatus of claim 6, wherein the live cover update module is further configured to determine at least two video frames to be selected from the plurality of video frames in the target sub-period in the target live room based on a second preset rule; sending the at least two video frames to be selected to a target anchor client corresponding to the target live broadcast room so that the target anchor client generates a target video frame selection instruction; and receiving and responding to a target video frame selection instruction sent by the target anchor client, and determining a target video frame from the at least two video frames to be selected.

9. An electronic device, comprising:

one or more processors;

a memory;

one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more programs configured to perform the method of any of claims 1-5.

10. A computer-readable storage medium, having stored thereon program code that can be invoked by a processor to perform the method according to any one of claims 1 to 5.

Images