WO2023116233A1 - 视频卡顿预测方法、装置、设备和介质 - Google Patents

视频卡顿预测方法、装置、设备和介质 Download PDF

Info

Publication number
WO2023116233A1
WO2023116233A1 PCT/CN2022/129721 CN2022129721W WO2023116233A1 WO 2023116233 A1 WO2023116233 A1 WO 2023116233A1 CN 2022129721 W CN2022129721 W CN 2022129721W WO 2023116233 A1 WO2023116233 A1 WO 2023116233A1
Authority
WO
WIPO (PCT)
Prior art keywords
video
current
downloaded
freeze
information
Prior art date
Application number
PCT/CN2022/129721
Other languages
English (en)
French (fr)
Inventor
孙黎阳
秦彦源
马茜
Original Assignee
北京字节跳动网络技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京字节跳动网络技术有限公司 filed Critical 北京字节跳动网络技术有限公司
Publication of WO2023116233A1 publication Critical patent/WO2023116233A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
    • H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
    • H04N21/24—Monitoring of processes or resources, e.g. monitoring of server load, available bandwidth, upstream requests
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
    • H04N21/24—Monitoring of processes or resources, e.g. monitoring of server load, available bandwidth, upstream requests
    • H04N21/2402—Monitoring of the downstream path of the transmission network, e.g. bandwidth available
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
    • H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
    • H04N21/44209—Monitoring of downstream path of the transmission network originating from a server, e.g. bandwidth variations of a wireless network

Definitions

  • the present disclosure relates to the technical field of the Internet, for example, to a video freeze prediction method, device, device and medium.
  • the present disclosure provides a video freezing prediction method, device, equipment and medium, to predict the video freezing of the current video to be downloaded, so that the result of the video freezing can be known in advance, thereby effectively ensuring the fluency of video playback, and improving user experience. viewing experience.
  • An embodiment of the present disclosure provides a video freezing prediction method, including:
  • the current playing information and the current service information determine the freeze prediction result when playing the currently to-be-downloaded video.
  • An embodiment of the present disclosure also provides a video freezing prediction device, including:
  • the current playback information acquisition module is configured to obtain the current playback information when playing the currently downloaded video
  • the current service information acquisition module is configured to obtain the current service information of the server
  • the freeze prediction module is configured to determine the freeze prediction result when playing the current video to be downloaded according to the preset freeze prediction model, the current playing information and the current service information.
  • An embodiment of the present disclosure also provides an electronic device, including:
  • processors one or more processors
  • memory configured to store one or more programs
  • the one or more processors are made to implement the video freezing prediction method provided in any embodiment of the present disclosure.
  • An embodiment of the present disclosure further provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the method for predicting video freezing as provided in any embodiment of the present disclosure is implemented.
  • FIG. 1 is a flow chart of a video freeze prediction method provided by Embodiment 1 of the present disclosure
  • FIG. 2 is an example of a video freezing prediction process involved in Embodiment 1 of the present disclosure
  • FIG. 3 is a structural example of a preset freeze prediction model involved in Embodiment 1 of the present disclosure
  • FIG. 4 is a flow chart of a video freeze prediction method provided in Embodiment 2 of the present disclosure.
  • FIG. 5 is an example of a video freezing prediction process involved in Embodiment 2 of the present disclosure.
  • FIG. 6 is a schematic structural diagram of a video freezing prediction device provided by Embodiment 3 of the present disclosure.
  • FIG. 7 is a schematic structural diagram of an electronic device provided by Embodiment 4 of the present disclosure.
  • the term “comprise” and its variations are open-ended, ie “including but not limited to”.
  • the term “based on” is “based at least in part on”.
  • the term “one embodiment” means “at least one embodiment”; the term “another embodiment” means “at least one further embodiment”; the term “some embodiments” means “at least some embodiments.” Relevant definitions of other terms will be given in the description below.
  • Fig. 1 is a flow chart of a video freeze prediction method provided by Embodiment 1 of the present disclosure.
  • This embodiment is applicable to the situation of predicting in advance whether playback freeze occurs when playing the currently to-be-downloaded video, for example, it can be used In the application scenario of freeze prediction for live video or on-demand video.
  • the method can be executed by a device for predicting video freezing, which can be realized by software and/or hardware, and integrated in a client or a server. As shown in Figure 1, the method includes the following steps.
  • the currently downloaded video may refer to a downloaded video currently being played by the client.
  • the live video or on-demand video can be segmented according to time to obtain multiple video clips in the live video or on-demand video, so that the client can download and play multiple video clips in sequence Play live video or on-demand video.
  • the currently downloaded video may refer to a downloaded live video segment or a downloaded on-demand video segment currently being played by the client.
  • the current playing information may include: the current network information of the client, the length of the currently cached video, and the information of the video currently being downloaded.
  • the current network information may include: packet loss rate, current bandwidth, maximum bandwidth, maximum network delay, minimum network delay, current round-trip delay, minimum round-trip delay, and maximum round-trip delay;
  • video information currently being downloaded may Including: video frame rate, video code rate and video data size.
  • the client When the client is playing the currently downloaded video, it can obtain the current playback information corresponding to the currently downloaded video in real time, so as to make real-time prediction of video freeze; it can also set the corresponding time interval based on the length of the currently cached video, and based on this The current playback information corresponding to the currently downloaded video is obtained regularly at intervals, so as to reduce the number of freeze predictions on the basis of ensuring the smoothness of video playback, thereby saving device resources. If this method is applied to the client, the client can directly obtain the current playback information. If the method is applied to the server, the client can send the obtained current playing information to the server, so that the server can receive the current playing information sent by the client. By performing video freeze prediction operation on the server based on the current playback information sent by the client, server resources can be fully utilized to quickly and accurately predict video freezes for the current video to be downloaded, improving prediction accuracy and efficiency.
  • the current service information may refer to information of the current server itself that may affect the video download speed.
  • the current service information may include: current server load, content delivery network (Content Delivery Network, CDN) vendor information, and downloadable video volume information cached by the CDN.
  • CDN Content Delivery Network
  • downloadable video volume information cached by the CDN may refer to the length of the downloadable live video segment or the length of the downloadable on-demand video segment currently cached in the CDN node of the server.
  • the server can send the currently collected current service information to the client, so that the client can receive the current service information sent by the server. If the method is applied to the server, the server can directly obtain the current service information, so as to perform stall prediction based on the current service information.
  • current playback information and current service information determine a freeze prediction result when playing the currently to-be-downloaded video.
  • the current video to be downloaded may refer to a video currently waiting to be downloaded by the client.
  • the video currently to be downloaded may refer to a live video segment or an on-demand video segment currently waiting to be downloaded by the client.
  • the freezing prediction result may include: whether or the probability of freezing occurs within a preset time period after the current moment.
  • the preset time period may be a time period set in advance based on business requirements, and the preset time period may include the time period from the current moment to the playing moment corresponding to the video to be downloaded.
  • the freeze prediction result may also include: freeze prediction duration, that is, the length of time during which freeze is predicted to occur within a preset time period, so as to meet different business needs and scenarios.
  • the preset freezing prediction model may be a neural network model that is preset and used to predict whether video freezing occurs when the video currently to be downloaded is subsequently played.
  • the preset freeze prediction model in this embodiment may be obtained through pre-training based on sample data.
  • the current playback information and current service information can be input into the pre-trained preset freeze prediction model to predict video freeze, and based on the output of the preset freeze prediction model, the client can obtain the time when the client is playing the current video to be downloaded.
  • the preset lagging prediction model can predict lagging based on current playback information and current service information at the same time, thereby improving the accuracy of lagging prediction.
  • S130 may include: preprocessing the current playback information to obtain the target playback information, and preprocessing the current service information to obtain the target service information; inputting the target playback information and the target service information into the preset freeze prediction The freeze prediction is performed in the model to obtain the freeze prediction result when playing the video currently to be downloaded.
  • performing preprocessing on the current playback information to obtain the target playback information may include: performing data standardization processing on the current playback information based on a preset normalization model to obtain the target playback information.
  • Preprocessing the current service information to obtain the target service information may include: performing data standardization processing on the current service information based on a preset standardized model to obtain the target service information.
  • Figure 2 shows an example of a video freeze prediction process.
  • the preset standardization model can be used to quickly perform data standardization processing on current broadcast information and current service information, so as to improve processing efficiency and ensure the uniformity of input information.
  • the preset standardized model regularly updates parameters based on the playback information and service information obtained within the latest preset duration.
  • this method is applied to the server, and the server can store the current playback information reported by the client each time and the current service information collected each time in the database, so that the latest stored playback information and service information can be used , to update the parameters of the preset standardized model.
  • the broadcast information and service information stored in the latest day can be used to update the average value and variance used for normalizing each information in the preset normalization model, so that the accuracy of freeze prediction can be improved.
  • this embodiment can periodically collect the current playback information and the current service information of the acquisition server reported by the clients of some users (that is, users who have not performed freeze prediction), and store the current playback information and current service information in the database middle. These users can report whether there is a freeze when actually playing the video to be downloaded, that is, the actual freeze result is reported to the server, and the server stores the actual freeze result in the database.
  • this embodiment can update the weight of the preset freeze prediction model based on the playback information and service information stored in the latest preset duration and the corresponding actual freeze results, for example, the latest day or
  • the multi-day broadcast information and service information and the corresponding actual lagging results conduct a new round of training on the preset lagging prediction model in order to improve the accuracy of lagging prediction.
  • the preset freeze prediction model may include: an output layer and at least one residual network (Residual Network, ResNet); each ResNet residual network includes: a first fully connected layer, a first activation layer , the second fully connected layer and the second activation layer; wherein, the output of the first fully connected layer is used as the input of the first activated layer; the output of the first activated layer is used as the input of the second fully connected layer, and the output of the second fully connected layer The output is skipped to the input of the first fully connected layer, and the summed result after the skipped connection is used as the input of the second activation layer.
  • ResNet residual network
  • the activation function used by the first activation layer and the second activation layer may be a Rectified Linear Unit (ReLU) activation function.
  • the activation function used in the output layer may be a Sigmoid function.
  • the information output by the output layer can be the probability of freezing. When the probability is less than 0.5, it means that the prediction will not occur. When the probability is greater than 0.5, it means that the prediction will occur.
  • Figure 3 shows an example of the structure of a preset freeze prediction model.
  • the preset freeze prediction model includes two ResNet residual networks
  • the accuracy of freeze prediction is the highest, so the preset freeze prediction model composed of two ResNet residual networks can be selected.
  • the prediction process of the preset freeze prediction model in Figure 3 is: input the target playback information and target service information into the first fully connected layer in the first ResNet residual network to obtain the first fully connected information , input the first fully connected information to the first activation layer to obtain the first activation information, input the first activation information to the second fully connected layer to obtain the second fully connected information, and combine the second fully connected information with the input
  • the target playback information is added to the target service information, and the obtained addition result is input into the second activation layer to obtain the second activation information.
  • FIG. 4 is a flow chart of a video freeze prediction method provided by Embodiment 2 of the present disclosure.
  • this embodiment describes the process of performing video processing on a video to be downloaded that is predicted to freeze.
  • the explanations of terms that are the same as or corresponding to those in the above-mentioned multiple embodiments will not be repeated here.
  • the video freezing prediction method provided by this embodiment includes the following steps.
  • freeze prediction result is: freeze occurs within a preset time period after the current moment, then based on the preset freeze processing method, perform video processing on the video to be downloaded corresponding to the preset time period, so as to reduce the number of videos to be downloaded.
  • the preset freezing processing method can be preset, and is used to reduce the amount of video data transmitted between the server and the client, so as to avoid unnecessary video freezing and ensure the smoothness of video playback.
  • the preset stall handling methods may include multiple methods.
  • the preset freezing processing method may refer to a method of reducing a video bit rate or a method of reducing a video frame rate.
  • Figure 5 shows an example of a video freeze prediction process.
  • the freeze avoidance policy model in the server can make corresponding decisions based on the freeze prediction results output by the preset freeze prediction model. For example, when the stuttering prediction result is that stuttering occurs within a preset period of time after the current moment, it can be determined to reduce the transmission video bit rate or reduce the stuttering processing method of the transmitted video frame rate, so that the live broadcast/on-demand server based on the stuttering
  • the processing method is to perform video processing on the videos to be downloaded that need to be downloaded within the preset time period, thereby reducing the amount of video data transmitted between the server and the client, avoiding unnecessary video freezes, and ensuring the smoothness of video playback , which improves the user experience.
  • the freezing process of the video to be downloaded that is predicted to be stuck can be implemented in the following three ways:
  • the "processing the video to be downloaded corresponding to the preset time period" in S440 may include: based on the first video bit rate corresponding to the video currently being downloaded, determine A second video bit rate less than the first video bit rate; based on the pre-stored video to be downloaded corresponding to the preset time period under each video bit rate, determine the target to be downloaded corresponding to the preset time period under the second video bit rate video.
  • each video to be downloaded with a different video bit rate may be pre-generated and stored.
  • a video bit rate lower than the currently downloaded first video bit rate can be selected from the existing video bit rates as the second video bit rate, and each video bit rate to be downloaded can be selected from the pre-stored
  • the target video to be downloaded corresponding to the preset time period under the second video bit rate can be quickly obtained, so that the client can download the target video to be downloaded, thereby improving the efficiency of freeze processing, and reducing the video bit rate by reducing
  • the amount of downloaded video data is reduced, so that the target video to be downloaded can be downloaded more quickly, and the situation of video freeze in subsequent playback is avoided.
  • the "processing the video to be downloaded corresponding to the preset time period" in S440 may include: extracting video frames from the video to be downloaded corresponding to the preset time period , to obtain the extracted target video to be downloaded.
  • the video frame rate of the target video to be downloaded can be reduced by extracting the video frame of each video to be downloaded corresponding to the preset time period, thereby reducing the amount of downloaded video data and avoiding subsequent occurrences of Video freezes.
  • the "processing the video to be downloaded corresponding to the preset time period" in S440 may include: based on the target video encoding method, performing video processing on the video to be downloaded corresponding to the preset time period Download the video and perform transcoding processing to obtain the processed target video to be downloaded; wherein, the video bit rate of the target video coding method is lower than the video bit rate of the current video coding method; or, the video frame rate of the target video coding method is lower than the current video coding method mode video frame rate.
  • the target video encoding method with a lower video bit rate and/or lower video frame rate can be used to transcode the video to be downloaded, thereby reducing the video bit rate and/or video frame of the processed target video to be downloaded rate, thereby reducing the amount of transmitted video data, and avoiding subsequent video freezes.
  • video processing can be performed on the video to be downloaded corresponding to the preset time period based on the preset freeze processing method, Reduce the amount of video data transmitted between the server and the client by reducing the video bit rate or video frame rate corresponding to the video to be downloaded, avoiding unnecessary video freezes, ensuring smooth video playback, and improving user viewing experience .
  • the freeze prediction result also includes: the freeze prediction duration when freeze occurs within a preset time period after the current moment.
  • step S440 "based on the preset freezing processing method, perform video processing on the video to be downloaded corresponding to the preset time period" may include: based on the preset freezing processing method, performing video processing on the video to be downloaded corresponding to the freezing prediction duration Video for video processing.
  • the freeze prediction result can also predict the freeze prediction time corresponding to the freeze prediction result when it is predicted that there will be freeze.
  • the video bit rate or video frame rate of the video to be downloaded corresponding to the frame prediction duration does not need to process other videos to be downloaded, so that the video viewing quality can be guaranteed and the user experience can be improved.
  • the following is an embodiment of the video freezing prediction device provided by the embodiment of the present disclosure.
  • This device belongs to the same concept as the video freezing prediction method of the above-mentioned embodiment, and the details are not described in detail in the embodiment of the video freezing prediction device. , you can refer to the above-mentioned embodiments.
  • Fig. 6 is a schematic structural diagram of a video freeze prediction device provided by Embodiment 3 of the present disclosure. This embodiment is applicable to the situation of predicting in advance whether playback freeze occurs when playing the video currently to be downloaded, for example, it can be used In the application scenario of freeze prediction for live video or on-demand video.
  • the device includes: a current playback information acquisition module 610 , a current service information acquisition module 620 and a freeze prediction module 630 .
  • the current play information acquisition module 610 is set to acquire the current play information when playing the currently downloaded video; the current service information acquisition module 620 is set to acquire the current service information of the server; The freeze prediction model, current playback information, and current service information determine the freeze prediction result when playing the current video to be downloaded.
  • the current service information includes: the current load of the server, CDN vendor information, and downloadable video volume information cached by the CDN.
  • the preset freeze prediction model includes: an output layer and at least one ResNet; each ResNet includes: a first fully connected layer, a first activation layer, a second fully connected layer and a second activation layer;
  • the output of the first fully connected layer is used as the input of the first activation layer; the output of the first activation layer is used as the input of the second fully connected layer, and the output of the second fully connected layer is skip-connected with the input of the first fully connected layer , the addition result after the skip connection is used as the input of the second activation layer.
  • the freeze prediction module 630 may include:
  • the preprocessing unit is configured to preprocess the current playback information to obtain the target playback information, and preprocess the current service information to obtain the target service information;
  • the stutter prediction unit is configured to input the target playback information and target service information into a preset stutter prediction model for stutter prediction, and obtain a stutter prediction result when playing the currently to-be-downloaded video.
  • the preprocessing unit is set to: based on the preset standardized model, perform data standardization processing on the current playback information to obtain the target playback information; wherein, the preset standardized model timing is based on the latest preset duration Update the parameters of the playback information obtained in the file.
  • the device also includes:
  • the freeze processing module is set to if the freeze prediction result is: freeze occurs within the preset time period after the current moment, then based on the preset freeze processing method, video processing is performed on the video to be downloaded corresponding to the preset time period, To reduce the video bit rate or video frame rate corresponding to the video to be downloaded.
  • the stuck processing module is set as:
  • the stuck processing module is set as:
  • the video frame extraction is performed on the video to be downloaded corresponding to the preset time period, and the extracted target video to be downloaded is obtained.
  • the stuck processing module is set as:
  • the video to be downloaded corresponding to the preset time period is transcoded to obtain the processed target video to be downloaded; wherein, the video code rate of the target video coding method is less than the video code rate of the current video coding method; Or, the video frame rate of the target video encoding mode is lower than the video frame rate of the current video encoding mode.
  • the freeze prediction result also includes: the freeze prediction duration when freeze occurs within the preset time period after the current moment;
  • the freeze processing module is configured to: perform video processing on the video to be downloaded corresponding to the freeze prediction duration based on the preset freeze processing method.
  • the current playing information includes: current network information of the client, currently cached video length, and currently downloading video information.
  • the current network information includes: packet loss rate, current bandwidth, maximum bandwidth, maximum network delay, minimum network delay, current round-trip delay, minimum round-trip delay, and maximum round-trip delay;
  • the video information currently being downloaded includes: video frame rate, video code rate and video data size.
  • the video freeze prediction device provided in the embodiments of the present disclosure can execute the video freeze prediction method provided in any embodiment of the present disclosure, and has corresponding functional modules and effects for executing the video freeze prediction method.
  • the multiple units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be realized; in addition, multiple functions
  • the names of the units are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the present disclosure.
  • FIG. 7 shows a schematic structural diagram of an electronic device 900 suitable for implementing the embodiments of the present disclosure.
  • the electronic device shown in FIG. 7 is only an example, and should not limit the functions and application scope of the embodiments of the present disclosure.
  • an electronic device 900 may include a processing device (such as a central processing unit, a graphics processing unit, etc.)
  • the storage device 908 loads programs in the random access memory (Random Access Memory, RAM) 903 to execute various appropriate actions and processes.
  • RAM Random Access Memory
  • various programs and data necessary for the operation of the electronic device 900 are also stored.
  • the processing device 901, ROM 902, and RAM 903 are connected to each other through a bus 904.
  • An input/output (Input/Output, I/O) interface 905 is also connected to the bus 904 .
  • the following devices can be connected to the I/O interface 905: an input device 906 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; including, for example, a liquid crystal display (Liquid Crystal Display, LCD), a speaker , an output device 907 such as a vibrator; a storage device 908 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 909.
  • the communication means 909 may allow the electronic device 900 to perform wireless or wired communication with other devices to exchange data.
  • FIG. 7 shows electronic device 900 having various means, it is not a requirement to implement or possess all of the means shown. More or fewer means may alternatively be implemented or provided.
  • embodiments of the present disclosure include a computer program product, which includes a computer program carried on a non-transitory computer readable medium, where the computer program includes program code for executing the method shown in the flowchart.
  • the computer program may be downloaded and installed from a network via communication means 909, or from storage means 908, or from ROM 902.
  • the processing device 901 When the computer program is executed by the processing device 901, the above-mentioned functions defined in the methods of the embodiments of the present disclosure are performed.
  • the electronic device provided by the embodiment of the present disclosure belongs to the same idea as the video freezing prediction method provided by the above-mentioned embodiment.
  • the technical details not described in detail in the embodiment of the present disclosure please refer to the above-mentioned embodiment, and the embodiment of the present disclosure and the above-mentioned embodiment has the same effect.
  • An embodiment of the present disclosure provides a computer storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the video freeze prediction method provided in the foregoing embodiments is implemented.
  • the computer-readable medium mentioned above in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the above two.
  • a computer-readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof.
  • the computer readable storage medium may include: an electrical connection with one or more wires, a portable computer disk, a hard disk, RAM, ROM, Erasable Programmable Read-Only Memory (EPROM), flash memory, optical fiber , portable compact disk read-only memory (Compact Disc Read-Only Memory, CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.
  • a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.
  • a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave carrying computer-readable program code therein. Such propagated data signals may take many forms, including electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • a computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can transmit, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device .
  • the program code contained on the computer readable medium can be transmitted by any appropriate medium, including: electric wire, optical cable, radio frequency (Radio Frequency, RF), etc., or any appropriate combination of the above.
  • the storage medium may be a non-transitory storage medium.
  • the client and the server can communicate using any currently known or future network protocols such as Hypertext Transfer Protocol (HyperText Transfer Protocol, HTTP), and can communicate with digital data in any form or medium
  • the communication eg, communication network
  • Examples of communication networks include local area networks (Local Area Network, LAN), wide area networks (Wide Area Network, WAN), internetworks (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently existing networks that are known or developed in the future.
  • the above-mentioned computer-readable medium may be included in the above-mentioned server; or it may exist independently without being incorporated into the server.
  • the above-mentioned computer-readable medium carries one or more programs, and when the above-mentioned one or more programs are executed by the server, the server: obtains the current playback information when playing the currently downloaded video; obtains the current service information of the server; According to the preset freeze prediction model, the current playing information and the current service information, determine the freeze prediction result when playing the currently to-be-downloaded video.
  • Computer program code for carrying out the operations of the present disclosure can be written in one or more programming languages, or combinations thereof, including object-oriented programming languages—such as Java, Smalltalk, C++, and conventional Procedural Programming Language - such as "C" or a similar programming language.
  • the program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server.
  • the remote computer can be connected to the user computer through any kind of network, including a LAN or WAN, or it can be connected to an external computer (eg via the Internet using an Internet Service Provider).
  • each block in a flowchart or block diagram may represent a module, program segment, or portion of code that contains one or more logical functions for implementing specified executable instructions.
  • the functions noted in the block may occur out of the order noted in the figures.
  • two blocks shown in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved.
  • Each block in the block diagrams and/or flowcharts, and combinations of blocks in the block diagrams and/or flowcharts can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by dedicated hardware implemented in combination with computer instructions.
  • the units involved in the embodiments described in the present disclosure may be implemented by software or by hardware. Wherein, the name of the unit does not constitute a limitation on the unit itself in one case, for example, the editable content display unit may also be described as an "editing unit".
  • FPGA Field Programmable Gate Array
  • ASIC Application Specific Integrated Circuit
  • ASSP Application Specific Standard Parts
  • SOC System on Chip
  • Complex Programmable Logic Device Complex Programmable Logic Device, CPLD
  • a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device.
  • a machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium.
  • a machine-readable medium may comprise an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing.
  • Machine-readable storage media include one or more wire-based electrical connections, portable computer discs, hard drives, RAM, ROM, EPROM, flash memory, optical fiber, portable CD-ROMs, optical storage devices, magnetic storage devices, or Any suitable combination of content.
  • Example 1 provides a video freezing prediction method, including:
  • the current playing information and the current service information determine the freeze prediction result when playing the currently to-be-downloaded video.
  • Example 2 provides a video freezing prediction method, which is applied to a server, and further includes:
  • the current service information includes: the current load of the server, CDN vendor information, and downloadable video volume information cached by the CDN.
  • Example 3 provides a video freezing prediction method, which is applied to a server, and further includes:
  • the preset freeze prediction model includes: an output layer and at least one ResNet residual network;
  • Each ResNet includes: the first fully connected layer, the first activation layer, the second fully connected layer and the second activation layer;
  • the output of the first fully connected layer is used as the input of the first activation layer; the output of the first activated layer is used as the input of the second fully connected layer, and the output of the second fully connected layer is connected with the The input of the first fully-connected layer is skip-connected, and the addition result after the skip-connection is used as the input of the second activation layer.
  • Example 4 provides a video freezing prediction method, which is applied to a server, and further includes:
  • the current playback information and the current service information includes:
  • Preprocessing the current playback information to obtain target playback information and preprocessing the current service information to obtain target service information;
  • the target playback information and the target service information are input into a preset freeze prediction model for freeze prediction, and a freeze prediction result when playing the currently to-be-downloaded video is obtained.
  • Example 5 provides a video freezing prediction method, which is applied to a server, and further includes:
  • the preprocessing of the current playback information to obtain target playback information includes:
  • the preset standardized model regularly updates parameters based on the playback information acquired within the latest preset duration.
  • Example 6 provides a video freezing prediction method, which is applied to a server, and further includes:
  • the method also includes:
  • freeze prediction result is: freeze occurs within the preset time period after the current moment
  • video processing is performed on the video to be downloaded corresponding to the preset time period to reduce the Describe the video bit rate or video frame rate corresponding to the video to be downloaded.
  • Example 7 provides a video freeze prediction method, which is applied to a server, and further includes:
  • performing video processing on the video to be downloaded corresponding to the preset time period based on the preset freeze processing method includes:
  • Example 8 provides a video freezing prediction method, which is applied to a server, and further includes:
  • performing video processing on the video to be downloaded corresponding to the preset time period based on the preset freeze processing method includes:
  • Example 9 provides a video freezing prediction method, which is applied to a server, and further includes:
  • performing video processing on the video to be downloaded corresponding to the preset time period based on the preset freeze processing method includes:
  • the video bit rate of the target video coding mode is lower than the video bit rate of the current video coding mode; or, the video frame rate of the target video coding mode is lower than the video frame rate of the current video coding mode.
  • Example 10 provides a video freezing prediction method, which is applied to a server, and further includes:
  • the freeze prediction result further includes: the freeze prediction duration when freeze occurs within a preset time period after the current moment;
  • the video processing of the video to be downloaded corresponding to the preset time period based on the preset freeze processing method includes:
  • video processing is performed on the video to be downloaded corresponding to the freeze prediction duration.
  • Example 11 provides a video freezing prediction method, which further includes:
  • the current playing information includes: current network information of the client, currently cached video length, and currently downloading video information.
  • Example 12 provides a video freezing prediction method, which further includes:
  • the current network information of the client includes: packet loss rate, current bandwidth, maximum bandwidth, maximum network delay, minimum network delay, current round-trip delay, minimum round-trip delay, and maximum round-trip delay;
  • the video information currently being downloaded includes: video frame rate, video code rate and video data size.
  • Example 13 provides a video freezing prediction device, including:
  • the current playback information acquisition module is configured to obtain the current playback information when playing the currently downloaded video
  • the current service information acquisition module is configured to obtain the current service information of the server
  • the freeze prediction module is configured to determine the freeze prediction result when playing the current video to be downloaded according to the preset freeze prediction model, the current playing information and the current service information.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Databases & Information Systems (AREA)
  • Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)

Abstract

本公开提供了一种视频卡顿预测方法、装置、设备和介质。该视频卡顿预测方法包括：获取在播放当前已下载视频时的当前播放信息；获取服务器的当前服务信息；根据预设卡顿预测模型、当前播放信息和当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。

Description

视频卡顿预测方法、装置、设备和介质
本申请要求在2021年12月20日提交中国专利局、申请号为202111564883.5的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本公开涉及互联网技术领域，例如涉及一种视频卡顿预测方法、装置、设备和介质。
背景技术
随着互联网技术的快速发展，越来越多的用户喜欢在客户端上观看直播视频或者点播视频。在视频播放过程中，往往由于视频下载速度较慢而导致出现视频卡顿的情况，严重影响了用户的观看体验。可以在视频播放后，基于观看视频过程中是否出现卡顿的结果对视频进行处理，提高视频的传输速度和稳定性。可见，这种方式只能在视频实际播放后才能获知视频卡顿结果，无法提前获知视频卡顿结果，从而无法有效保证视频播放的流畅度，降低用户观看体验。
发明内容
本公开提供了一种视频卡顿预测方法、装置、设备和介质，以对当前待下载视频进行视频卡顿预测，从而可以提前获知视频卡顿结果，进而有效保证视频播放的流畅度，提升用户观看体验。
本公开实施例提供了一种视频卡顿预测方法，包括：
获取在播放当前已下载视频时的当前播放信息；
获取服务器的当前服务信息；
根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
本公开实施例还提供了一种视频卡顿预测装置，包括：
当前播放信息获取模块，设置为获取在播放当前已下载视频时的当前播放信息；
当前服务信息获取模块，设置为获取服务器的当前服务信息；
卡顿预测模块，设置为根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
本公开实施例还提供了一种电子设备，包括：
一个或多个处理器；
存储器，设置为存储一个或多个程序；
当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如本公开任意实施例所提供的视频卡顿预测方法。
本公开实施例还提供了一种计算机可读存储介质，存储有计算机程序，该计算机程序被处理器执行时实现如本公开任意实施例所提供的视频卡顿预测方法。
附图说明
图1是本公开实施例一提供的一种视频卡顿预测方法的流程图；
图2是本公开实施例一所涉及的一种视频卡顿预测过程的示例；
图3是本公开实施例一所涉及的一种预设卡顿预测模型的结构示例；
图4是本公开实施例二提供的一种视频卡顿预测方法的流程图；
图5是本公开实施例二所涉及的一种视频卡顿预测过程的示例；
图6是本公开实施例三提供的一种视频卡顿预测装置的结构示意图；
图7是本公开实施例四提供的一种电子设备的结构示意图。
具体实施方式
下面将参照附图描述本公开的实施例。虽然附图中显示了本公开的一些实施例，然而本公开可以通过多种形式来实现，而且不应该被解释为限于这里阐述的实施例，相反提供这些实施例是为了理解本公开。本公开的附图及实施例仅用于示例性作用，并非用于限制本公开的保护范围。
本公开的方法实施方式中记载的多个步骤可以按照不同的顺序执行，和/或并行执行。此外，方法实施方式可以包括附加的步骤和/或省略执行示出的步骤。本公开的范围在此方面不受限制。
本文使用的术语“包括”及其变形是开放性包括，即“包括但不限于”。术语“基于”是“至少部分地基于”。术语“一个实施例”表示“至少一个实施例”；术语“另一实施例”表示“至少一个另外的实施例”；术语“一些实施例”表示“至少一些实施例”。其他术语的相关定义将在下文描述中给出。
本公开中提及的“第一”、“第二”等概念仅用于对不同的装置、模块或单元进 行区分，并非用于限定这些装置、模块或单元所执行的功能的顺序或者相互依存关系。
本公开中提及的“一个”、“多个”的修饰是示意性而非限制性的，除非在上下文另有指出，否则应该理解为“一个或多个”。
本公开实施方式中的多个装置之间所交互的消息或者信息的名称仅用于说明性的目的，而并不是用于对这些消息或信息的范围进行限制。
实施例一
图1为本公开实施例一提供的一种视频卡顿预测方法的流程图，本实施例可适用于对在播放当前待下载视频时是否出现播放卡顿进行提前预测的情况，例如可以用于对直播视频或者点播视频进行卡顿预测的应用场景中。该方法可以由视频卡顿预测装置来执行，该装置可以由软件和/或硬件的方式来实现，集成于客户端或者服务器中。如图1所示，该方法包括以下步骤。
S110、获取在播放当前已下载视频时的当前播放信息。
当前已下载视频可以是指客户端当前正在播放的已下载视频。在直播场景或者点播场景中，可以将直播视频或者点播视频按时间进行分片处理，获得直播视频或者点播视频中的多个视频片段，使得客户端可以通过依次下载并播放多个视频片段的方式进行直播视频或点播视频的播放。例如，当前已下载视频可以是指客户端当前正在播放的已下载直播视频片段或者已下载点播视频片段。
当前播放信息可以包括：客户端的当前网络信息、当前已缓存的视频长度和当前正在下载的视频信息。示例性地，当前网络信息可以包括：丢包率、当前带宽、最大带宽、最大网络延迟、最小网络延迟、当前往返时延、最小往返时延和最大往返时延；当前正在下载的视频信息可以包括：视频帧率、视频码率和视频数据大小。
客户端在播放当前已下载视频时，可以实时获取当前已下载视频对应的当前播放信息，以便进行视频卡顿的实时预测；也可以基于当前已缓存的视频长度设置相应的时间间隔，并基于该时间间隔定时获取当前已下载视频对应的当前播放信息，以便在保证视频播放流畅度的基础上降低卡顿预测次数，从而节省设备资源。若该方法应用于客户端，则客户端可以直接获得当前播放信息。若该方法应用于服务器，则客户端可以将获得的当前播放信息发送至服务器，以使服务器可以接收到客户端发送的当前播放信息。通过在服务器上基于客户端发送的当前播放信息进行视频卡顿预测操作，可以充分利用服务器资源，快速准确地对当前待下载视频进行视频卡顿预测，提高了预测准确性和预测效率。
S120、获取服务器的当前服务信息。
当前服务信息可以是指当前服务器自身的可以影响视频下载速度的信息。例如，当前服务信息可以包括：服务器的当前负载量、内容分发网络(Content Delivery Network，CDN)厂商信息和CDN缓存的可下载视频量信息。其中，若CDN厂商信息与用户网络服务提供商信息相匹配，则会提高视频下载速度，否则会降低视频下载速度。CDN缓存的可下载视频量信息可以是指服务器的CDN节点中当前缓存的可下载直播视频片段长度或者可下载点播视频片段长度。
若该方法应用于客户端，则服务器可以将当前采集的当前服务信息发送至客户端，以使客户端可以接收到服务器发送的当前服务信息。若该方法应用于服务器，则服务器可以直接获取当前服务信息，以便基于当前服务信息进行卡顿预测。
S130、根据预设卡顿预测模型、当前播放信息和当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
当前待下载视频可以是指客户端当前等待下载的视频。例如，当前待下载视频可以是指客户端当前等待下载的直播视频片段或者点播视频片段。卡顿预测结果可以包括：当前时刻后的预设时间段内是否出现卡顿或者出现卡顿的概率。预设时间段可以是预先基于业务需求设置的一个时间段，该预设时间段可以包括从当前时刻到待下载视频对应的播放时刻之间的时间段。卡顿预测结果还可以包括：卡顿预测时长，也就是在预设时间段内预测会出现卡顿的时间长度，以便满足不同的业务需求和场景。预设卡顿预测模型可以是预先设置的，用于预测后续播放当前待下载视频时是否出现视频卡顿的神经网络模型。本实施例中的预设卡顿预测模型可以是预先基于样本数据训练获得的。
可以将当前播放信息和当前服务信息输入至预先训练好的预设卡顿预测模型中进行视频卡顿的预测，并基于预设卡顿预测模型的输出，获得客户端在播放当前待下载视频时的卡顿预测结果。预设卡顿预测模型可以同时基于当前播放信息和当前服务信息进行卡顿预测，从而可以提高卡顿预测的准确性。
示例性地，S130可以包括：对当前播放信息进行预处理，获得目标播放信息，并对当前服务信息进行预处理，获得目标服务信息；将目标播放信息和目标服务信息输入至预设卡顿预测模型中进行卡顿预测，获得在播放当前待下载视频时的卡顿预测结果。
通过对当前播放信息和当前服务信息进行预处理，比如对当前播放信息和当前服务信息进行归一化、以及过滤无效信息等，从而可以获得统一的目标播放信息和目标服务信息，进而保证输入预设卡顿预测模型的信息统一性，提高模型预测的准确性。
示例性地，对当前播放信息进行预处理，获得目标播放信息，可以包括：基于预设标准化模型，对当前播放信息进行数据标准化处理，获得目标播放信息。对当前服务信息进行预处理，获得目标服务信息，可以包括：基于预设标准化模型，对当前服务信息进行数据标准化处理，获得目标服务信息。
图2给出了一种视频卡顿预测过程的示例。如图2所示，可以利用预设标准化模型快速地对当前播放信息以及当前服务信息进行数据标准化处理，以便提高处理效率，并且保证输入信息的统一性。
预设标准化模型定时基于最新预设时长内获取到的播放信息和服务信息进行参数更新。如图2所示，该方法应用于服务器中，服务器可以将客户端每次上报的当前播放信息和每次采集的当前服务信息存储至数据库中，从而可以利用当前最新存储的播放信息和服务信息，对预设标准化模型进行参数更新。例如，可以利用最新一日存储的播放信息和服务信息，对预设标准化模型中用于标准化处理每种信息的平均值和方差进行更新，从而可以提高卡顿预测的准确性。
示例性地，本实施例可以定期收集一部分用户(即未进行卡顿预测的用户)的客户端上报的当前播放信息和采集服务器的当前服务信息，并将当前播放信息和当前服务信息存储至数据库中。这部分用户可以将实际播放待下载视频时是否出现卡顿的情况，即实际卡顿结果上报至服务器，服务器将实际卡顿结果存储至数据库中。如图2所示，本实施例可以基于最新预设时长内存储的播放信息和服务信息以及相应的实际卡顿结果，对预设卡顿预测模型进行权重更新，例如，可以利用最新一日或多日的播放信息和服务信息以及相应的实际卡顿结果对预设卡顿预测模型进行新一轮训练，以便提高卡顿预测的准确性。
本公开实施例的技术方案，通过获取在播放当前已下载视频时的当前播放信息以及服务器自身的当前服务信息，并根据预设卡顿预测模型、当前播放信息和当前服务信息，确定出在播放当前待下载视频时的卡顿预测结果，从而可以提前获知视频卡顿结果，进而有效保证视频播放的流畅度，提升用户观看体验。
在上述技术方案的基础上，预设卡顿预测模型可以包括：输出层和至少一个残差网络(Residual Network，ResNet)；每个ResNet残差网络包括：第一全连接层、第一激活层、第二全连接层和第二激活层；其中，第一全连接层的输出作为第一激活层的输入；第一激活层的输出作为第二全连接层的输入，第二全连接层的输出与第一全连接层的输入进行跳跃连接，跳跃连接后的相加结果作为第二激活层的输入。
第一激活层和第二激活层所使用的激活函数可以为修正线性单元(Rectified  Linear Unit，ReLU)激活函数。输出层所使用的激活函数可以为Sigmoid函数。输出层输出的信息可以是出现卡顿的概率，当概率小于0.5时表示预测不会出现卡顿，当概率大于0.5时表示预测会出现卡顿。
图3给出了一种预设卡顿预测模型的结构示例。如图3所示，在预设卡顿预测模型包括两个ResNet残差网络时的卡顿预测的精确率最高，从而可以选择由两个ResNet残差网络组成的预设卡顿预测模型。示例性地，图3中的预设卡顿预测模型的预测过程为：将目标播放信息和目标服务信息输入至第一个ResNet残差网络中的第一全连接层，获得第一全连接信息，将第一全连接信息输入至第一激活层，获得第一激活信息，将第一激活信息输入至第二全连接层，获得第二全连接信息，并将第二全连接信息与输入的目标播放信息和目标服务信息进行相加，将获得的相加结果输入至第二激活层中，获得第二激活信息。同理，将第一个ResNet残差网络输出的第二激活信息输入至第二个ResNet残差网络中进行类似处理，获得第二个ResNet残差网络输出的信息，并将该信息输入至输出层中进行映射处理，获得最终输出的卡顿预测结果。
实施例二
图4为本公开实施例二提供的一种视频卡顿预测方法的流程图，本实施例在上述实施例一的基础上，对预测出现卡顿的待下载视频进行视频处理的过程进行描述。其中与上述多个实施例相同或相应的术语的解释在此不再赘述。
参见图4，本实施例提供的视频卡顿预测方法包括以下步骤。
S410、获取在播放当前已下载视频时的当前播放信息。
S420、获取服务器的当前服务信息。
S430、根据预设卡顿预测模型、当前播放信息和当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
S440、若卡顿预测结果是：当前时刻后的预设时间段内出现卡顿，则基于预设卡顿处理方式，对预设时间段对应的待下载视频进行视频处理，以降低待下载视频对应的视频码率或者视频帧率。
预设卡顿处理方式可以是预先设置的，用于降低服务器到客户端之间传输的视频数据量大小的方式，以避免不必要的视频卡顿，保证视频播放流畅度。预设卡顿处理方式可以包括多种方式。例如，预设卡顿处理方式可以是指降低视频码率的方式或者降低视频帧率的方式。
图5给出了一种视频卡顿预测过程的示例。如图5所示，服务器中的卡顿 避免策略模型可以基于预设卡顿预测模型输出的卡顿预测结果进行相应的决策。例如，在卡顿预测结果是当前时刻后的预设时间段内出现卡顿时，可以确定出降低传输视频码率或者降低传输视频帧率的卡顿处理方式，使得直播/点播服务器基于该卡顿处理方式，对在预设时间段内需要下载的待下载视频进行视频处理，从而降低服务器到客户端之间传输的视频数据量大小，避免了不必要的视频卡顿，保证了视频播放流畅度，提升了用户体验。
示例性地，对预测出现卡顿的待下载视频的卡顿处理过程可以通过如下三种方式实现：
作为第一种实现方式，S440中的“基于预设卡顿处理方式，对预设时间段对应的待下载视频进行视频处理”可以包括：基于当前正在下载视频对应的第一视频码率，确定小于第一视频码率的第二视频码率；基于预先存储的每个视频码率下预设时间段对应的待下载视频，确定出第二视频码率下预设时间段对应的目标待下载视频。
在点播场景中，可以针对点播视频中的每个点播视频片段，预先生成并存储有具有不同视频码率的每个待下载视频。在对待下载视频进行卡顿处理时，可以从现有的视频码率中选择一个小于当前下载的第一视频码率的视频码率作为第二视频码率，并从预先存储的每个待下载视频中，可以快速地获得第二视频码率下预设时间段对应的目标待下载视频，以使客户端下载目标待下载视频，从而提高卡顿处理效率，并且通过降低视频码率的方式降低了下载视频数据量，使得目标待下载视频可以更加快速地进行下载，避免了后续播放出现视频卡顿的情况。
作为第二种实现方式，S440中的“基于预设卡顿处理方式，对预设时间段对应的待下载视频进行视频处理”可以包括：对预设时间段对应的待下载视频进行视频帧抽取，获得抽取后的目标待下载视频。
在直播场景或点播场景中，可以通过对预设时间段对应的每个待下载视频进行视频帧抽取的方式，降低目标待下载视频的视频帧率，从而降低下载视频数据量，避免了后续出现视频卡顿的情况。
作为第三种实现方式，S440中的“基于预设卡顿处理方式，对预设时间段对应的待下载视频进行视频处理”可以包括：基于目标视频编码方式，对预设时间段对应的待下载视频进行转码处理，获得处理后的目标待下载视频；其中，目标视频编码方式的视频码率小于当前视频编码方式的视频码率；或者，目标视频编码方式的视频帧率小于当前视频编码方式的视频帧率。
在直播场景中，需要对实时获得的视频片段进行实时转码处理，以保证视 频流畅度。本实施例可以利用视频码率较低的和/或视频帧率较低的目标视频编码方式对待下载视频进行转码处理，从而降低处理后的目标待下载视频的视频码率和/或视频帧率，进而降低传输视频数据量，避免了后续出现视频卡顿的情况。
本实施例的技术方案，通过在卡顿预测结果是当前时刻后的预设时间段内出现卡顿时，可以基于预设卡顿处理方式，对预设时间段对应的待下载视频进行视频处理，通过降低待下载视频对应的视频码率或者视频帧率的方式降低服务器到客户端之间传输的视频数据量大小，避免了不必要的视频卡顿，保证视频播放流畅度，提升了用户观看体验。
在上述技术方案的基础上，卡顿预测结果还包括：当前时刻后的预设时间段内出现卡顿时的卡顿预测时长。相应地，步骤S440中的“基于预设卡顿处理方式，对预设时间段对应的待下载视频进行视频处理”可以包括：基于预设卡顿处理方式，对卡顿预测时长对应的待下载视频进行视频处理。
对卡顿预测时长对应的待下载视频的卡顿处理过程可以参见上述三种实现方式，此处不再赘述。
卡顿预测结果在预测会出现卡顿时还可以预测出卡顿预测结果对应的卡顿预测时长，从而只需对在卡顿预测时长内需要下载的待下载视频进行视频处理即可，即降低卡顿预测时长对应的待下载视频的视频码率或者视频帧率，无需对其他的待下载视频进行处理，从而可以保证视频观看质量，提升用户体验。
以下是本公开实施例提供的视频卡顿预测装置的实施例，该装置与上述实施例的视频卡顿预测方法属于同一个构思，在视频卡顿预测装置的实施例中未详尽描述的细节内容，可以参考上述实施例。
实施例三
图6为本公开实施例三提供的一种视频卡顿预测装置的结构示意图，本实施例可适用于对在播放当前待下载视频时是否出现播放卡顿进行提前预测的情况，例如可以用于对直播视频或者点播视频进行卡顿预测的应用场景中。如图6所示，该装置包括：当前播放信息获取模块610、当前服务信息获取模块620和卡顿预测模块630。
当前播放信息获取模块610，设置为获取在播放当前已下载视频时的当前播放信息；当前服务信息获取模块620，设置为获取服务器的当前服务信息；卡顿预测模块730，设置为根据预设卡顿预测模型、当前播放信息和当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
本公开实施例的技术方案，通过获取在播放当前已下载视频时的当前播放信息以及服务器自身的当前服务信息，并根据预设卡顿预测模型、当前播放信息和当前服务信息，确定出在播放当前待下载视频时的卡顿预测结果，从而可以提前获知视频卡顿结果，进而有效保证视频播放流畅度，提升用户观看体验。
在上述技术方案的基础上，当前服务信息包括：服务器的当前负载量、CDN厂商信息和CDN缓存的可下载视频量信息。
在上述多个技术方案的基础上，预设卡顿预测模型包括：输出层和至少一个ResNet；每个ResNet包括：第一全连接层、第一激活层、第二全连接层和第二激活层；
其中，第一全连接层的输出作为第一激活层的输入；第一激活层的输出作为第二全连接层的输入，第二全连接层的输出与第一全连接层的输入进行跳跃连接，跳跃连接后的相加结果作为第二激活层的输入。
在上述多个技术方案的基础上，卡顿预测模块630可以包括：
预处理单元，设置为对当前播放信息进行预处理，获得目标播放信息，并对当前服务信息进行预处理，获得目标服务信息；
卡顿预测单元，设置为将目标播放信息和目标服务信息输入至预设卡顿预测模型中进行卡顿预测，获得在播放当前待下载视频时的卡顿预测结果。
在上述多个技术方案的基础上，预处理单元，是设置为：基于预设标准化模型，对当前播放信息进行数据标准化处理，获得目标播放信息；其中，预设标准化模型定时基于最新预设时长内获取到的播放信息进行参数更新。
在上述多个技术方案的基础上，该装置还包括：
卡顿处理模块，设置为若卡顿预测结果是：当前时刻后的预设时间段内出现卡顿，则基于预设卡顿处理方式，对预设时间段对应的待下载视频进行视频处理，以降低待下载视频对应的视频码率或者视频帧率。
在上述多个技术方案的基础上，卡顿处理模块，是设置为：
基于当前正在下载视频对应的第一视频码率，确定小于第一视频码率的第二视频码率；基于预先存储的每个视频码率下预设时间段对应的待下载视频，确定出第二视频码率下预设时间段对应的目标待下载视频。
在上述多个技术方案的基础上，卡顿处理模块，是设置为：
对预设时间段对应的待下载视频进行视频帧抽取，获得抽取后的目标待下载视频。
在上述多个技术方案的基础上，卡顿处理模块，是设置为：
基于目标视频编码方式，对预设时间段对应的待下载视频进行转码处理，获得处理后的目标待下载视频；其中，目标视频编码方式的视频码率小于当前视频编码方式的视频码率；或者，目标视频编码方式的视频帧率小于当前视频编码方式的视频帧率。
在上述多个技术方案的基础上，卡顿预测结果还包括：当前时刻后的预设时间段内出现卡顿时的卡顿预测时长；
卡顿处理模块，是设置为：基于预设卡顿处理方式，对卡顿预测时长对应的待下载视频进行视频处理。
在上述多个技术方案的基础上，所述当前播放信息包括：客户端的当前网络信息、当前已缓存的视频长度和当前正在下载的视频信息。
在上述多个技术方案的基础上，所述当前网络信息包括：丢包率、当前带宽、最大带宽、最大网络延迟、最小网络延迟、当前往返时延、最小往返时延和最大往返时延；
所述当前正在下载的视频信息包括：视频帧率、视频码率和视频数据大小。
本公开实施例所提供的视频卡顿预测装置可执行本公开任意实施例所提供的视频卡顿预测方法，具备执行视频卡顿预测方法相应的功能模块和效果。
上述视频卡顿预测装置的实施例中，所包括的多个单元和模块只是按照功能逻辑进行划分的，但并不局限于上述的划分，只要能够实现相应的功能即可；另外，多个功能单元的名称也只是为了便于相互区分，并不用于限制本公开的保护范围。
实施例四
下面参考图7，其示出了适于用来实现本公开实施例的电子设备900的结构示意图。图7示出的电子设备仅仅是一个示例，不应对本公开实施例的功能和使用范围带来任何限制。
如图7所示，电子设备900可以包括处理装置(例如中央处理器、图形处理器等)901，处理装置901可以根据存储在只读存储器(Read-only Memory，ROM)902中的程序或者从存储装置908加载到随机访问存储器(Random Access Memory，RAM)903中的程序而执行多种适当的动作和处理。在RAM 903中，还存储有电子设备900操作所需的多种程序和数据。处理装置901、ROM 902以及RAM 903通过总线904彼此相连。输入/输出(Input/Output，I/O)接口905也连接至总线904。
以下装置可以连接至I/O接口905：包括例如触摸屏、触摸板、键盘、鼠标、摄像头、麦克风、加速度计、陀螺仪等的输入装置906；包括例如液晶显示器(Liquid Crystal Display，LCD)、扬声器、振动器等的输出装置907；包括例如磁带、硬盘等的存储装置908；以及通信装置909。通信装置909可以允许电子设备900与其他设备进行无线或有线通信以交换数据。虽然图7示出了具有多种装置的电子设备900，但是并不要求实施或具备所有示出的装置。可以替代地实施或具备更多或更少的装置。
根据本公开的实施例，上文参考流程图描述的过程可以被实现为计算机软件程序。例如，本公开的实施例包括一种计算机程序产品，其包括承载在非暂态计算机可读介质上的计算机程序，该计算机程序包含用于执行流程图所示的方法的程序代码。在这样的实施例中，该计算机程序可以通过通信装置909从网络上被下载和安装，或者从存储装置908被安装，或者从ROM 902被安装。在该计算机程序被处理装置901执行时，执行本公开实施例的方法中限定的上述功能。
本公开实施例提供的电子设备与上述实施例提供的视频卡顿预测方法属于同一构思，未在本公开实施例中详尽描述的技术细节可参见上述实施例，并且本公开实施例与上述实施例具有相同的效果。
实施例五
本公开实施例提供了一种计算机存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现上述实施例所提供的视频卡顿预测方法。
本公开上述的计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质或者是上述两者的任意组合。计算机可读存储介质例如可以是电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。计算机可读存储介质可以包括：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、RAM、ROM、可擦式可编程只读存储器(Erasable Programmable Read-Only Memory，EPROM)、闪存、光纤、便携式紧凑磁盘只读存储器(Compact Disc Read-Only Memory，CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本公开中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。而在本公开中，计算机可读信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括电磁信号、光信号或上述的任意合适的组合。计算机可读信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该 计算机可读信号介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括：电线、光缆、射频(Radio Frequency，RF)等等，或者上述的任意合适的组合。存储介质可以是非暂态(non-transitory)存储介质。
在一些实施方式中，客户端、服务器可以利用诸如超文本传输协议(HyperText Transfer Protocol，HTTP)之类的任何当前已知或未来研发的网络协议进行通信，并且可以与任意形式或介质的数字数据通信(例如，通信网络)互连。通信网络的示例包括局域网(Local Area Network，LAN)，广域网(Wide Area Network，WAN)，网际网(例如，互联网)以及端对端网络(例如，ad hoc端对端网络)，以及任何当前已知或未来研发的网络。
上述计算机可读介质可以是上述服务器中所包含的；也可以是单独存在，而未装配入该服务器中。
上述计算机可读介质承载有一个或者多个程序，当上述一个或者多个程序被该服务器执行时，使得该服务器：获取在播放当前已下载视频时的当前播放信息；获取服务器的当前服务信息；根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
可以以一种或多种程序设计语言或其组合来编写用于执行本公开的操作的计算机程序代码，上述程序设计语言包括面向对象的程序设计语言—诸如Java、Smalltalk、C++，还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络——包括LAN或WAN—连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。
附图中的流程图和框图，图示了按照本公开多种实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段、或代码的一部分，该模块、程序段、或代码的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个接连地表示的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或操作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实 现。
描述于本公开实施例中所涉及到的单元可以通过软件的方式实现，也可以通过硬件的方式来实现。其中，单元的名称在一种情况下并不构成对该单元本身的限定，例如，可编辑内容显示单元还可以被描述为“编辑单元”。
本文中以上描述的功能可以至少部分地由一个或多个硬件逻辑部件来执行。例如，可以使用的示范类型的硬件逻辑部件包括：现场可编程门阵列(Field Programmable Gate Array，FPGA)、专用集成电路(Application Specific Integrated Circuit，ASIC)、专用标准产品(Application Specific Standard Parts，ASSP)、片上系统(System on Chip，SOC)、复杂可编程逻辑设备(Complex Programmable Logic Device，CPLD)等等。
在本公开的上下文中，机器可读介质可以是有形的介质，其可以包含或存储以供指令执行系统、装置或设备使用或与指令执行系统、装置或设备结合地使用的程序。机器可读介质可以是机器可读信号介质或机器可读储存介质。机器可读介质可以包括电子的、磁性的、光学的、电磁的、红外的、或半导体系统、装置或设备，或者上述内容的任何合适组合。机器可读存储介质包括基于一个或多个线的电气连接、便携式计算机盘、硬盘、RAM、ROM、EPROM、快闪存储器、光纤、便捷式CD-ROM、光学储存设备、磁储存设备、或上述内容的任何合适组合。
根据本公开的一个或多个实施例，【示例一】提供了一种视频卡顿预测方法，包括：
获取在播放当前已下载视频时的当前播放信息；
获取服务器的当前服务信息；
根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
根据本公开的一个或多个实施例，【示例二】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述当前服务信息包括：所述服务器的当前负载量、CDN厂商信息和CDN缓存的可下载视频量信息。
根据本公开的一个或多个实施例，【示例三】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述预设卡顿预测模型包括：输出层和至少一个ResNet残差网络；
每个ResNet包括：第一全连接层、第一激活层、第二全连接层和第二激活层；
其中，所述第一全连接层的输出作为第一激活层的输入；所述第一激活层的输出作为所述第二全连接层的输入，所述第二全连接层的输出与所述第一全连接层的输入进行跳跃连接，所述跳跃连接后的相加结果作为所述第二激活层的输入。
根据本公开的一个或多个实施例，【示例四】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频的卡顿预测结果，包括：
对所述当前播放信息进行预处理，获得目标播放信息，并对所述当前服务信息进行预处理，获得目标服务信息；
将所述目标播放信息和所述目标服务信息输入至预设卡顿预测模型中进行卡顿预测，获得在播放当前待下载视频时的卡顿预测结果。
根据本公开的一个或多个实施例，【示例五】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述对所述当前播放信息进行预处理，获得目标播放信息，包括：
基于预设标准化模型，对所述当前播放信息进行数据标准化处理，获得目标播放信息；
其中，所述预设标准化模型定时基于最新预设时长内获取到的播放信息进行参数更新。
根据本公开的一个或多个实施例，【示例六】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述方法还包括：
若所述卡顿预测结果是：当前时刻后的预设时间段内出现卡顿，则基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，以降低所述待下载视频对应的视频码率或者视频帧率。
根据本公开的一个或多个实施例，【示例七】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
基于当前正在下载视频对应的第一视频码率，确定小于所述第一视频码率的第二视频码率；
基于预先存储的每个视频码率下所述预设时间段对应的待下载视频，确定出所述第二视频码率下所述预设时间段对应的目标待下载视频。
根据本公开的一个或多个实施例，【示例八】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
对所述预设时间段对应的待下载视频进行视频帧抽取，获得抽取后的目标待下载视频。
根据本公开的一个或多个实施例，【示例九】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
基于目标视频编码方式，对所述预设时间段对应的待下载视频进行转码处理，获得处理后的目标待下载视频；
其中，所述目标视频编码方式的视频码率小于当前视频编码方式的视频码率；或者，所述目标视频编码方式的视频帧率小于当前视频编码方式的视频帧率。
根据本公开的一个或多个实施例，【示例十】提供了一种视频卡顿预测方法，应用于服务器，还包括：
可选的，所述卡顿预测结果还包括：当前时刻后的预设时间段内出现卡顿时的卡顿预测时长；
所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
基于预设卡顿处理方式，对所述卡顿预测时长对应的待下载视频进行视频处理。
根据本公开的一个或多个实施例，【示例十一】提供了一种视频卡顿预测方法，还包括：
可选的，所述当前播放信息包括：客户端的当前网络信息、当前已缓存的视频长度和当前正在下载的视频信息。
根据本公开的一个或多个实施例，【示例十二】提供了一种视频卡顿预测方法，还包括：
可选的，所述客户端的当前网络信息包括：丢包率、当前带宽、最大带宽、最大网络延迟、最小网络延迟、当前往返时延、最小往返时延和最大往返时延；
所述当前正在下载的视频信息包括：视频帧率、视频码率和视频数据大小。
根据本公开的一个或多个实施例，【示例十三】提供了一种视频卡顿预测装置，包括：
当前播放信息获取模块，设置为获取在播放当前已下载视频时的当前播放信息；
当前服务信息获取模块，设置为获取服务器的当前服务信息；
卡顿预测模块，设置为根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频时的卡顿预测结果。
此外，虽然采用特定次序描绘了多个操作，但是这不应当理解为要求这些操作以所示出的特定次序或以顺序次序执行来执行。在一定环境下，多任务和并行处理可能是有利的。同样地，虽然在上面论述中包含了多个实现细节，但是这些不应当被解释为对本公开的范围的限制。在单独的实施例的上下文中描述的一些特征还可以组合地实现在单个实施例中。相反地，在单个实施例的上下文中描述的多种特征也可以单独地或以任何合适的子组合的方式实现在多个实施例中。
尽管已经采用特定于结构特征和/或方法逻辑动作的语言描述了本主题，但是应当理解所附权利要求书中所限定的主题未必局限于上面描述的特定特征或动作。相反，上面所描述的特定特征和动作仅仅是实现权利要求书的示例形式。

Claims (15)

  1. 一种视频卡顿预测方法，包括：
    获取在播放当前已下载视频的情况下的当前播放信息；
    获取服务器的当前服务信息；
    根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频的情况下的卡顿预测结果。
  2. 根据权利要求1所述的方法，其中，所述当前服务信息包括：所述服务器的当前负载量、内容分发网络CDN厂商信息和CDN缓存的可下载视频量信息。
  3. 根据权利要求1所述的方法，其中，所述预设卡顿预测模型包括：输出层和至少一个残差网络ResNet；
    每个ResNet包括：第一全连接层、第一激活层、第二全连接层和第二激活层；
    其中，所述第一全连接层的输出作为所述第一激活层的输入，所述第一激活层的输出作为所述第二全连接层的输入，所述第二全连接层的输出与所述第一全连接层的输入进行跳跃连接，所述跳跃连接后的相加结果作为所述第二激活层的输入。
  4. 根据权利要求1所述的方法，其中，所述根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频的情况下的卡顿预测结果，包括：
    对所述当前播放信息进行预处理，获得目标播放信息，并对所述当前服务信息进行预处理，获得目标服务信息；
    将所述目标播放信息和所述目标服务信息输入至所述预设卡顿预测模型中进行卡顿预测，获得在播放所述当前待下载视频的情况下的卡顿预测结果。
  5. 根据权利要求4所述的方法，其中，所述对所述当前播放信息进行预处理，获得目标播放信息，包括：
    基于预设标准化模型，对所述当前播放信息进行数据标准化处理，获得所述目标播放信息；
    其中，所述预设标准化模型定时基于更新后的预设时长内获取到的播放信息进行参数更新。
  6. 根据权利要求1所述的方法，还包括：
    在所述卡顿预测结果是：当前时刻后的预设时间段内出现卡顿的情况下， 基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，以降低所述待下载视频对应的视频码率或者视频帧率。
  7. 根据权利要求6所述的方法，其中，所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
    基于当前正在下载视频对应的第一视频码率，确定小于所述第一视频码率的第二视频码率；
    基于预先存储的每个视频码率下所述预设时间段对应的待下载视频，确定出所述第二视频码率下所述预设时间段对应的目标待下载视频。
  8. 根据权利要求6所述的方法，其中，所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
    对所述预设时间段对应的待下载视频进行视频帧抽取，获得抽取后的目标待下载视频。
  9. 根据权利要求6所述的方法，其中，所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
    基于目标视频编码方式，对所述预设时间段对应的待下载视频进行转码处理，获得处理后的目标待下载视频；
    其中，所述目标视频编码方式的视频码率小于当前视频编码方式的视频码率，或者，所述目标视频编码方式的视频帧率小于当前视频编码方式的视频帧率。
  10. 根据权利要求6所述的方法，其中，所述卡顿预测结果还包括：在所述当前时刻后的预设时间段内出现卡顿的情况下的卡顿预测时长；
    所述基于预设卡顿处理方式，对所述预设时间段对应的待下载视频进行视频处理，包括：
    基于所述预设卡顿处理方式，对所述卡顿预测时长对应的待下载视频进行视频处理。
  11. 根据权利要求1所述的方法，其中，所述当前播放信息包括：客户端的当前网络信息、当前已缓存的视频长度和当前正在下载的视频信息。
  12. 根据权利要求11所述的方法，其中，所述客户端的当前网络信息包括：丢包率、当前带宽、最大带宽、最大网络延迟、最小网络延迟、当前往返时延、最小往返时延和最大往返时延；
    所述当前正在下载的视频信息包括：视频帧率、视频码率和视频数据大小。
  13. 一种视频卡顿预测装置，包括：
    当前播放信息获取模块，设置为获取在播放当前已下载视频的情况下的当前播放信息；
    当前服务信息获取模块，设置为获取服务器的当前服务信息；
    卡顿预测模块，设置为根据预设卡顿预测模型、所述当前播放信息和所述当前服务信息，确定在播放当前待下载视频的情况下的卡顿预测结果。
  14. 一种电子设备，包括：
    至少一个处理器；
    存储器，设置为存储至少一个程序；
    当所述至少一个程序被所述至少一个处理器执行，使得所述至少一个处理器实现如权利要求1-12中任一项所述的视频卡顿预测方法。
  15. 一种计算机可读存储介质，存储有计算机程序，其中，所述计算机程序被处理器执行时实现如权利要求1-12中任一项所述的视频卡顿预测方法。
PCT/CN2022/129721 2021-12-20 2022-11-04 视频卡顿预测方法、装置、设备和介质 WO2023116233A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111564883.5 2021-12-20
CN202111564883.5A CN114401447A (zh) 2021-12-20 2021-12-20 一种视频卡顿预测方法、装置、设备和介质

Publications (1)

Publication Number Publication Date
WO2023116233A1 true WO2023116233A1 (zh) 2023-06-29

Family

ID=81227534

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/129721 WO2023116233A1 (zh) 2021-12-20 2022-11-04 视频卡顿预测方法、装置、设备和介质

Country Status (2)

Country Link
CN (1) CN114401447A (zh)
WO (1) WO2023116233A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117241071A (zh) * 2023-11-15 2023-12-15 北京浩瀚深度信息技术股份有限公司 一种基于机器学习算法感知视频卡顿质差的方法

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114401447A (zh) * 2021-12-20 2022-04-26 北京字节跳动网络技术有限公司 一种视频卡顿预测方法、装置、设备和介质
CN114979799A (zh) * 2022-05-20 2022-08-30 北京字节跳动网络技术有限公司 一种全景视频处理方法、装置、设备和存储介质
CN115002557B (zh) * 2022-05-23 2024-01-30 北京字跳网络技术有限公司 一种网速预测方法、装置、设备和存储介质
CN114827750B (zh) * 2022-05-31 2023-12-22 脸萌有限公司 视角的预测方法、装置、设备及存储介质
CN116226702B (zh) * 2022-09-09 2024-04-26 武汉中数医疗科技有限公司 一种基于生物电阻抗的甲状腺采样数据识别方法
CN117750140A (zh) * 2023-12-27 2024-03-22 北京大麦星云科技有限公司 基于机顶盒的视频点播交互方法及系统

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110958491A (zh) * 2018-09-27 2020-04-03 中兴通讯股份有限公司 视频卡顿模型训练方法、预测方法、服务器及存储介质
CN111405327A (zh) * 2020-04-03 2020-07-10 广州市百果园信息技术有限公司 网络带宽预测模型训练方法、视频数据播放方法及装置
CN112584417A (zh) * 2019-09-27 2021-03-30 中国移动通信有限公司研究院 无线业务质量确定方法、装置、计算机设备和存储介质
CN113055745A (zh) * 2021-03-30 2021-06-29 脸萌有限公司 视频播放的处理方法、装置、设备及存储介质
WO2021190090A1 (zh) * 2020-03-27 2021-09-30 北京金山云网络技术有限公司 判断播放卡顿的方法、装置和电子终端
CN114401447A (zh) * 2021-12-20 2022-04-26 北京字节跳动网络技术有限公司 一种视频卡顿预测方法、装置、设备和介质

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108900856B (zh) * 2018-07-26 2020-04-28 腾讯科技（深圳）有限公司 一种视频帧率预测方法、装置及设备
CN110225417B (zh) * 2019-05-09 2022-06-10 网宿科技股份有限公司 数据处理方法及服务器、检测卡顿的方法及服务器
CN111031403B (zh) * 2019-11-05 2022-01-18 网宿科技股份有限公司 一种卡顿检测方法、系统及设备
CN111263225A (zh) * 2020-01-08 2020-06-09 恒安嘉新（北京）科技股份公司 视频卡顿预测方法、装置、计算机设备及存储介质
CN112788360B (zh) * 2020-12-30 2023-06-20 北京达佳互联信息技术有限公司 直播方法、直播装置与计算机程序产品
CN112887795B (zh) * 2021-01-26 2023-04-21 脸萌有限公司 视频播放方法、装置、设备和介质
CN112995712A (zh) * 2021-02-10 2021-06-18 北京字节跳动网络技术有限公司 卡顿因素的确定方法、装置、设备及存储介质
CN113784216B (zh) * 2021-08-24 2024-05-31 咪咕音乐有限公司 视频卡顿识别方法、装置、终端设备以及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110958491A (zh) * 2018-09-27 2020-04-03 中兴通讯股份有限公司 视频卡顿模型训练方法、预测方法、服务器及存储介质
CN112584417A (zh) * 2019-09-27 2021-03-30 中国移动通信有限公司研究院 无线业务质量确定方法、装置、计算机设备和存储介质
WO2021190090A1 (zh) * 2020-03-27 2021-09-30 北京金山云网络技术有限公司 判断播放卡顿的方法、装置和电子终端
CN111405327A (zh) * 2020-04-03 2020-07-10 广州市百果园信息技术有限公司 网络带宽预测模型训练方法、视频数据播放方法及装置
CN113055745A (zh) * 2021-03-30 2021-06-29 脸萌有限公司 视频播放的处理方法、装置、设备及存储介质
CN114401447A (zh) * 2021-12-20 2022-04-26 北京字节跳动网络技术有限公司 一种视频卡顿预测方法、装置、设备和介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117241071A (zh) * 2023-11-15 2023-12-15 北京浩瀚深度信息技术股份有限公司 一种基于机器学习算法感知视频卡顿质差的方法
CN117241071B (zh) * 2023-11-15 2024-02-06 北京浩瀚深度信息技术股份有限公司 一种基于机器学习算法感知视频卡顿质差的方法

Also Published As

Publication number Publication date
CN114401447A (zh) 2022-04-26

Similar Documents

Publication Publication Date Title
WO2023116233A1 (zh) 视频卡顿预测方法、装置、设备和介质
WO2022028304A1 (zh) 一种多媒体数据处理方法、装置、设备及可读存储介质
CN108848060B (zh) 一种多媒体文件处理方法、处理系统及计算机可读存储介质
EP1438673B1 (en) System and method for communicating media signals
WO2023165372A1 (zh) 视频流获取方法、装置、系统、设备和介质
US10305955B1 (en) Streaming decision in the cloud
WO2023051243A1 (zh) 视频码率切换方法、装置、电子设备及存储介质
CN114389975B (zh) 网络带宽预估方法、装置、系统、电子设备及存储介质
WO2023093339A1 (zh) 一种基于智能数字视网膜的视频处理方法和装置
CN113891132A (zh) 一种音视频同步监控方法、装置、电子设备及存储介质
CN110248195B (zh) 用于输出信息的方法和装置
CN111385574B (zh) 视频编码中的码率控制方法、装置、移动终端及存储介质
CN111385576B (zh) 视频编码方法、装置、移动终端及存储介质
CN113726801A (zh) 应用于服务端的ab实验方法、装置、设备及介质
CN117714741A (zh) 视频文件处理方法、视频管理平台及存储介质
WO2022237427A1 (zh) 视频处理方法、装置、设备及存储介质
WO2022242470A1 (zh) Cdn的调度方法、装置、设备及存储介质
CN107920275B (zh) 视频播放方法、装置、终端及存储介质
CN112153322B (zh) 数据分发方法、装置、设备及存储介质
CN117336521A (zh) 视频流传输的控制方法及装置、设备、介质
CN113242446B (zh) 视频帧的缓存方法、转发方法、通信服务器及程序产品
CN112057832A (zh) 一种基于移动流媒体技术的实景骑行训练系统
CN113364672B (zh) 媒体网关信息确定方法、装置、设备和计算机可读介质
WO2022237461A1 (zh) 视频的预加载方法、装置、设备及存储介质
CN115442277B (zh) 一种提升5g溯源关联正确性的方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22909550

Country of ref document: EP

Kind code of ref document: A1