A video path optimization method based on AI
Through the AI-based video path optimization method, combined with monitoring data and user data, and using artificial intelligence algorithm evaluation model to perform video path switching, the reliability and stability problems during video path switching in the existing technology are solved, and video playback is realized Reliable and balanced switching.
Patent Information
- Application Number
- CN202410398806.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-03
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-04-03
AI Technical Summary
The prior art cannot effectively ensure the reliability of video playback and the viewing effect of the audience when switching video paths, especially after the target video path fails to play differentiated switches in time, resulting in the stability and balance of video playback. .
Using AI-based video path optimization method, by monitoring the number of playback failures and user data of the target video path, combining the monitoring data of the backup video path, the comprehensive evaluation volume of paths is determined using an artificial intelligence algorithm evaluation model, and differentiated user switching allocation is carried out. Ensure the reliability and stability of video playback.
By combining the credibility of monitoring data and user data, reliable switching of video paths is achieved, the impact of untrusted monitoring data on evaluation results is avoided, the stability and balance of video playback is ensured, and the video path switching is improved. Effect.
Smart Images

Figure CN118301392B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of artificial intelligence technology, and in particular relates to a video path optimization method based on AI. Background Art
[0002] In order to ensure the smoothness of online video playback, the existing technical solutions often set multiple video paths for different online videos, and automatically switch the video path when the video smoothness of the current video path cannot meet the requirements. However, at the same time, how to achieve automatic switching of video paths has become a technical problem that needs to be solved urgently.
[0003] In order to solve the above technical problems, in the invention patent CN201911319531.6 "A video path switching processing method and device", when the video of the target video source node fails to be obtained through the first target video path information, the video of the target video source node is obtained according to the second target video path information in the multiple video path information, thereby reducing the probability of video playback failure. However, it is not difficult to find through analysis that there are the following technical problems:
[0004] When the target video path fails to play, switching the target video path will inevitably affect the viewing experience of the target video audience to a certain extent. Therefore, if the differentiated switching timing cannot be judged based on the monitoring results of the target video path, the reliability of the target video playback cannot be guaranteed.
[0005] In response to the above technical problems, the present invention provides an AI-based video path optimization method. Summary of the invention
[0006] To achieve the purpose of the present invention, the present invention adopts the following technical solutions:
[0007] According to one aspect of the present invention, a video path optimization method based on AI is provided.
[0008] A video path optimization method based on AI, characterized by comprising:
[0009] S1 determines the change data of different types of monitoring data of the target video path within a preset time under different numbers of playback failures based on the playback failure data, and determines the credibility of the monitoring data of the target video path under different numbers of playback failures and the number of credible failures in combination with the change data of different types of monitoring data of other backup video paths within a preset time under different numbers of playback failures;
[0010] S2 obtains the change data of different types of monitoring data and user data of the target video path within a preset time, and determines the playback failure probability of the target video path in combination with the change data of different types of monitoring data under the number of credible failures, user data, and the credibility of the monitoring data. When the playback failure probability does not meet the requirement, proceed to the next step;
[0011] S3 determines the path comprehensive evaluation amounts of different candidate video paths and the valid video paths among the candidate video paths by using the evaluation model based on the artificial intelligence algorithm according to the user data of the candidate video paths and the probability of playback failure;
[0012] S4 performs switching allocation of users of the target video path according to the comprehensive evaluation amounts of different valid video paths.
[0013] A further technical solution is that the playback failure data is determined based on the analysis result of the operation log of the target video path.
[0014] A further technical solution is that the change data of the monitoring data includes the monitoring data of the target video path at different moments and the change amount of the monitoring data between different adjacent moments.
[0015] A further technical solution is that when the credibility of the monitoring data of the target video path under a specific number of playback failures meets the requirement, the specific number of playback failures is determined as a credible number of failures.
[0016] A further technical solution is to perform switching allocation of users of the target video path, specifically including:
[0017] The user allocation ratios of different effective video paths are determined based on the comprehensive evaluation amounts of the paths of different effective video paths, and the switching allocation of users of the target video path is performed in combination with the number of users of the target video path.
[0018] The beneficial effects of the present invention are:
[0019] 1. By determining the credibility of the monitoring data of the target video path under different numbers of playback failures and the number of credible failures, it is possible to screen the credible number of playback failures of the monitoring data from the perspective of the change data of the monitoring data of the target video path, and also lay the foundation for the differentiated evaluation of the current playback failure probability, avoiding the technical problem that the unreliable monitoring data will not be accurate enough for the final evaluation result.
[0020] 2. Switching allocation of users of the target video path is performed according to the comprehensive evaluation amount of the paths of different effective video paths, which not only takes into account the number of users of different effective video paths, but also ensures the reliability of switching allocation of users of the target video path by combining it with the playback failure probability of the effective video path, thus ensuring the balance of users of different effective video paths and the stability of the switching results of the target video path.
[0021] Other features and advantages will be described in the following description. The objects and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description and drawings.
[0022] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.
[0024] Figure 1 is a flow chart of an AI-based video path optimization method;
[0025] Figure 2 is a flow chart of a method for determining the credibility of monitoring data of a target video path under different playback failure times;
[0026] Figure 3 It is a flow chart of a method for determining a playback failure probability of a target video path. DETAILED DESCRIPTION
[0027] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.
[0028] The following will further elaborate from two perspectives: method-based embodiments and system-based embodiments.
[0029] Method Example
[0030] To solve the above problems, according to one aspect of the present invention, Figure 1 As shown, a video path optimization method based on AI is provided, which is characterized by specifically including:
[0031] S1 determines the change data of different types of monitoring data of the target video path within a preset time under different numbers of playback failures based on the playback failure data, and determines the credibility of the monitoring data of the target video path under different numbers of playback failures and the number of credible failures in combination with the change data of different types of monitoring data of other backup video paths within a preset time under different numbers of playback failures;
[0032] Furthermore, the change data of the monitoring data includes the monitoring data of the target video path at different moments and the change amount of the monitoring data between different adjacent moments.
[0033] Specifically, Figure 2 As shown, the method for determining the credibility of the monitoring data of the target video path under different playback failure times is:
[0034] Determine the change amount of the monitoring data at different moments within a preset time and the monitoring data between different adjacent moments based on the change data of different types of monitoring data under a specific number of playback failures of the target video path;
[0035] Determine, based on monitoring data at different moments and changes in monitoring data between different adjacent moments, a change similarity between the specific number of playback failures and different types of monitoring data of other number of playback failures of the target video path, and a change similarity between the specific number of playback failures and different types of monitoring data of other backup video paths;
[0036] The similarity of changes in the monitoring data is used to determine the number of similar failures under other playback failures of the target video path, and the credibility of the monitoring data of the target video path itself is determined by combining the similarity of changes under different similar failures;
[0037] The similar number of failures under the number of playback failures of the other backup video paths is determined by the similarity of changes in the monitoring data, and the credibility of the backup monitoring data of the target video path is determined in combination with the similarity of changes in the number of failures under the number of playback failures of the other backup video paths and the number of other backup video paths with similar number of failures;
[0038] The credibility of the monitoring data of the target video path under a specific number of playback failures is determined based on the credibility of the target video path's own monitoring data and the credibility of the backup monitoring data.
[0039] Further, when the credibility of the monitoring data of the target video path under a specific number of playback failures meets the requirement, the specific number of playback failures is determined to be a credible number of failures.
[0040] In another embodiment, the method for determining the credibility of the monitoring data of the target video path under different playback failure times is:
[0041] Determine the change amount of the monitoring data at different moments within a preset time and the monitoring data between different adjacent moments based on the change data of different types of monitoring data under a specific number of playback failures of the target video path;
[0042] Determine the change similarity between the specific number of playback failures and different types of monitoring data of other playback failures of the target video path and the change similarity between the specific number of playback failures and different types of monitoring data of other backup video paths according to the monitoring data at different times and the change amount of the monitoring data between different adjacent times, and judge whether there is a number of playback failures whose change similarity meets the requirement, if so, proceed to the next step, if not, determine that the specific number of playback failures does not belong to the credible number of failures;
[0043] The similarity of changes in the monitoring data is used to determine the number of similar failures under other playback failures of the target video path, and the credibility of the monitoring data of the target video path itself is determined by combining the similarity of changes under different similar failures;
[0044] The similar number of failures under the number of playback failures of the other backup video paths is determined by the similarity of changes in the monitoring data, and the credibility of the backup monitoring data of the target video path is determined in combination with the similarity of changes in the number of failures under the number of playback failures of the other backup video paths and the number of other backup video paths with similar number of failures;
[0045] When the credibility of the target video path's own monitoring data and the credibility of the backup monitoring data do not meet the requirements, it is determined that the specific number of playback failures does not belong to the credible number of failures;
[0046] When any one of the credibility of the target video path's own monitoring data and the credibility of the backup monitoring data meets the requirements, the credibility of the monitoring data of the target video path under a specific number of playback failures is determined based on the credibility of the target video path's own monitoring data and the credibility of the backup monitoring data.
[0047] In another embodiment, the method for determining the credibility of the monitoring data of the target video path under different playback failure times is:
[0048] Determine the change amount of the monitoring data at different moments within a preset time and the monitoring data between different adjacent moments based on the change data of different types of monitoring data under a specific number of playback failures of the target video path;
[0049] Determine, according to the monitoring data at different moments and the variation of the monitoring data between different adjacent moments, the variation similarity between the specific playback failure times and the different types of monitoring data of other playback failure times of the target video path and the variation similarity between the specific playback failure times and the different types of monitoring data of other backup video paths;
[0050] Determine the data similarity between the specific playback failure number and other playback failure numbers of the target video path and the data similarity between the specific playback failure number and other backup video paths through the change similarity of different types of monitoring data;
[0051] When it is determined according to the data similarity that other playback failure times of the target video path and other playback failure times of the backup video paths do not have similar failure times:
[0052] It is determined that the specific number of playback failures does not belong to the credible number of failures;
[0053] When determining, according to the data similarity, that any one of other playback failure times of the target video path and other playback failure times of other backup video paths has a similar number of failure times:
[0054] When the number of similar failures does not meet the preset number threshold requirement, it is determined that the specific number of playback failures does not belong to the credible number of failures;
[0055] When the number of similar failures meets the preset number threshold requirement, the similar failure number evaluation amount is determined according to the similar failure number of the target video path and the path weight value of the target video path, the similar failure number of other backup video paths and the path weight values of other backup video paths, and it is judged whether the similar failure number evaluation amount meets the requirement, if so, proceed to the next step, if not, it is determined that the specific playback failure number does not belong to the credible failure number;
[0056] The similarity of changes in the monitoring data is used to determine the number of similar failures under other playback failures of the target video path, and the credibility of the monitoring data of the target video path itself is determined by combining the similarity of changes under different similar failures;
[0057] The similar number of failures under the number of playback failures of the other backup video paths is determined by the similarity of changes in the monitoring data, and the credibility of the backup monitoring data of the target video path is determined in combination with the similarity of changes in the number of failures under the number of playback failures of the other backup video paths and the number of other backup video paths with similar number of failures;
[0058] The credibility of the monitoring data of the target video path under a specific number of playback failures is determined based on the credibility of the target video path's own monitoring data and the credibility of the backup monitoring data.
[0059] S2 obtains the change data of different types of monitoring data and user data of the target video path within a preset time, and determines the playback failure probability of the target video path in combination with the change data of different types of monitoring data under the number of credible failures, user data, and the credibility of the monitoring data. When the playback failure probability does not meet the requirement, proceed to the next step;
[0060] Specifically, the playback failure data is determined according to the analysis result of the operation log of the target video path.
[0061] Specifically, Figure 3 As shown, the method for determining the playback failure probability of the target video path is:
[0062] Determine the number of users of the target video path according to the user data of the target video path, determine the user condition similarity between the target video path and different trusted failure times based on the number of users, and determine the monitoring data similarity between the target video path and different trusted failure times based on the change data of different types of monitoring data of the target video path within a preset time;
[0063] Determine the comprehensive similarity between the target video path and different credible failure times based on the user condition similarity between the target video path and different credible failure times, the monitoring data similarity, and the credibility of the monitoring data of different credible failure times;
[0064] The playback failure probability of the target video path is determined by the comprehensive similarity between the target video path and different credible failure times.
[0065] Furthermore, the playback failure probability of the target video path ranges from 0 to 1, wherein when the playback failure probability of the target video path is greater than a preset probability, it is determined that the playback failure probability of the target video path does not meet the requirement.
[0066] In another embodiment, the method for determining the playback failure probability of the target video path is:
[0067] Determine the similarity between the target video path and the monitoring data of different trusted failure times based on the change data of different types of monitoring data of the target video path within a preset time, and judge whether there is a trusted failure time whose monitoring data similarity does not meet the requirement, if so, proceed to the next step, if not, determine whether the playback failure probability of the target video path meets the requirement;
[0068] The number of credible failures for which the monitoring data similarity meets the requirements is taken as the number of data similarities, the number of users of the target video path is determined according to the user data of the target video path, and the user condition similarity of the target video path with different data similarity times is determined based on the number of users, and it is judged whether there is a data similarity number for which the user condition similarity meets the requirements, if so, proceed to the next step, if not, it is determined that the playback failure probability of the target video path meets the requirements;
[0069] Determine the comprehensive similarity between the target video path and different credible failure times based on the user condition similarity between the target video path and different credible failure times, the monitoring data similarity, and the credibility of the monitoring data of different credible failure times;
[0070] The playback failure probability of the target video path is determined by the comprehensive similarity between the target video path and different credible failure times.
[0071] In another embodiment, the method for determining the playback failure probability of the target video path is:
[0072] Determine the number of users of the target video path according to the user data of the target video path, determine the user condition similarity between the target video path and different credible failure times based on the user number, determine the number of condition similarity times in the credible failure times based on the user condition similarity, and judge whether the number of condition similarity times meets the requirement, if so, proceed to the next step, if not, determine that the playback failure probability of the target video path meets the requirement;
[0073] Determine the monitoring data similarity between the target video path and different condition similarity times based on the change data of different types of monitoring data within the preset time, determine the number of data similarity times in the condition similarity times through the monitoring data similarity, and determine the comprehensive data similarity of the target video path in combination with the monitoring data similarity between the target video path and different data similarity times, and judge whether the comprehensive data similarity is less than a preset similarity threshold, if so, determine that the playback failure probability of the target video path meets the requirement, if not, proceed to the next step;
[0074] Determine the comprehensive similarity between the target video path and different credible failure times based on the user condition similarity, monitoring data similarity and the credibility of the monitoring data of different credible failure times, and judge whether there is a credible failure time whose comprehensive similarity meets the requirement, if so, proceed to the next step, if not, determine whether the playback failure probability of the target video path meets the requirement;
[0075] The number of matching failures in the credible number of failures is determined by using the comprehensive similarity, and the playback failure probability of the target video path is determined in combination with the comprehensive similarities between the target video path and different matching failure times.
[0076] S3 determines the path comprehensive evaluation amounts of different candidate video paths and the valid video paths among the candidate video paths by using the evaluation model based on the artificial intelligence algorithm according to the user data of the candidate video paths and the probability of playback failure;
[0077] Specifically, the method for determining the path comprehensive evaluation amount of the candidate video path is:
[0078] Determining the number of users of different alternative video paths based on the user data of the alternative video paths;
[0079] The comprehensive evaluation value of the path of the alternative video path is determined by an evaluation model based on an artificial intelligence algorithm through the number of users, the number of users that can be carried, and the probability of playback failure of the alternative video path.
[0080] Further, when the comprehensive evaluation value of the candidate video path meets the requirement, the candidate video path is determined to be a valid video path.
[0081] In another embodiment, the method for determining the path comprehensive evaluation amount of the candidate video path is:
[0082] Determine the number of users of the alternative video path based on the user data of the alternative video path, and determine whether the number of users of the alternative video path meets the requirement in combination with the number of users that the alternative video path can carry, if yes, proceed to the next step, if no, determine that the alternative video path does not belong to a valid video path;
[0083] Determine whether the probability of playback failure of the candidate video path is greater than a preset probability threshold, if so, determine that the candidate video path does not belong to a valid video path, if not, proceed to the next step;
[0084] The comprehensive evaluation value of the path of the alternative video path is determined by an evaluation model based on an artificial intelligence algorithm through the number of users, the number of users that can be carried, and the probability of playback failure of the alternative video path.
[0085] Furthermore, the artificial intelligence algorithm is constructed using one or more of a BP neural network, a CNN convolutional neural network, and an RNN recurrent neural network.
[0086] S4 performs switching allocation of users of the target video path according to the comprehensive evaluation amounts of different valid video paths.
[0087] Specifically, performing the switching allocation of users of the target video path specifically includes:
[0088] The user allocation ratios of different effective video paths are determined based on the comprehensive evaluation amounts of the paths of different effective video paths, and the switching allocation of users of the target video path is performed in combination with the number of users of the target video path.
[0089] Through the above embodiments, the present invention achieves the following beneficial effects:
[0090] 1. By determining the credibility of the monitoring data of the target video path under different numbers of playback failures and the number of credible failures, it is possible to screen the credible number of playback failures of the monitoring data from the perspective of the change data of the monitoring data of the target video path, and also lay the foundation for the differentiated evaluation of the current playback failure probability, avoiding the technical problem that the unreliable monitoring data will not be accurate enough for the final evaluation result.
[0091] 2. Switching allocation of users of the target video path is performed according to the comprehensive evaluation amount of the paths of different effective video paths, which not only takes into account the number of users of different effective video paths, but also ensures the reliability of switching allocation of users of the target video path by combining it with the playback failure probability of the effective video path, thus ensuring the balance of users of different effective video paths and the stability of the switching results of the target video path.
[0092] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0093] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0094] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.
Claims
1. A video path optimization method based on AI, It is characterized in that Specifically include: Determine the change data of different types of monitoring data of the target video path within a preset time under different numbers of playback failures based on the playback failure data, and determine the credibility of the monitoring data of the target video path under different numbers of playback failures and the number of credible failures in combination with the change data of different types of monitoring data of other alternative video paths within a preset time under different numbers of playback failures; Obtaining the change data of different types of monitoring data and user data of the target video path within a preset time, and determining the playback failure probability of the target video path in combination with the change data of different types of monitoring data and user data under the number of credible failures and the credibility of the monitoring data, and when the playback failure probability does not meet the requirement, proceeding to the next step; Using user data of alternative video paths and playback failure probabilities, an evaluation model based on an artificial intelligence algorithm is used to determine comprehensive evaluation quantities of different alternative video paths and valid video paths among the alternative video paths; Perform switching allocation of users of the target video path according to the comprehensive evaluation quantities of different effective video paths; The method for determining the playback failure probability of the target video path is: Determine the number of users of the target video path according to the user data of the target video path, determine the user condition similarity between the target video path and different trusted failure times based on the number of users, and determine the monitoring data similarity between the target video path and different trusted failure times based on the change data of different types of monitoring data of the target video path within a preset time; Determine the comprehensive similarity between the target video path and different credible failure times based on the user condition similarity between the target video path and different credible failure times, the monitoring data similarity, and the credibility of the monitoring data of different credible failure times; Determining the playback failure probability of the target video path through the comprehensive similarity between the target video path and different credible failure times; The playback failure probability of the target video path ranges from 0 to 1, wherein when the playback failure probability of the target video path is greater than a preset probability, it is determined that the playback failure probability of the target video path does not meet the requirement.
2. The AI-based video path optimization method according to claim 1, It is characterized in that The playback failure data is determined based on the analysis result of the operation log of the target video path.
3. The AI-based video path optimization method according to claim 1, It is characterized in that The change data of the monitoring data includes the monitoring data of the target video path at different moments and the change amount of the monitoring data between different adjacent moments.
4. The AI-based video path optimization method according to claim 1, It is characterized in that The method for determining the credibility of the monitoring data of the target video path under different playback failure times is as follows: Determine the change amount of the monitoring data at different moments within a preset time and the monitoring data between different adjacent moments based on the change data of different types of monitoring data under a specific number of playback failures of the target video path; Determine, based on monitoring data at different moments and changes in monitoring data between different adjacent moments, a change similarity between the specific number of playback failures and different types of monitoring data of other number of playback failures of the target video path, and a change similarity between the specific number of playback failures and different types of monitoring data of other number of playback failures of other candidate video paths; The similarity of changes in the monitoring data is used to determine the number of similar failures under other playback failures of the target video path, and the credibility of the monitoring data of the target video path itself is determined by combining the similarity of changes under different similar failures; The similar number of failures under the number of playback failures of the other candidate video paths is determined by the similarity of changes in the monitoring data, and the credibility of the candidate monitoring data of the target video path is determined in combination with the similarity of changes in the number of failures under the number of playback failures of the other candidate video paths and the number of other candidate video paths with similar number of failures; The credibility of the monitoring data of the target video path under a specific number of playback failures is determined based on the credibility of the target video path's own monitoring data and the credibility of the alternative monitoring data.
5. The AI-based video path optimization method according to claim 4, It is characterized in that When the credibility of the monitoring data of the target video path under a specific number of playback failures meets the requirement, the specific number of playback failures is determined to be a credible number of failures.
6. The AI-based video path optimization method according to claim 1, It is characterized in that The method for determining the path comprehensive evaluation amount of the candidate video path is: Determining the number of users of different candidate video paths based on the user data of the candidate video paths; The comprehensive evaluation value of the path of the alternative video path is determined by an evaluation model based on an artificial intelligence algorithm through the number of users, the number of users that can be carried, and the probability of playback failure of the alternative video path.
7. The AI-based video path optimization method according to claim 1, It is characterized in that When the comprehensive evaluation value of the candidate video path meets the requirement, the candidate video path is determined to be a valid video path.
8. The AI-based video path optimization method according to claim 1, It is characterized in that Performing switching allocation of users of the target video path specifically includes: The user allocation ratios of different effective video paths are determined based on the comprehensive evaluation amounts of the paths of different effective video paths, and the switching allocation of users of the target video path is performed in combination with the number of users of the target video path.
Citation Information
Patent Citations
Video path switching processing method and device
CN110996139A
Video playing method and device, terminal device and server
CN103458305A
Watching channel path counting method and watching channel path counting device
CN106658083A
Non-reference adaptive streaming media quality evaluation method and system based on neural network optimization
CN111083125A
Code rate self-adaption method, device and equipment and readable storage medium
CN113726759A
