Special effect frame rate prediction method and device, electronic equipment and storage medium
By obtaining and inputting special effects and terminal feature data into the prediction model, the problem of the difference in special effects operation frame rate on different terminal devices is solved, efficient and accurate frame rate prediction is achieved, and user experience is improved.
Patent Information
- Application Number
- CN202311568413.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-11-22
- Publication Date
- 2025-05-23
AI Technical Summary
Since there are differences in the running frame rates of the same special effects on different terminal devices, how to quickly understand and predict the running frame rates of the special effects on different devices has become an urgent problem.
By obtaining the static feature data and dynamic feature data of the target special effects, as well as the model data of the target type terminal, these data are input into the special effects frame rate prediction model to obtain the running frame rate prediction result of the target special effects in the target type terminal.
It realizes accurate operation frame rate prediction for any model terminal and special effects, improves prediction efficiency and accuracy, and is suitable for video editing, game development and other fields.
Smart Images
Figure CN120029862A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of Internet technology, and in particular to a special effect frame rate prediction method, device, electronic device and storage medium. Background Art
[0002] With the development of technology and the increasing demand for entertainment, video editing applications provide special effects editing functions to meet users' needs for video special effects. Users can use these applications to add various special effects to videos to increase the fun of video content.
[0003] However, the running frame rate of special effects on terminal devices has an important impact on user experience. The higher the running frame rate of special effects, the clearer and smoother the picture, and the better the user's visual experience. But at the same time, the high running frame rate also increases the performance requirements of terminal devices accordingly. Therefore, in order to enable users to have a good experience when using special effects, it is necessary to evaluate the running frame rate of special effects on different terminal devices before the special effects go online. However, since the running frame rates of the same special effect on different terminal devices are different, how to quickly understand and predict the running frame rates of special effects on different devices has become a problem that needs to be solved urgently. Summary of the invention
[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the present disclosure provides a special effect frame rate prediction method, device, electronic device and storage medium.
[0005] In a first aspect, the present disclosure provides a special effect frame rate prediction method, comprising:
[0006] Obtain static feature data and dynamic feature data of the target special effect, as well as model data of the target type terminal;
[0007] The static feature data, the dynamic feature data and the model data are input into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
[0008] In a second aspect, the present disclosure further provides a special effect frame rate prediction device, comprising:
[0009] An acquisition module, used to acquire static feature data and dynamic feature data of a target special effect, as well as model data of a target type terminal;
[0010] The prediction module is used to input the static feature data, the dynamic feature data and the model data into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
[0011] In a third aspect, the present disclosure further provides an electronic device, the electronic device comprising:
[0012] one or more processors;
[0013] A storage device for storing one or more programs;
[0014] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described above.
[0015] In a fourth aspect, the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, which implements the method described above when the program is executed by a processor.
[0016] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has the following advantages:
[0017] The technical solution provided by the embodiment of the present disclosure obtains static feature data and dynamic feature data of the target special effect, as well as the model data of the target type terminal by setting; the static feature data, dynamic feature data and model data are input into the special effect frame rate prediction model to obtain the predicted result of the running frame rate of the target special effect on the target type terminal. This process is essentially to use the model to predict the running frame rate of the special effect on different devices. Considering that there are many types of terminal devices on the market, such as mobile phones, tablets, personal computers, game consoles, etc., and the terminals of the same category also have many models and are constantly updated. It is very difficult to understand the running frame rate of special effects on different devices through manual testing. Through the above technical solution, regardless of the terminal model, the running frame rate of the special effect on it can be accurately predicted, which has wide applicability and high precision. This technical solution solves the problem of special effect frame rate prediction, and can efficiently and accurately predict the running frame rate for any model of terminal and special effect, improve the prediction efficiency and accuracy, and provide strong support for video editing, game development, educational visualization and other fields. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.
[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0020] Figure 1 A flowchart of a special effects frame rate prediction method provided by an embodiment of the present disclosure;
[0021] Figure 2A schematic diagram of the working principle of a special effects frame rate prediction model provided in an embodiment of the present disclosure;
[0022] Figure 3 A flowchart of a special effects frame rate prediction model training method provided in an embodiment of the present disclosure;
[0023] Figure 4 A schematic diagram of the training principle of a special effects frame rate prediction model provided in an embodiment of the present disclosure;
[0024] Figure 5 is a structural schematic diagram of a special effect frame rate prediction device in an embodiment of the present disclosure;
[0025] Figure 6 It is a structural schematic diagram of an electronic device in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0026] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.
[0027] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.
[0028] Figure 1 This is a flowchart of a special effect frame rate prediction method provided in an embodiment of the present disclosure. This embodiment can be applied to the case where special effect frame rate prediction is performed in a client or server. The method can be performed by a special effect frame rate prediction device, which can be implemented in software and / or hardware, and can be configured in an electronic device, such as a terminal or a server. The terminal specifically includes but is not limited to a smart phone, a PDA, a tablet computer, a wearable device with a display screen, a desktop computer, a laptop computer, an all-in-one machine, a smart home device, etc.
[0029] like Figure 1 As shown, the method may specifically include:
[0030] S110: Obtain static feature data and dynamic feature data of a target special effect, and model data of a target type terminal.
[0031] The target special effect can be, for example, a special effect for which the running frame rate on the terminal device is desired to be evaluated. The target special effect can be, for example, at least one of the following: a special effect of adding a sticker or label to an image frame; a special effect of beautifying an object in an image frame, a special effect of stylizing an object in an image frame, a special effect of blurring an image frame, a special effect of sharpening an image frame, a special effect of adding noise to an image frame, a special effect of changing the hue of an image frame, a special effect of adding a halo to an image frame, a special effect of changing the perspective of an image frame, a special effect of making an image frame present a 3D effect. Among them, the special effect of the sticker or label specifically includes, but is not limited to, text-based special effects, contour-based special effects, and graffiti-based special effects. These special effects can be used in different fields such as video editing, game development, and educational visualization.
[0032] The target type terminal can be, for example, the terminal that "runs" the target special effect when evaluating the running frame rate of the target special effect. The target type terminal can be a terminal device such as a mobile phone, a tablet computer, a personal computer, a game console, etc.
[0033] The target type terminal represents a specific type of terminal, and its brand, model, and configuration are all clear. In each process of executing the special effect frame rate prediction method provided by this application, any one of the brand, model, and configuration of the target type terminal remains unchanged. However, when the special effect frame rate prediction method provided by this application is executed at different times, if the target special effect is the same, at least one of the brand, model, and configuration of the target type terminal is different.
[0034] Exemplarily, assume that it is desired to evaluate the running frame rate of special effect A on a terminal device of model B. Special effect A is the target special effect, and the terminal device of model B is the target type terminal.
[0035] It should be emphasized that using the technical solution provided by this application to evaluate the running frame rate of a special effect is obtained by predicting through a special effect frame rate prediction model, rather than actually running the target special effect on the target type terminal.
[0036] The static feature data of the target special effect may be, for example, data for describing the static properties of the special effect. Static properties refer to properties that remain unchanged regardless of the scenario, the model of the target type terminal, or the characteristics of the video or image to which the target special effect is to be added. Exemplarily, the static feature data of the target special effect may include, for example, at least one of the following: the number of filters included in the target special effect, the number of beauty makeup included in the target special effect, the number of masks included in the target special effect, the number of textures included in the target special effect, and the number of model resources involved in the target special effect. The dynamic feature data of the target special effect may be, for example, data for describing the dynamic properties of the special effect. Dynamic properties refer to properties that will change accordingly when the usage scenario is different, or the model of the target type terminal is different, or the characteristics of the video or image to which the target special effect is to be added are different. Exemplarily, the dynamic feature data of the target special effect may include, for example, at least one of the following: using the target special effect rendering in the target type terminal to render the video, the time consumed to render each frame of the picture and the total time consumed to render the entire video, the resolution of the video to be rendered, the number of faces processed using the target special effect, whether to turn on the beauty and whether to turn on the filter, and whether to reduce the resolution. It should be noted that the dynamic feature data of the target special effect can be determined according to the usage scenario to be simulated. For example, the dynamic feature data of the target special effect includes: turning on beauty, the number of faces processed by the target special effect is 1, no filter is turned on, and no resolution is reduced.
[0037] The model data of the target type terminal may be, for example, data used to describe the hardware performance of the target type terminal. The model data of the target type terminal is static data and does not change with the usage scenario, target special effects, or the video or image to which the target special effects need to be added. Exemplarily, the model data of the target type terminal may include, for example, at least one of the following: the system type, number of CPU cores, CPU model, GPU model, memory size, and model name of the target type terminal.
[0038] S120, inputting the static feature data, the dynamic feature data and the model data into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
[0039] The special effects frame rate prediction model is a pre-trained model that can predict the running frame rate of the target special effects on the target type of terminal based on static feature data, dynamic feature data and model data.
[0040] In some embodiments, the special effects frame rate prediction model is a gradient boosting tree regression model.
[0041] In practice, the specific structure of the special effect frame rate prediction model is different, and the specific implementation method of this step is different. This application does not limit this.
[0042] Exemplarily, if the special effects frame rate prediction model includes a first prediction model and a second prediction model; the target special effect includes multiple static feature data and multiple dynamic feature data, and the target type terminal includes multiple model data; the specific implementation method of this step may include: determining the target data from multiple static feature data, multiple dynamic feature data and multiple model data, and the importance level of the target data is greater than or equal to the set level; inputting multiple static feature data, multiple dynamic feature data and multiple model data into the first prediction model to obtain an intermediate prediction result; inputting the intermediate prediction result and the target data into the second prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
[0043] The target data is one or more data selected from multiple static feature data, multiple dynamic feature data, and multiple model data, which have a greater impact on the special effect running frame rate.
[0044] The importance level can be, for example, an indicator used to characterize the degree of influence of data (including static feature data, dynamic feature data, and model data) on the running frame rate. For example, if the importance level of a piece of data (the data can be static feature data, dynamic feature data, or model data) is higher, it means that the data has a higher degree of influence on the running frame rate.
[0045] The set level is a pre-specified level. This application does not limit which level the set level specifically refers to. For example, if the importance level of a static feature data is greater than or equal to the set level, it means that the static feature data has a greater impact on the target special effect running frame rate, and it needs to be considered when predicting the running frame rate of the target special effect on the target type terminal.
[0046] It should be noted that which static feature data, dynamic feature data and model data have an importance level greater than or equal to the set level have been clarified before executing the technical method provided by this application.
[0047] In addition, if the target data contains multiple data, when actually determining which data to use as the target data, it is necessary to consider the relationship between these data to ensure that the overall target data has a greater impact on the running frame rate of the target special effect. In other words, it is necessary to consider how to select, process or combine these data in order to accurately predict the running frame rate of the target special effect.
[0048] For example, suppose you want to evaluate the running frame rate of special effect A on terminal device model B, and you have obtained 5 static feature data and 5 dynamic feature data, as well as 3 model data of the target type terminal. According to the pre-setting, the importance levels of dynamic feature data 2, dynamic feature data 3, and model data 1 are all greater than or equal to the set level, which are the target data. When making predictions, refer to Figure 2, input 5 static feature data, 5 dynamic feature data, and 3 model data into the first prediction model to obtain the intermediate prediction result. Then input dynamic feature data 2, dynamic feature data 3, model data 1 and the intermediate prediction result into the second prediction model to obtain the final prediction result, that is, the running frame rate prediction result of the target special effect on the target type terminal.
[0049] By setting the special effect frame rate prediction model to include the first prediction model and the second prediction model, the essence is to use the two prediction models to cooperate with each other to predict the special effect frame rate. The benefits of doing so are that the complexity and generalization ability of the special effect frame rate prediction model can be increased, the prediction accuracy can be improved, the robustness can be increased, and the convergence speed can be accelerated.
[0050] The above technical solution obtains the static feature data and dynamic feature data of the target special effect and the model data of the target type terminal by setting; the static feature data, dynamic feature data and model data are input into the special effect frame rate prediction model to obtain the predicted result of the running frame rate of the target special effect on the target type terminal. This process is essentially to use the model to predict the running frame rate of the special effect on different devices. Considering that there are many types of terminal devices on the market, such as mobile phones, tablets, personal computers, game consoles, etc., and the terminals of the same category also have many models and are constantly updated. It is very difficult to understand the running frame rate of special effects on different devices through manual testing. Through the above technical solution, regardless of the terminal model, the running frame rate of the special effect on it can be accurately predicted, which has wide applicability and high precision. This technical solution solves the problem of special effect frame rate prediction, and can efficiently and accurately predict the running frame rate for any model of terminal and special effect, improve the prediction efficiency and accuracy, and provide strong support for video editing, game development, educational visualization and other fields.
[0051] On the basis of the above technical solution, optionally, after S120, the method further includes: if the predicted result of the running frame rate of the target special effect on the target type terminal is greater than or equal to the preset frame rate, determining that the target special effect meets the online conditions of the target type terminal.
[0052] Among them, the preset frame rate is pre-set and is used to measure whether the target special effect can be launched on the target type terminal. If the running frame rate of the target special effect on the target type terminal is lower than the preset frame rate, it may cause problems such as screen freeze and unsmoothness, affecting the user experience. If the running frame rate of the target special effect on the target type terminal is greater than or equal to the preset frame rate, it means that the target special effect can run smoothly on the target type terminal and meets the launch conditions. The target special effect can be deployed to the target type terminal later and released for users to use.
[0053] Based on the above technical solution, before S120, it is necessary to train the special effect frame rate prediction model. There are many methods for training the special effect frame rate prediction model, and this application does not limit this. Figure 3 ,The special effects frame rate prediction model training method includes:
[0054] S310. Obtain an original training data set, where the original training data set includes multiple original test records; the original test records include static feature data and dynamic feature data of corresponding sample special effects, model data of the sample type terminal, and actual running frame rate of the sample special effects on the sample type terminal.
[0055] The meanings of sample special effects and target special effects are similar, the only difference is that sample special effects are special effects used in the special effect frame rate prediction model training phase, while target special effects are special effects used in the special effect frame rate prediction model use phase.
[0056] The meanings of sample type terminals and target type terminals are similar, with two main differences: First, sample type terminals are terminals used in the special effect frame rate prediction model training phase; while target type terminals are terminals used in the special effect frame rate prediction model usage phase. Second, in order to obtain the original test record, it is necessary to actually run the sample special effect on the sample type terminal. That is, the actual running frame rate in the original test record is the actual measured result of running the target special effect on the target type terminal.
[0057] Each original test record records the actual test results of a sample special effect on a sample type terminal. These test results include the static and dynamic feature data of the sample special effect, as well as the model data of the sample type terminal and the actual running frame rate of the sample special effect on the sample type terminal. These data are all records of the actual test results, and each data is directly related to its corresponding actual test result.
[0058] The original training data set is a set of multiple original test records. In practice, different original test records have different static feature data, dynamic feature data or model data; or different original test records are generated at different times.
[0059] S320: Train a special effects frame rate prediction model using the original training data set.
[0060] In actual tests, the inventors found that when the same special effect is run on the same type of device, its actual running frame rate is not constant, but fluctuates. In order to optimize the training process of the special effect frame rate prediction model and provide clear guidance for it, it is necessary to establish a target running frame rate. This target running frame rate is a fixed value, which can be regarded as the most likely running frame rate of the special effect on a specific device. In this way, in the process of training the special effect frame rate prediction model, the target running frame rate is used as a benchmark to help the special effect frame rate prediction model better learn and predict the frame rate of the special effect. Based on this, optionally, after S310, the method includes: aggregating the original test records in the original training data set to obtain multiple test record groups; in the same test record group, the static feature data, dynamic feature data and model data of each original test record are the same; processing each test record group as training data respectively; the training data includes static feature data and dynamic feature data of sample special effects with corresponding relationships, model data of the sample type terminal, and target operating frame rate; the target operating frame rate is obtained according to the actual operating frame rate of each original test record in the test record group; S320 includes: using the training data to train the special effects frame rate prediction model.
[0061] Table 1 Several original test records and training data
[0062]
[0063]
[0064] Exemplarily, referring to Table 1, it is assumed that the original training data set includes 20 original test records. Among them, the static feature data, dynamic feature data and model data of original test records 1-3 and original test record 9 are the same, and original test records 1-3 and test record 9 can be regarded as elements in a test record group (hereinafter referred to as test record group 1). The static feature data, dynamic feature data and model data of original test records 4-8 and original test records 10-14 are the same, and original test records 4-8 and original test records 10-14 can be regarded as elements in a test record group (hereinafter referred to as test record group 2). The static feature data, dynamic feature data and model data of original test records 15-20 are the same, and original test records 15-20 can be regarded as elements in a test record group (hereinafter referred to as test record group 3).
[0065] In actual operation, for any test record group, one actual running frame rate can be randomly selected from the actual running frame rates of the original test records included therein as the target running frame rate; or the actual running frame rate that accounts for the largest proportion of the actual running frame rates of the original test records included therein can be selected as the target running frame rate.
[0066] Optionally, the target running frame rate can also be set to the actual running frame rate with the largest proportion within the preset distribution range in the ascending queue of the actual running frame rate in the test record group corresponding to it. The preset distribution range is a pre-specified range, and the present application does not limit the specific range referred to therein. Exemplarily, the preset distribution range is the range between 50% and 100%. Exemplarily, if a test record group contains 10 original test records, there are 10 actual running frame rates in total. Arrange the actual running frame rates in order from small to large according to the actual running frame rate to obtain the actual running frame rate queue. In the set consisting of the 5th to 10th actual running frame rates in the actual running frame rate queue, the actual running frame with the largest number of occurrences is the target running frame rate. The purpose of doing this is to better reflect the actual operation of the special effects in the test record group and provide it with a more clear and representative guiding value. And the target running frame rate determined in this way is more accurate, which is conducive to improving the prediction accuracy of the model trained subsequently.
[0067] Exemplarily, referring to Table 1, test record group 1 is processed into training data 1. The static feature data, dynamic feature data, and model data in training data 1 are consistent with the static feature data, dynamic feature data, and model data of any original test data included in the corresponding test record group 1. The target operating frame rate in training data 1 is the actual operating frame rate with the largest proportion between 50% and 100% in the ascending queue of the actual operating frame rate in test record group 1.
[0068] Similarly, test record group 2 is processed as training data 2, and test record group 3 is processed as training data 3.
[0069] Furthermore, the sample special effects include multiple static feature data and multiple dynamic feature data, and the sample type terminal includes multiple model data; before using the training data to train the special effects frame rate prediction model, the method also includes: determining the importance level of each static feature data, each static feature data and each model data; S320 includes: based on the importance level of each static feature data, each dynamic feature data and each model data, using the training data to train the special effects frame rate prediction model.
[0070] The importance level can be, for example, an indicator used to characterize the degree of influence of data (including static feature data, dynamic feature data, and model data) on the running frame rate. For example, if the importance level of a piece of data (the data can be static feature data, dynamic feature data, or model data) is higher, it means that the data has a higher degree of influence on the running frame rate.
[0071] Furthermore, the special effects frame rate prediction model includes a first prediction model and a second prediction model; based on the importance levels of each static feature data, each dynamic feature data and each model data, the special effects frame rate prediction model is trained using training data, including: taking static feature data, static feature data and model data whose importance level is greater than or equal to the set level as target sample data; inputting all static feature data, dynamic feature data and model data in the training data into the first prediction model to obtain a first frame rate prediction value; inputting the first frame rate prediction value and the target sample data into the second prediction model to obtain a second frame rate prediction value; based on the first frame rate prediction value and the target operating frame rate, adjusting the parameters in the first prediction model; and / or, based on the second frame rate prediction value and the target operating frame rate, adjusting the parameters in the second prediction model.
[0072] The set level is a pre-specified level. This application does not limit which level the set level specifically refers to. For example, if the importance level of a static feature data is greater than or equal to the set level, it means that the static feature data has a greater impact on the target special effect running frame rate and needs to be considered.
[0073] The target sample data has a similar meaning to the target data and includes the same data. The only difference is that the target sample data is the data used in the special effects frame rate prediction model training stage, while the target data is the data used in the special effects frame rate prediction model use stage.
[0074] Similarly, if the target sample data includes multiple data, when actually determining which data to use as the target sample data, the relationship between these data needs to be considered to ensure that the overall target sample data has a greater impact on the running frame rate of the target special effect.
[0075] Specifically, all static feature data, static feature data and model data can be deduplicated first to remove static feature data, static feature data and model data with high similarity; then the correlation between each data in the remaining data and the running frame rate, and / or the correlation between the combination of each data and the running frame rate is calculated; based on the correlation with the running frame rate, determine which data to use as target sample data.
[0076] For example, suppose a training data reflects the actual measured results of the running frame rate of special effect C on the terminal device of model D. The training data includes 5 static feature data and 5 dynamic feature data, 3 model data and the target running frame rate. Since the importance levels of dynamic feature data 2, dynamic feature data 3 and model data 1 are all greater than or equal to the set level, they are the target sample data. When training, refer to Figure 4, input 5 static feature data, 5 dynamic feature data, and 3 model data into the first prediction model to obtain the first frame rate prediction value. Then input dynamic feature data 2, dynamic feature data 3, model data 1, and the first frame rate prediction value into the second prediction model to obtain the second frame rate prediction value. Based on the first frame rate prediction value and the target operating frame rate, adjust the parameters in the first prediction model; based on the second frame rate prediction value and the target operating frame rate, adjust the parameters in the second prediction model.
[0077] When adjusting the parameters in the first prediction model and / or the second prediction model, a performance evaluation function for evaluating the model may be defined. Based on the output result of the model performance evaluation function of the first prediction model and / or the output result of the model performance evaluation function of the second prediction model, the parameters in the first prediction model and / or the second prediction model are adjusted. Exemplarily, the performance evaluation function may be the proportion of errors less than 3fps.
[0078] Optionally, the R2 index or Bayesian optimization algorithm can also be used to adjust the parameters of the model. The R2 index is a regression model evaluation index, which is used to reflect the degree of fit between the model prediction value and the actual value.
[0079] Optionally, based on the importance levels of each static feature data, each dynamic feature data and each model data, the static feature data, each dynamic feature data and each model data are preprocessed before the special effect frame rate prediction model is trained using the training data. The purpose of the preprocessing is to unify the scales of features of different scales, eliminate the influence of the dimensions and / or numerical values between data features, thereby enhancing the stability of the model; and convert the data into a format that is easy for the model to understand and use.
[0080] Optionally, in the process of training the special effects frame rate prediction model, the special effects frame rate prediction model can also be pruned. The purpose of model pruning is to reduce the complexity of the model, thereby avoiding overfitting of the model. There are many methods for model pruning. Such as early stopping method, reducing model depth method, increasing data iteration step method and increasing data splitting node requirement method. Among them, the early stopping method refers to stopping the training of the model in advance during the training process. The method of reducing the depth of the model can reduce the complexity of the model by reducing the structural level of the model. Increasing the data iteration step can better explore the parameter space of the model, thereby finding better model parameters. The method of increasing the data splitting node requirement can reduce the number of nodes by raising the standard for splitting nodes, thereby reducing the complexity of the model.
[0081] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.
[0082] Figure 5 Schematic diagram of the structure of a special effect frame rate prediction device in an embodiment of the present disclosure. The special effect frame rate prediction device provided in the embodiment of the present disclosure can be configured in a client, or can be configured in a server. Figure 5 , the special effect frame rate prediction device specifically includes:
[0083] An acquisition module 510 is used to acquire static feature data and dynamic feature data of a target special effect, and model data of a target type terminal;
[0084] The prediction module 520 is used to input the static feature data, the dynamic feature data and the model data into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
[0085] Furthermore, the device also includes a determination module for determining that the target special effect meets the online condition of the target type terminal if the predicted result of the running frame rate of the target special effect on the target type terminal is greater than or equal to a preset frame rate.
[0086] Furthermore, the special effect frame rate prediction model includes a first prediction model and a second prediction model; the target special effect includes a plurality of static feature data and a plurality of dynamic feature data, and the target type terminal includes a plurality of model data;
[0087] The prediction module 520 is used to:
[0088] Determine target data from the plurality of static feature data, the plurality of dynamic feature data, and the plurality of machine model data, wherein the importance level of the target data is greater than or equal to a set level;
[0089] Inputting the plurality of static feature data, the plurality of dynamic feature data and the plurality of aircraft model data into the first prediction model to obtain an intermediate prediction result;
[0090] The intermediate prediction result and the target data are input into the second prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
[0091] Furthermore, the device also includes a training module, which is used to:
[0092] Acquire an original training data set, wherein the original training data set includes a plurality of original test records; the original test records include static feature data and dynamic feature data of a sample special effect having a corresponding relationship, model data of a sample type terminal, and an actual running frame rate of the sample special effect on the sample type terminal;
[0093] The special effects frame rate prediction model is trained using the original training data set.
[0094] Furthermore, the training module is used to:
[0095] After obtaining the original training data set, the original test records in the original training data set are aggregated to obtain multiple test record groups; in the same test record group, the static feature data, dynamic feature data and model data of each original test record are the same;
[0096] Processing each of the test record groups into training data respectively; the training data includes static feature data and dynamic feature data of the sample special effects with corresponding relationships, model data of the sample type terminal, and target running frame rate; the target running frame rate is obtained according to the actual running frame rate of each of the original test records in the test record group;
[0097] The special effects frame rate prediction model is trained using the training data.
[0098] Furthermore, the target running frame rate is the actual running frame rate with the largest proportion in the ascending queue of actual running frame rates within a preset distribution range in the corresponding test record group.
[0099] Further, the sample special effects include a plurality of static feature data and a plurality of dynamic feature data, and the sample type terminal includes a plurality of model data;
[0100] Training modules for:
[0101] Before training the special effect frame rate prediction model using the training data, determining the importance level of each of the static feature data, each of the static feature data, and each of the model data;
[0102] Based on the importance levels of the static feature data, the dynamic feature data and the model data, the special effects frame rate prediction model is trained using the training data.
[0103] Further, the special effect frame rate prediction model includes a first prediction model and a second prediction model;
[0104] Training modules for:
[0105] The static feature data, the static feature data and the machine model data whose importance level is greater than or equal to the set level are used as target sample data;
[0106] Inputting all the static feature data, the dynamic feature data and the model data in the training data into the first prediction model to obtain a first frame rate prediction value;
[0107] Inputting the first frame rate prediction value and the target sample data into the second prediction model to obtain a second frame rate prediction value;
[0108] Based on the first frame rate prediction value and the target operating frame rate, parameters in the first prediction model are adjusted; and / or based on the second frame rate prediction value and the target operating frame rate, parameters in the second prediction model are adjusted.
[0109] The special effects frame rate prediction device provided in the embodiment of the present disclosure can execute the steps executed by the client or server in the special effects frame rate prediction method provided in the embodiment of the method of the present disclosure, and has execution steps and beneficial effects, which will not be repeated here.
[0110] Figure 6 Schematic diagram of the structure of an electronic device in the embodiment of the present disclosure. Figure 6 , which shows a schematic diagram of the structure of an electronic device 1000 suitable for implementing the embodiment of the present disclosure. The electronic device 1000 in the embodiment of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), wearable electronic devices, etc., and fixed terminals such as digital TVs, desktop computers, smart home devices, etc. Figure 6 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.
[0111] like Figure 6 As shown, the electronic device 1000 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 1001, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1008 to a random access memory (RAM) 1003 to implement the special effect frame rate prediction method of the embodiment described in the present disclosure. In the RAM 1003, various programs and information required for the operation of the electronic device 1000 are also stored. The processing device 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.
[0112] Typically, the following devices may be connected to the I / O interface 1005: an input device 1006 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 1007 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 1008 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 may allow the electronic device 1000 to communicate with other devices wirelessly or by wire to exchange information. Although Figure 6 The electronic device 1000 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.
[0113] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart, thereby implementing the special effects frame rate prediction method as described above. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 1009, or installed from the storage device 1008, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.
[0114] It should be noted that the computer-readable medium described above can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present disclosure, the computer-readable signal medium can include an information signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated information signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and the computer-readable signal medium can send, 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 using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.
[0115] In some embodiments, the client and the server can communicate using any known or future-developed network protocol such as HTTP (HyperText Transfer Protocol), and can be interconnected with digital information communication in any form or medium (e.g., a communication network). Examples of communication networks include local area networks ("LANs"), wide area networks ("WANs"), the Internet (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any known or future-developed networks.
[0116] The above computer-readable medium can be included in the above electronic device; or it can exist separately without being assembled into the electronic device.
[0117] The above computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device is caused to:
[0118] Obtain static feature data and dynamic feature data of the target special effect, as well as model data of the target type terminal;
[0119] The static feature data, the dynamic feature data and the model data are input into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
[0120] Optionally, when the above one or more programs are executed by the electronic device, the electronic device may also execute other steps described in the above embodiments.
[0121] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).
[0122] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0123] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, wherein the name of a unit does not, in some cases, constitute a limitation on the unit itself.
[0124] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.
[0125] In the context of the present disclosure, 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, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0126] According to one or more embodiments of the present disclosure, the present disclosure provides an electronic device, including:
[0127] one or more processors;
[0128] A memory for storing one or more programs;
[0129] When the one or more programs are executed by the one or more processors, the one or more processors implement any special effect frame rate prediction method provided in the present disclosure.
[0130] According to one or more embodiments of the present disclosure, the present disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the special effect frame rate prediction methods provided by the present disclosure.
[0131] The embodiments of the present disclosure also provide a computer program product, which includes a computer program or instructions. When the computer program or instructions are executed by a processor, the special effect frame rate prediction method as described above is implemented.
[0132] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0133] The above description is only a specific embodiment of the present disclosure, so that those skilled in the art can understand or implement the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A special effects frame rate prediction method, It is characterized in that include: Obtain static feature data and dynamic feature data of the target special effect, as well as model data of the target type terminal; The static feature data, the dynamic feature data and the model data are input into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
2. The method according to claim 1, It is characterized in that Also includes: If the predicted result of the running frame rate of the target special effect on the target type terminal is greater than or equal to the preset frame rate, it is determined that the target special effect meets the online condition of the target type terminal.
3. The method according to claim 1, It is characterized in that The special effect frame rate prediction model includes a first prediction model and a second prediction model; the target special effect includes a plurality of static feature data and a plurality of dynamic feature data, and the target type terminal includes a plurality of model data; The step of inputting the static feature data, the dynamic feature data, and the model data into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal includes: Determine target data from the plurality of static feature data, the plurality of dynamic feature data, and the plurality of machine model data, wherein the importance level of the target data is greater than or equal to a set level; Inputting the plurality of static feature data, the plurality of dynamic feature data and the plurality of aircraft model data into the first prediction model to obtain an intermediate prediction result; The intermediate prediction result and the target data are input into the second prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
4. The method according to claim 1, It is characterized in that The special effects frame rate prediction model training method comprises: Acquire an original training data set, wherein the original training data set includes a plurality of original test records; the original test records include static feature data and dynamic feature data of a sample special effect having a corresponding relationship, model data of a sample type terminal, and an actual running frame rate of the sample special effect on the sample type terminal; The special effects frame rate prediction model is trained using the original training data set.
5. The method according to claim 4, It is characterized in that After obtaining the original training data set, the method further includes: Aggregating the original test records in the original training data set to obtain a plurality of test record groups; in the same test record group, the static feature data, dynamic feature data and model data of each of the original test records are the same; Processing each of the test record groups into training data respectively; the training data includes static feature data and dynamic feature data of the sample special effects with corresponding relationships, model data of the sample type terminal, and target running frame rate; the target running frame rate is obtained according to the actual running frame rate of each of the original test records in the test record group; The using the original training data set to train the special effects frame rate prediction model includes: The special effects frame rate prediction model is trained using the training data.
6. The method according to claim 5, It is characterized in that The target running frame rate is the actual running frame rate with the largest proportion within a preset distribution range in the ascending queue of actual running frame rates in the corresponding test record group.
7. The method according to claim 5, It is characterized in that The sample special effect includes a plurality of static feature data and a plurality of dynamic feature data, and the sample type terminal includes a plurality of model data; before using the training data to train the special effect frame rate prediction model, the method further includes: Determine the importance level of each of the static feature data, each of the static feature data, and each of the model data; The using the training data to train the special effects frame rate prediction model includes: Based on the importance levels of the static feature data, the dynamic feature data and the model data, the special effects frame rate prediction model is trained using the training data.
8. The method according to claim 7, It is characterized in that The special effect frame rate prediction model includes a first prediction model and a second prediction model; The step of training the special effect frame rate prediction model based on the importance level of each of the static feature data, each of the dynamic feature data, and each of the model data using the training data includes: The static feature data, the static feature data and the machine model data whose importance level is greater than or equal to the set level are used as target sample data; Inputting all the static feature data, the dynamic feature data and the model data in the training data into the first prediction model to obtain a first frame rate prediction value; Inputting the first frame rate prediction value and the target sample data into the second prediction model to obtain a second frame rate prediction value; Based on the first frame rate prediction value and the target operating frame rate, parameters in the first prediction model are adjusted; and / or based on the second frame rate prediction value and the target operating frame rate, parameters in the second prediction model are adjusted.
9. A special effects frame rate prediction device, It is characterized in that include: An acquisition module, used to acquire static feature data and dynamic feature data of a target special effect, as well as model data of a target type terminal; The prediction module is used to input the static feature data, the dynamic feature data and the model data into a special effect frame rate prediction model to obtain a prediction result of the running frame rate of the target special effect on the target type terminal.
10. An electronic device, It is characterized in that The electronic device comprises: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 8.
11. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.