Judgment method and system for automatically editing video material based on AI

Abstract

The invention discloses a judgment method and system for automatically editing video materials based on AI, and relates to the technical field of artificial intelligence, and the judgment method comprises the following steps: obtaining multi-modal video data; generating an editing exception fragment set; the influence degree of the editing abnormal fragment set on the finished product editing quality is evaluated; and judging whether the abnormal editing fragment set conforms to a video editing standard or not, screening out materials conforming to the video editing standard, and generating a high-quality material library for editing for storage. According to the method, the influence of the abnormal clip on the finished product clip quality is evaluated, the quality of the materials can be accurately controlled in the clip process, bad materials are prevented from influencing the final effect, the materials meeting the video clip standard are automatically screened based on the evaluation result, and the high-quality material library is generated for storage, so that the subjectivity of manual selection of the materials is avoided, and the efficiency is improved. Uniformity and quality of the materials are guaranteed, and then a high-standard material library is provided for subsequent editing.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and system for determining automatic editing of video materials based on AI. Background Art

[0002] Video editing is a process of nonlinear editing of video sources and is part of multimedia production software. Through video editing software, users can perform various complex operations on videos, including cutting, merging, adjusting speed, adding special effects, color grading, dubbing, etc., to achieve the desired performance effect. The editing process is not just a simple deletion and splicing. It requires in-depth processing of video materials, removing unnecessary parts, and reorganizing and recreating according to creative needs. After the video is collected, it usually contains a lot of waste clips, which may be caused by shooting errors, repeated shots, blank screens or irrelevant content. The existence of waste clips seriously affects the fluency and expressiveness of the video. Therefore, the primary task of the editing software is to screen and delete these waste clips. Waste clips include but are not limited to redundant shots, wrong shooting content, meaningless blank screens, etc. These parts need to be cleaned up in the preliminary editing stage.

[0003] However, in the prior art, it is not convenient to quickly screen out invalid clips and abnormal materials in video materials, which increases the time and labor costs of video editing, and it is not convenient to compare the generated abnormal editing clips with the high-quality video material standards, and the quality of the final product cannot be ensured, so that the quality of the material cannot be accurately controlled during the editing process to avoid bad materials affecting the final effect. At the same time, it is not convenient to automatically screen materials that meet the video editing standards, thereby reducing the efficiency and quality of video editing and increasing the difficulty of manual operation.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes a judgment method and system for automatic editing of video materials based on AI, which solves the problem proposed in the above background technology that it is not convenient to quickly screen out invalid clips and abnormal materials in video materials, increases the time and manpower cost of video editing, and is not convenient to compare the generated abnormal editing clips with the high-quality video material standards, and cannot ensure the quality of the final product, so that the quality of the material cannot be accurately controlled during the editing process to avoid the influence of bad materials on the final effect. At the same time, it is not convenient to automatically screen materials that meet the video editing standards, thereby reducing the efficiency and quality of video editing and increasing the difficulty of manual operation.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: According to one aspect of the present invention, a method for determining whether a video material is automatically edited based on AI is provided, and the method comprises the following steps: S1. Obtaining video material data to be edited, and preprocessing the video material data to obtain multimodal video data; S2. Analyze the obtained multimodal video data based on an abnormality recognition algorithm, mark abnormal material data, and generate a set of abnormal clips; S3, comparing the generated abnormal editing segment set with known high-quality video material standards to evaluate the impact of the abnormal editing segment set on the quality of the finished editing; S4. Based on the evaluation results, determine whether the set of abnormal clips meets the video editing standard, select materials that meet the video editing standard, and generate a high-quality material library for editing for storage; Based on the anomaly recognition algorithm, the obtained multimodal video data is analyzed, and the abnormal material data is marked. The generation of a clip abnormality segment set includes the following steps: S21, quantizing the obtained multimodal video data, dividing the quantized multimodal video data into a training set and a test set, and setting a maximum number of iterations and a candidate parameter set size of the anomaly recognition algorithm; S22. Generate an initial parameter combination of the anomaly recognition algorithm using a random initialization strategy combined with the distribution characteristics of the feature space, and configure key parameters of the anomaly detection model; S23, dynamically adjusting the weight of the objective function of the anomaly detection model according to the changing rules of the abnormal features of the video clips, combining the temporal characteristics of the abnormal clips, and setting the feature priority; S24, using the anomaly detection model objective function, evaluating the recognition accuracy of the anomaly detection model for video data anomalies under each set of parameter combinations, obtaining a fitness value, and sorting the fitness values ​​of all parameter combinations to select the current optimal anomaly detection model parameters; S25. Based on the distribution of abnormal features, the nonlinear deviation of the abnormal detection model is calculated. If the nonlinear deviation is less than the threshold, a local optimization strategy is adopted to adjust the parameter combination of the abnormal detection model. Otherwise, a full local optimization strategy is adopted to adjust the direction of all parameter combinations of the abnormal detection model and adjust the weights of the auxiliary features. S26. Based on the parameter selection algorithm, compare the optimization result of the current anomaly detection model with the known optimal parameter combination, and select the anomaly detection model parameters with the best performance; S27, randomly perturbing the current optimal anomaly detection model parameters to generate new candidate anomaly detection model parameter combinations, optimizing the perturbed anomaly detection model parameters using a model optimization algorithm, and obtaining local optimal anomaly detection model parameters in the perturbed area; S28. If the local optimal anomaly detection model parameters are better than the current optimal anomaly detection model parameters, the current optimal anomaly detection model is recorded and the anomaly segment set is updated. Otherwise, return to step S22 to continue optimization. S29: output the optimal anomaly detection model parameters and the anomaly marking results of the anomaly detection model, and generate a set of clip anomaly segments according to the anomaly marking results of the anomaly detection model.

[0007] Furthermore, based on the parameter selection algorithm, the optimization result of the current anomaly detection model is compared with the known optimal parameter combination, and the selection of the anomaly detection model parameters with the best performance includes the following steps: S261. Define the key parameter search range and feature selection range of the anomaly detection model; S262, randomly generate a number of key parameter configurations and feature subsets, each combination of the key parameter configuration and the feature subset represents a candidate parameter configuration, and all combinations constitute a candidate parameter configuration group; S263, using each candidate parameter configuration to train the anomaly detection model, calculating its fitness value, and recording the performance result of the current candidate parameter configuration; S264, performing a neighborhood search near the key parameter configuration with the highest fitness in the candidate parameter configuration set, updating and recording the key parameter configuration and feature subset with the highest fitness; S265, distributing the remaining candidate parameter configurations to random key parameter configurations and feature subsets in the global scope, calculating their fitness values, and updating the global optimal key parameter configuration; S266. Repeat steps S263 to S265 until the maximum number of iterations is reached, and compare the current optimal key parameter configuration with the known optimal parameter configuration. If the current optimal key parameter configuration is better than the known optimal parameter configuration, replace it with the new optimal parameter configuration as the anomaly detection model parameter with the best performance.

[0008] Further, performing a neighborhood search near the key parameter configuration with the highest fitness in the candidate parameter configuration set, and updating and recording the key parameter configuration and feature subset with the highest fitness include the following steps: S2641. In the current candidate parameter configuration set, identify the key parameter configuration with the highest fitness value as the center of the neighborhood search; S2642, defining a neighborhood range according to a preset rule, taking the current optimal parameter configuration as the center, setting a disturbance amplitude for each dimension of the key parameter, and calculating a neighborhood separately for each key parameter configuration to form a multi-dimensional search space; S2643. Randomly generate a number of new key parameter configurations within the neighborhood, retrain the anomaly detection model using the new parameter configurations, and calculate the fitness value of each new key parameter configuration; S2644. Compare the fitness values ​​of all new key parameter configurations within the domain with the fitness value of the current optimal parameter configuration. If the fitness value of the new key parameter configuration is greater than the fitness value of the current optimal parameter configuration, the new key parameter configuration is used as the key parameter configuration with the highest fitness, and the feature subset corresponding to the key parameter configuration with the highest fitness is updated and recorded. Otherwise, the current optimal parameter configuration is kept unchanged.

[0009] Furthermore, the current optimal anomaly detection model parameters are randomly perturbed to generate new candidate anomaly detection model parameter combinations, and the perturbed anomaly detection model parameters are optimized using a model optimization algorithm. Obtaining the local optimal anomaly detection model parameters in the perturbed area includes the following steps: S271, initializing the parameters of the model optimization algorithm, and randomly selecting a number of candidate anomaly detection model parameter configurations in the disturbance area; S272, inputting the candidate anomaly detection model parameter configuration into the anomaly detection model, and calculating the fitness value of each candidate anomaly detection model parameter configuration through the objective function; S273, selecting the anomaly detection model parameter configuration with the highest fitness value from all candidate anomaly detection model parameter configurations as the current optimal anomaly detection model parameter solution; S274, updating the position of each anomaly detection model parameter configuration according to the position update formula; S275. If the fitness value of any candidate parameter is greater than the fitness value of the current optimal anomaly detection model parameter, the optimal anomaly detection model parameter scheme is updated, the anomaly detection model parameter configuration with the lowest fitness value is removed from the current candidate anomaly detection model parameter set, and a new anomaly detection model parameter configuration is randomly generated within the disturbance region of the current optimal anomaly detection model parameter. S276. Check whether the maximum number of iterations is reached. If reached, output the current optimal anomaly detection model parameter configuration as the local optimal anomaly detection model parameter of the disturbance area.

[0010] Furthermore, the position update formula is: ; In the formula, Y i ( s +1) indicates the i The anomaly detection model parameters are configured in s +1 position at iteration; rand Represents a random number; Y Pc represents the location of the optimal anomaly detection model parameter scheme within the disturbance region; PCRepresents the optimal anomaly detection model parameter scheme.

[0011] Furthermore, comparing the generated abnormal editing segment set with known high-quality video material standards to evaluate the impact of the abnormal editing segment set on the quality of the finished editing includes the following steps: S31, importing the generated abnormal editing segment set and known high-quality video material standards, setting evaluation parameters, and preprocessing the abnormal editing segment set and the high-quality video material; S32, screening out the editing features that have the greatest impact on the quality of the finished editing through a feature screening algorithm; S33, dividing the abnormal clip set into a training set and a test set; S34, randomly initializing the parameter configuration of the clip evaluation model, defining the parameter search space as a multi-dimensional space, and using each group of parameter configurations as a combination of evaluation criteria; S35, using a cross-validation method to evaluate the fitness value of each parameter configuration in the training set, recording the fitness value of the current parameter configuration, and marking the best fitness value among all parameter configurations; S36, using a heuristic algorithm to adjust the evaluation model parameter configuration and generate a new parameter configuration; S37, evaluating the training set using the new parameter configuration, calculating the fitness value of each clip abnormal segment set under the new parameter configuration; and updating the optimal fitness value and its corresponding parameter configuration among all parameter configurations; S38, check whether the maximum number of iterations has been reached, if so, proceed to step S39, otherwise, return to step S36, continue to adjust the evaluation model parameter configuration; output the evaluation model parameter configuration with the highest fitness value; S39. Use the optimal parameter configuration to evaluate the abnormal editing segments in the test set, quantify the matching degree between the abnormal editing segments and the high-quality standards, calculate the impact score of the editing abnormality segment set on the quality of the finished editing, and output the evaluation results.

[0012] Furthermore, the following steps are included to screen out the editing features that have the greatest impact on the quality of the finished editing through the feature screening algorithm: S321, initializing all parameters in the feature screening algorithm; S322, randomly selecting a feature combination as an initial feature candidate set, and initializing the feature weights using chaos thinking; S323, evaluating the influence of each candidate feature combination on the editing quality, calculating the fitness value of the candidate feature combination, and finding the feature combination with the best performance at present; S324, updating the weight of the current feature combination to generate a new candidate feature combination; S325, checking whether the updated feature combination meets the feature quantity threshold, if so, updating the feature combination according to the update rule, otherwise, keeping the current feature combination unchanged; S326, recalculate the fitness value of the updated feature combination, and compare it with the previous feature combination to find the feature combination with the highest fitness value; S327, comparing the fitness values ​​of the current feature combination and the historical optimal feature combination, and recording the feature combination with the highest fitness as the new optimal feature combination; S328, check whether the maximum number of iterations has been reached. If so, output the optimal feature combination as the editing feature that has the greatest impact on the quality of the finished editing. Otherwise, return to step S324 and continue iterative optimization.

[0013] Furthermore, the heuristic algorithm is used to adjust the evaluation model parameter configuration, and generating a new parameter configuration includes the following steps: S361, setting the parameters of the heuristic algorithm, randomly generating a parameter configuration population matrix, each row represents a candidate parameter configuration, and each column represents an evaluation model parameter; S362, evaluate the fitness value of each parameter configuration in turn, check whether the parameter configuration exceeds the upper limit and the lower limit, if exceeded, adjust the parameter configuration to the boundary value, otherwise, retain the current parameter configuration, and calculate the fitness value of each parameter configuration through the fitness function; S363, when the initial iteration is performed, the parameter configurations are sorted from small to large according to the fitness value, the sorted parameter configuration population is the optimal parameter configuration population, and the first parameter configuration in the optimal parameter configuration population is selected as the optimal parameter configuration. When the subsequent iteration is performed, the current optimal parameter configuration population is merged with the previous optimal parameter configuration population, and they are re-sorted according to the fitness value to select a new optimal parameter configuration population, and the first parameter configuration in the new optimal parameter configuration population is selected as the optimal parameter configuration; S364, updating the parameter configurations in the parameter configuration population one by one according to the updating rule of the heuristic algorithm; S365. Check whether the maximum number of iterations has been reached. If so, output the optimal parameter configuration and the corresponding fitness value as the new parameter configuration. Otherwise, return to step S362 and continue to execute until the maximum number of iterations has been reached.

[0014] Furthermore, the formula for updating the rule is: ; In the formula, A e Indicates e Parameter configuration; M n Indicates n Target parameter configuration;G e Indicates e The parameter configuration and n The difference between the target parameter configurations; b represents the constant that defines the shape of the logarithmic spiral; t Represents a random number; B represents the spiral function.

[0015] According to another aspect of the present invention, a judgment system for automatically editing video materials based on AI is also provided, and the judgment system includes: A data acquisition module is used to acquire video material data to be edited and pre-process the video material data to obtain multimodal video data; An anomaly recognition module is used to analyze the obtained multimodal video data based on an anomaly recognition algorithm, mark the abnormal material data, and generate a set of clip abnormality segments; A quality comparison module is used to compare the generated abnormal editing segment set with the known high-quality video material standard to evaluate the impact of the abnormal editing segment set on the quality of the finished editing; An evaluation and judgment module is used to judge whether the set of abnormal clips meets the video editing standard based on the evaluation results, screen out the materials that meet the video editing standard, and generate a high-quality material library for editing for storage; Among them, the data acquisition module is connected with the quality comparison module through the abnormality recognition module, and the quality comparison module is connected with the evaluation and judgment module.

[0016] The beneficial effects of the present invention are: 1. The present invention can quickly screen out invalid clips and abnormal materials in video materials through automated multimodal video data processing and abnormality recognition algorithms, greatly saving the time and labor costs of video editing. The generated abnormal editing clips are compared with the high-quality video material standards, which helps to ensure the quality of the final product. By evaluating the impact of abnormal editing clips on the editing quality of the finished product, the quality of the material can be accurately controlled during the editing process to avoid the impact of poor material on the final effect. Based on the evaluation results, the material that meets the video editing standards is automatically screened, and a high-quality material library is generated for storage, thereby avoiding the subjectivity of manual selection of materials, ensuring the uniformity and quality of the materials, and providing a high-standard material library for subsequent editing.

[0017] 2. The present invention uses an anomaly recognition algorithm to perform a detailed analysis of the video data, and adjusts the parameter combination of the anomaly detection model through multiple rounds of optimization, so that it can accurately identify abnormal segments in the video material, and ensure the quality of the video material by marking and removing these parts. By using a random initialization strategy and multiple iterative optimizations, the optimal anomaly detection model parameters are generated, making the screening of abnormal materials in the editing process more efficient and accurate. The final generated set of editing abnormal segments can automatically screen out irrelevant or low-quality materials, thereby avoiding omissions during manual inspection and reducing manual intervention.

[0018] 3. The present invention uses a feature screening algorithm to screen out the editing features that have the greatest impact on the quality of the finished editing, which can ensure that the most critical quality factors are focused on during the editing process. This not only improves the accuracy of the editing process, but also makes the evaluation results more targeted, thereby ensuring that the quality of the finished video meets the expected standards. By comparing with known high-quality video material standards, it can ensure that the video editing results meet industry standards and ensure the quality consistency of the finished video. This not only improves the professionalism of the editing, but also avoids artificial quality differences and inconsistencies, thereby ensuring the reliability of the finished video. BRIEF DESCRIPTION OF THE DRAWINGS

[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 will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 is a flow chart of a method for determining a video material based on AI automatic editing according to an embodiment of the present invention; Figure 2 It is a principle block diagram of a judgment system based on AI automatic editing of video materials according to an embodiment of the present invention.

[0021] In the figure: 1. Data acquisition module; 2. Abnormal identification module; 3. Quality comparison module; 4. Evaluation and judgment module. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0023] In the description of the present invention, unless otherwise specified, the meaning of "plurality" is two or more. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0024] According to an embodiment of the present invention, a method and system for determining automatic editing of video materials based on AI are provided.

[0025] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the judgment method based on AI automatic editing of video materials in an embodiment of the present invention, the judgment method includes the following steps: S1. Obtaining video material data to be edited, and preprocessing the video material data to obtain multimodal video data; Specifically, the video material data to be edited refers to the original video clips obtained from the video source or video library, including: Video frame data: Each frame of video material is usually stored in an image format (such as JPEG, PNG, etc.) and contains the visual information of the video.

[0026] Audio data: The audio track in the video (background music, dialogue, sound effects, etc.), stored in an audio format (such as WAV, MP3, AAC, etc.).

[0027] Subtitles / text data: Subtitles, captions, or other text information in a video, which may exist in the form of text files (such as SRT, VTT, etc.) or embedded in the video.

[0028] Metadata: Additional information associated with a video, such as timestamp, shooting location, camera settings, video resolution, frame rate, duration, etc.

[0029] Specifically, multimodal video data refers to the combination of multiple types of information (i.e., multiple "modalities") of video materials. This data includes not only traditional visual and auditory data, but also other types of data. Specifically, it includes: Visual information (image data): video frames, image features (such as color, texture, shape, etc.), objects and people in the image, etc. These features can be extracted through image recognition algorithms or deep learning.

[0030] Audio information: The audio data itself, including audio waveform, frequency information, audio features (such as pitch, rhythm, tone, etc.). The speech features of the audio (such as the speaker's emotions, language content, background noise, etc.) can be extracted through sound processing and analysis algorithms.

[0031] Text information: subtitles, speech-to-text, annotations in the video, or other related text documents. These text information can provide additional context to help more accurately analyze and understand the video content.

[0032] Sentiment and emotion analysis data: Analysis of emotional or sentiment information in videos, inference of emotional states based on audio (speech sentiment analysis) or visual information (facial expression analysis, gesture recognition, etc.), to help understand the emotional tone of the video.

[0033] Object and scene detection information in images or videos: Object detection algorithms (such as YOLO, Faster R-CNN, etc.) are used to analyze objects, scenes, or actions in video frames and extract specific content information, such as people, places, objects, activities, etc.

[0034] Time series data: The timing information of video frames and audio, the synchronization relationship on the timeline, and especially the analysis of editing, transitions, and rhythm.

[0035] Structured and unstructured metadata: Additional information related to the video is extracted through analysis of video materials, such as visual attributes such as shooting angle, focal length, light intensity, or meta-information such as the theme, content, genre, etc.

[0036] Special effects / transition data: Special effects data or shot switching information that the original video may contain, especially for video clips with special effects or animations.

[0037] Video editing information: If the material is obtained from an edited video, it may include pre-editing markers and editing point information, indicating which parts of the video material have been selected or removed.

[0038] S2. Analyze the obtained multimodal video data based on an abnormality recognition algorithm, mark abnormal material data, and generate a set of abnormal clips; Specifically, abnormal material data includes: abnormal visual data, abnormal time series data, irrelevant or low-quality clips, materials that do not meet editing standards, scene and character anomalies, abnormal audio data, technical issues, etc.

[0039] S3, comparing the generated abnormal editing segment set with known high-quality video material standards to evaluate the impact of the abnormal editing segment set on the quality of the finished editing; S4. Based on the evaluation results, determine whether the set of abnormal clips meets the video editing standard, select materials that meet the video editing standard, and generate a high-quality material library for editing for storage; It should be explained that the evaluation results of the collection of abnormal editing clips are compared with the known standards of high-quality video materials to screen out clips that meet the standards. For example, if a clip meets the standard requirements in terms of picture quality, sound quality, rhythm, emotional expression, etc., then the clip is considered qualified material; by comparing with the standards, those abnormal clips that fail to meet the editing standards are identified. These clips may be excluded due to poor quality, incoherent plots, unclear images, distorted sounds, etc.; according to the set editing standards (such as video clarity, sense of rhythm, effectiveness of information transmission, emotional expression, etc.), the matching degree of each clip is evaluated to screen out materials that meet these standards; after evaluation and comparison, the remaining clips that meet the editing standards will be screened as high-quality materials; the screened collection of materials that meet the standards will be stored in a special high-quality material library for subsequent video editing.

[0040] Based on the anomaly recognition algorithm, the obtained multimodal video data is analyzed, and the abnormal material data is marked. The generation of a clip abnormality segment set includes the following steps: S21, quantizing the obtained multimodal video data, dividing the quantized multimodal video data into a training set and a test set, and setting a maximum number of iterations and a candidate parameter set size of the anomaly recognition algorithm; S22. Generate an initial parameter combination of the anomaly recognition algorithm by using a random initialization strategy (i.e., convex lens imaging reverse learning) combined with the distribution characteristics of the feature space, and configure the key parameters of the anomaly detection model; S23, dynamically adjust the weight of the objective function of the anomaly detection model according to the changing rules of the abnormal features of the video clips (i.e., the changing mode of the nonlinear s-shaped features of the Sigmoid function), combine the temporal characteristics of the abnormal clips, and set the feature priority; S24, using the anomaly detection model objective function, evaluating the recognition accuracy of the anomaly detection model for video data anomalies under each set of parameter combinations, obtaining a fitness value, and sorting the fitness values ​​of all parameter combinations to select the current optimal anomaly detection model parameters; S25. Based on the distribution of abnormal features, calculate the nonlinear deviation of the abnormal detection model (i.e., the escape energy of the prey in the improved golden jackal optimization algorithm). If the nonlinear deviation is less than the threshold, adopt a local optimization strategy to adjust the combination of abnormal detection model parameters. Otherwise, adopt a full local optimization strategy to adjust the combination direction of all parameters of the abnormal detection model and adjust the weight of the auxiliary features. S26. Based on the parameter selection algorithm, compare the optimization result of the current anomaly detection model with the known optimal parameter combination, and select the anomaly detection model parameters with the best performance; S27, randomly perturbing the current optimal anomaly detection model parameters to generate new candidate anomaly detection model parameter combinations, optimizing the perturbed anomaly detection model parameters using a model optimization algorithm, and obtaining local optimal anomaly detection model parameters in the perturbed area; S28. If the local optimal anomaly detection model parameters are better than the current optimal anomaly detection model parameters, the current optimal anomaly detection model is recorded and the anomaly segment set is updated. Otherwise, return to step S22 to continue optimization. S29: output the optimal anomaly detection model parameters and the anomaly marking results of the anomaly detection model, and generate a set of clip anomaly segments according to the anomaly marking results of the anomaly detection model.

[0041] Specifically, the anomaly recognition algorithm is an improved golden jackal optimization algorithm. First, the convex lens imaging reverse learning strategy is integrated to explore a better initial solution position; the Sigmoid function is introduced to change the escape energy and balance the exploration and development stages; the spiral encirclement mechanism of the whale optimization algorithm is integrated to enhance the exploration ability and improve the convergence accuracy. In the traditional golden jackal optimization algorithm, in the golden jackal combination, the male golden jackal is in the first leader position, which is the optimal exploration position; the female golden jackal is a follower and is in a suboptimal position. In the golden jackal optimization algorithm, the hunting process of the golden jackal is divided into three stages: (1) searching for prey and approaching the prey; (2) encircling the prey and stimulating the prey until it stops moving; (3) attacking the prey.

[0042] Specifically, in the basic convex lens imaging reverse learning, the scale factor is generally used as a fixed value, which is easy to fall into the local optimum. The present invention proposes a scale factor strategy based on dynamic decreasing, which can obtain a larger value in the early iteration, so that the golden jackal can search a large range in different dimensional areas and improve the diversity of the group; obtain a smaller value in the later iteration, and conduct a fine search near the optimal individual to improve the local optimization ability.

[0043] Specifically, the escape energy of the prey is the decisive factor controlling the update of the prey's position, which is responsible for balancing the global search and local development behavior of the algorithm. The nonlinear s-shaped characteristic change of the Sigmoid function can continuously perform global search to find the optimal area, thereby improving the disadvantage of slow convergence speed in the early stage and increasing the convergence speed.

[0044] In order to facilitate understanding of the above technical solution of the present invention, the present invention will be described in detail below in the actual process based on an abnormality recognition algorithm to analyze the obtained multimodal video data, mark abnormal material data, and generate a set of edited abnormal segments.

[0045] Step 1: Quantization and data division of multimodal video data 1) Multimodal video data acquisition: Assume that a set of multimodal video data from a video editing system has been obtained, including visual data (video frames) and audio data (sound waveforms).

[0046] Visual data: The image size of each frame of the video is 1280×720, the total length of the video is 100 seconds, the frame rate is 30fps, and there are 3000 frames in total.

[0047] Audio data: Assume that the audio sampling rate is 44.1kHz, 44,100 sampling points per second, the total duration of the audio data is 100 seconds, and the total number of sampling points is 4,410,000.

[0048] 2) Quantification For visual data, each frame of the image is converted into a grayscale image, and edge detection (such as the Canny algorithm) is performed to quantify the edge features of each frame of the image.

[0049] Perform spectral analysis on audio data to extract spectral features of the audio, such as MFCC (Mel Frequency Cepstral Coefficient).

[0050] 3) Data division: All data are divided into a training set (80%) and a test set (20%). That is, the training set contains 2400 frames of video and corresponding audio data, and the test set contains 600 frames of video and corresponding audio data.

[0051] 4) Set parameters Maximum number of iterations: set to 100 iterations.

[0052] Candidate parameter set size: set to 10 candidate parameter sets.

[0053] Step 2: Initialize the anomaly recognition algorithm parameters 1) Random initialization: Use a random initialization strategy to generate an initial parameter combination from the distribution characteristics of the feature space. Assume that a support vector machine (SVM)-based anomaly detection model is used, and its key parameters include: C (Penalty parameter): The initial value is 1.0; c (Kernel function parameters): The initial value is 0.01.

[0054] Randomly generate 10 sets of candidate parameter combinations, which are: Parameter combination 1: C =1.0, c =0.01.

[0055] Parameter combination 2: C =2.0, c =0.05.

[0056] Parameter combination 3:C =0.5, c =0.02.

[0057] And so on, 10 candidate combinations are generated.

[0058] 2) Configure the model: Apply these initial parameter combinations to the anomaly detection model.

[0059] Step 3: Dynamically adjust objective function weights and feature priorities 1) Objective function weight adjustment: According to the changing rules of clip abnormality features, dynamically adjust the weight of the objective function in the anomaly detection model. Assume that the importance of visual features (image clarity, edge detection) is higher, and the importance of audio features (spectral features) is lower. Therefore, the weight can be adjusted as follows: Visual feature weight: 0.7.

[0060] Audio feature weight: 0.3.

[0061] 2) Temporal feature merging: Combined with the temporal characteristics of the clips, a time window parameter can be added to model the continuity between video frames and increase the priority of temporal features. Assume that the time window size is 10 frames.

[0062] Step 4: Evaluate model fitness value Fitness evaluation: Based on the training set data, evaluate the recognition accuracy of each parameter combination. Suppose we use the cross-validation method to evaluate the accuracy of the model, and the results are as follows: Parameter combination 1: 95% accuracy.

[0063] Parameter combination 2: 92% accuracy.

[0064] Parameter combination 3: 89% accuracy.

[0065] Similarly, after the fitness values ​​of all parameter combinations are sorted, the optimal parameter combination is selected.

[0066] Step 5: Nonlinear deviation calculation and optimization strategy selection 1) Nonlinear deviation calculation: Based on the distribution of abnormal features, the nonlinear deviation is calculated. Assume that the calculation result is: deviation value = 0.05, which is lower than the set threshold (0.1).

[0067] 2) Local optimization strategy: Since the deviation is less than the threshold, a local optimization strategy is used to further adjust the parameter combination of the anomaly detection model. Assume that after local optimization, the following updated parameter combination is obtained: C =1.2, c =0.03.

[0068] Step 6: Compare with the optimal parameter combination 1) Optimal parameter comparison: Compare the current optimal parameters with the known optimal parameter combination. Assume that the optimal parameter combination is known from historical data: C =1.0, c =0.02.

[0069] Current optimal parameter combination C =1.2, c =0.03 has a small improvement in accuracy, so the current parameter combination is selected.

[0070] Step 7: Parameter perturbation and local optimum acquisition 1) Parameter perturbation: Perturb the parameters of the current optimal anomaly detection model to generate new candidate parameter combinations. For example: Parameter perturbation 1: C =1.0, c =0.04 Parameter perturbation 2: C =1.3, c =0.01.

[0071] 2) Local optimal calculation: Use the optimization algorithm to optimize the parameters after disturbance. Assume that the parameter combination obtained by local optimization is: C =1.1, c =0.025, the accuracy rate is increased to 96%.

[0072] Step 8: Local Optimal Parameter Update 1) Update the optimal model: Since the accuracy of the local optimal parameter combination is better than the current optimal model, update the optimal model and record the current optimal parameter combination: C =1.1, c =0.025.

[0073] 2) Update the abnormal clip set: According to the current optimal parameters, apply the anomaly detection model to analyze the video data and generate a new set of abnormal clips. Assume that the model marks 200 abnormal clips (such as blurred images, excessive background noise, editing errors, etc.).

[0074] Step 9: Output the optimal model parameters and anomaly marking results 1) Output the optimal parameter combination: C =1.1, c =0.025.

[0075] 2) Output abnormal marking results: Marking abnormal segments: 200 segments (e.g., numbered 105, 205, 405, etc.) were identified as abnormal clip segments.

[0076] 3) Generate a set of clip anomaly segments: Based on the model labeling results, generate a set of clip anomaly segments, which includes all segments marked as abnormal, for further processing (such as removal or replacement).

[0077] Preferably, based on the parameter selection algorithm, the optimization result of the current anomaly detection model is compared with the known optimal parameter combination, and the selection of the anomaly detection model parameters with the best performance includes the following steps: S261. Define the key parameter search range and feature selection range of the anomaly detection model; S262, randomly generate a number of key parameter configurations and feature subsets, each combination of the key parameter configuration and the feature subset represents a candidate parameter configuration, and all combinations constitute a candidate parameter configuration group; S263, using each candidate parameter configuration to train the anomaly detection model, calculating its fitness value, and recording the performance result of the current candidate parameter configuration; S264, performing a neighborhood search near the key parameter configuration with the highest fitness in the candidate parameter configuration set, updating and recording the key parameter configuration and feature subset with the highest fitness; S265, distributing the remaining candidate parameter configurations to random key parameter configurations and feature subsets in the global scope, calculating their fitness values, and updating the global optimal key parameter configuration; S266. Repeat steps S263 to S265 until the maximum number of iterations is reached, and compare the current optimal key parameter configuration with the known optimal parameter configuration. If the current optimal key parameter configuration is better than the known optimal parameter configuration, replace it with the new optimal parameter configuration as the anomaly detection model parameter with the best performance.

[0078] Specifically, the parameter selection algorithm is the bee algorithm, which is a new cluster intelligent search algorithm that finds the optimal solution to the problem by simulating the foraging behavior of bees in nature. An optimization algorithm proposed. In the algorithm, the location of the food source represents a set of solutions to the optimization problem. The location of the food source and its immediate neighborhood form a small search area, called a flower cluster. In the process of searching for food, the bee colony first sends out scout bees to search. The scout bees continuously fly from one food source to another to search for all possible food sources. When the profitability of the searched food source (such as the sugar content) exceeds a certain amount, the scout bee returns to the hive, unloads the honey and dances a magical "swing dance" on the "dance board". This "swing dance" is a necessary tool for the bee colony to communicate information, which enables the entire bee colony to understand the location of the discovered food source, the distance to the hive and the profitability. After dancing, the scout bee leads the follower bees waiting outside the hive to fly back to the corresponding food source to collect honey, and the food source with high profitability will recruit more follower bees to collect honey. During the honey-gathering process, the bees calculate the profitability of the food source they are collecting honey from, and return to the hive to perform a "swing dance" to convey the information of the current food source profitability to the bee colony. If the food source profitability is still high, more follower bees will be recruited to collect honey; if the food source profitability is low, the current food source will be abandoned.

[0079] Preferably, performing a neighborhood search near the key parameter configuration with the highest current fitness in the candidate parameter configuration set, and updating and recording the key parameter configuration and feature subset with the highest fitness comprises the following steps: S2641. In the current candidate parameter configuration set, identify the key parameter configuration with the highest fitness value as the center of the neighborhood search; S2642, defining a neighborhood range (i.e., nearest neighbor search) according to preset rules, taking the current optimal parameter configuration as the center, setting a disturbance amplitude for each dimension of the key parameter, and calculating a neighborhood separately for each key parameter configuration to form a multi-dimensional search space; S2643. Randomly generate a number of new key parameter configurations within the neighborhood, retrain the anomaly detection model using the new parameter configurations, and calculate the fitness value of each new key parameter configuration; S2644. Compare the fitness values ​​of all new key parameter configurations within the domain with the fitness value of the current optimal parameter configuration. If the fitness value of the new key parameter configuration is greater than the fitness value of the current optimal parameter configuration, the new key parameter configuration is used as the key parameter configuration with the highest fitness, and the feature subset corresponding to the key parameter configuration with the highest fitness is updated and recorded. Otherwise, the current optimal parameter configuration is kept unchanged.

[0080] Preferably, randomly perturbing the current optimal anomaly detection model parameters to generate new candidate anomaly detection model parameter combinations, optimizing the perturbed anomaly detection model parameters using a model optimization algorithm, and obtaining the local optimal anomaly detection model parameters in the perturbed area includes the following steps: S271, initializing the parameters of the model optimization algorithm, and randomly selecting a number of candidate anomaly detection model parameter configurations in the disturbance area; S272, inputting the candidate anomaly detection model parameter (i.e., each star in the black hole optimization algorithm) configuration into the anomaly detection model, and calculating the fitness value of each candidate anomaly detection model parameter configuration through the objective function; S273, selecting the anomaly detection model parameter configuration with the highest fitness value from all candidate anomaly detection model parameter configurations as the current optimal anomaly detection model parameter solution (i.e., the black hole in the black hole optimization algorithm); S274, updating the position of each anomaly detection model parameter configuration according to the position update formula; S275. If the fitness value of any candidate parameter is greater than the fitness value of the current optimal anomaly detection model parameter, the optimal anomaly detection model parameter scheme is updated, the anomaly detection model parameter configuration with the lowest fitness value is removed from the current candidate anomaly detection model parameter set, and a new anomaly detection model parameter configuration is randomly generated within the disturbance region of the current optimal anomaly detection model parameter. S276. Check whether the maximum number of iterations is reached. If reached, output the current optimal anomaly detection model parameter configuration as the local optimal anomaly detection model parameter of the disturbance area.

[0081] Specifically, the model optimization algorithm is a black hole optimization algorithm, and the idea of ​​the algorithm relatively completely describes the general characteristics of the black hole phenomenon in nature; based on these characteristics of the black hole optimization algorithm, the optimization search principle of the black hole optimization algorithm is relatively simple and easy to implement. The black hole optimization algorithm mainly simulates the actual black hole phenomenon, randomly arranges a certain number of stars in a certain search space, and determines and evaluates the fitness function of each star in the search space through statistical means, and selects a star with the best fitness value as the black hole. The boundary of this black hole is regarded as the area where the current global optimal solution is located, and the black hole itself is regarded as the current global optimal solution.

[0082] Specifically, the anomaly detection model is a kernel extreme learning machine (KELM), which is a new type of artificial intelligence prediction model. It introduces the kernel function idea on the basis of the extreme learning machine (ELM), thereby effectively overcoming the problem of low output stability caused by random generation of initial weights and thresholds in ELM, and improving the output stability. However, its prediction performance is usually affected by parameter selection. Therefore, the present invention uses variational mode decomposition to decompose the actual time series signal, and then adds gray correlation analysis to perform correlation analysis on the input and output variables of the variable pitch coordinated control model to increase the correlation of the data. Based on the KLEM model, a model optimization algorithm (i.e., black hole optimization algorithm) is used to optimize its parameters. Variational mode decomposition (VMD) is a signal processing technology that is widely used to analyze non-stationary and nonlinear signals. It decomposes the signal into a set of intrinsic mode functions with different center frequencies through a variational method. These IMFs represent different frequency components in the signal and can be used for feature extraction, predictive modeling, and pattern recognition.

[0083] Preferably, the location update formula is: ; In the formula, Y i ( s +1) indicates the i The anomaly detection model parameters are configured in s +1 position at iteration; rand Represents a random number between [0, 1]; Y Pc represents the location of the optimal anomaly detection model parameter scheme within the disturbance region; PC Represents the optimal anomaly detection model parameter scheme.

[0084] Preferably, comparing the generated abnormal editing segment set with a known high-quality video material standard to evaluate the influence of the abnormal editing segment set on the quality of the finished editing includes the following steps: S31, importing the generated abnormal editing segment set and known high-quality video material standards, setting evaluation parameters, and preprocessing the abnormal editing segment set and the high-quality video material; S32, screening out the editing features that have the greatest impact on the quality of the finished editing through a feature screening algorithm; S33, dividing the abnormal clip set into a training set and a test set; S34, randomly initializing the parameter configuration of the clip evaluation model, defining the parameter search space as a multi-dimensional space, and using each group of parameter configurations as a combination of evaluation criteria; S35, using a cross-validation method to evaluate the fitness value of each parameter configuration in the training set, recording the fitness value of the current parameter configuration, and marking the best fitness value among all parameter configurations; S36, using a heuristic algorithm to adjust the evaluation model parameter configuration and generate a new parameter configuration; S37, evaluating the training set using the new parameter configuration, calculating the fitness value of each clip abnormal segment set under the new parameter configuration; and updating the optimal fitness value and its corresponding parameter configuration among all parameter configurations; S38, check whether the maximum number of iterations has been reached, if so, proceed to step S39, otherwise, return to step S36, continue to adjust the evaluation model parameter configuration; output the evaluation model parameter configuration with the highest fitness value; S39. Use the optimal parameter configuration to evaluate the abnormal editing segments in the test set, quantify the matching degree between the abnormal editing segments and the high-quality standards, calculate the impact score of the editing abnormality segment set on the quality of the finished editing, and output the evaluation results.

[0085] It needs to be explained that the relevant feature data of abnormal clips are collected, including visual, audio and timing features, such as picture clarity, background noise intensity, shot switching smoothness, etc.; the relevant feature data of abnormal clips are collected, including visual, audio and timing features, such as picture clarity, background noise intensity, shot switching smoothness, etc.; the weight distribution and evaluation range of key features are defined; the collection of abnormal editing clips and high-quality video materials are normalized and feature extracted for unified format analysis; the feature screening algorithm is used to extract the core features that have the greatest impact on the quality of the finished editing; according to the algorithm results, the most important features in visual, audio and timing features are screened out as the basis for subsequent analysis; the collection of abnormal editing clips is divided into training set and test set; the training set is used for model optimization, and the test set is used to verify the effect of the final evaluation model; select a suitable editing evaluation model, such as a model based on machine learning or statistical analysis; define the search space of model parameters, including the range of feature weights and the specific hyperparameters of the model; randomly generate initial parameter configurations to provide a basis for subsequent optimization; evaluate the fitness value of each set of parameter configurations on the training set to measure the performance of the editing evaluation model for abnormal fragment features; record the current fitness value and select the parameter configuration with the best performance; use heuristic algorithms (such as genetic algorithms, particle swarm optimization, crow search algorithms, etc.) to generate new parameter combinations based on the current parameter configurations; the parameter configurations will be further evaluated in subsequent steps; use the newly generated parameter configurations to evaluate the training set and calculate the fitness value of each set of parameters; if the fitness value of the new parameter configuration is better than the previous optimal value, update the optimal configuration; check whether the number of iterations has reached the maximum value; if not, return to the parameter adjustment step to continue optimization; if reached, end the iteration and lock the current optimal parameter configuration; use the optimal parameter configuration to evaluate the abnormal clips in the test set; quantify the matching degree of the abnormal clips with the high-quality video material standards, and output the score of the finished clip quality and the final evaluation report.

[0086] Specifically, the present invention compares the generated abnormal editing clip set with the known high-quality video material standard by improving the longhorn beetle group algorithm to evaluate the degree of influence of the abnormal editing clip set on the quality of the finished editing. The traditional longhorn beetle group algorithm only considers the behavior of individuals learning from the global extreme value and the local extreme value when updating the speed, and ignores the judgment made by the longhorn beetle individuals on the surrounding environment. The present invention comprehensively considers various influencing factors when updating the longhorn beetle speed. In order to better coordinate global and local searches, a feature screening algorithm is introduced for feature screening.

[0087] Preferably, screening out the editing features that have the greatest impact on the quality of the finished editing by a feature screening algorithm comprises the following steps: S321, initializing all parameters in the feature screening algorithm; S322, randomly selecting a feature combination as an initial feature candidate set, and initializing the feature weights using chaos thinking; S323, evaluating the influence of each candidate feature combination on the editing quality, calculating the fitness value of the candidate feature combination, and finding the feature combination with the best performance at present; S324, updating the weight of the current feature combination to generate a new candidate feature combination; S325, checking whether the updated feature combination meets the feature quantity threshold, if so, updating the feature combination according to the update rule, otherwise, keeping the current feature combination unchanged; S326, recalculate the fitness value of the updated feature combination, and compare it with the previous feature combination to find the feature combination with the highest fitness value; S327, comparing the fitness values ​​of the current feature combination and the historical optimal feature combination, and recording the feature combination with the highest fitness as the new optimal feature combination; S328, check whether the maximum number of iterations has been reached. If so, output the optimal feature combination as the editing feature that has the greatest impact on the quality of the finished editing. Otherwise, return to step S324 and continue iterative optimization.

[0088] Specifically, the feature screening algorithm is an improved crow search algorithm, which is a group intelligence optimization algorithm inspired by the memory behavior and food hiding strategy of crows. Crows search for food in groups, remember the location of food and prevent other crows from stealing food. CSA solves complex optimization problems by simulating these behaviors of crows. When initializing the crow population, the present invention selects cubic chaotic mapping to produce a more evenly distributed population, thereby increasing the diversity of the population. Compared with the Logistic mapping, the cubic mapping has a simpler structure. The most important thing is that a more evenly distributed population can be obtained, and the range of activities of the crows in the entire iterative optimization process will also become larger.

[0089] Preferably, adjusting the evaluation model parameter configuration using a heuristic algorithm to generate a new parameter configuration includes the following steps: S361, setting the parameters of the heuristic algorithm, randomly generating a parameter configuration population matrix, each row represents a candidate parameter configuration, and each column represents an evaluation model parameter; S362, evaluate the fitness value of each parameter configuration in turn, check whether the parameter configuration exceeds the upper limit and the lower limit, if exceeded, adjust the parameter configuration to the boundary value, otherwise, retain the current parameter configuration, and calculate the fitness value of each parameter configuration through the fitness function; S363, when the initial iteration is performed, the parameter configurations are sorted from small to large according to the fitness value, the sorted parameter configuration population is the optimal parameter configuration population, and the first parameter configuration in the optimal parameter configuration population is selected as the optimal parameter configuration. When the subsequent iteration is performed, the current optimal parameter configuration population is merged with the previous optimal parameter configuration population, and they are re-sorted according to the fitness value to select a new optimal parameter configuration population, and the first parameter configuration in the new optimal parameter configuration population is selected as the optimal parameter configuration; S364, updating the parameter configurations in the parameter configuration population one by one according to the updating rule of the heuristic algorithm; S365. Check whether the maximum number of iterations has been reached. If so, output the optimal parameter configuration and the corresponding fitness value as the new parameter configuration. Otherwise, return to step S362 and continue to execute until the maximum number of iterations has been reached.

[0090] Specifically, the heuristic algorithm is an improved moth-to-fire algorithm. The main inspiration of the traditional moth-to-fire algorithm comes from the lateral navigation flight mechanism of moths. When flying at night, moths maintain a fixed angle to the moon. This behavior allows them to always fly in a straight line when flying long distances. In daily life, moths mistake artificial light for moonlight, so some moths move in a spiral curve around artificial light. In order to improve the performance of the traditional moth-to-fire algorithm, the present invention proposes some improvement measures. After one iteration of the moth-to-fire algorithm is completed, the simplex search strategy is used to select several moth positions with poor positions for optimization. The simplex search method is an unconstrained optimization method that was developed relatively early. The basic idea of ​​this algorithm is similar to blind people climbing a mountain. Each time you move forward, you first push the stick forward to test a few points, and then move forward when you find the highest point.

[0091] Preferably, the formula for the update rule is: ; In the formula, A e Indicates e parameter configuration (i.e. the moth in the moth-to-fire algorithm); M n Indicates n target parameter configuration (i.e. the flame in the moth-to-fire algorithm); G e Indicates e The parameter configuration and n The difference between the target parameter configurations (i.e., the distance from the moth to the flame in the moth-to-fire algorithm); b represents a constant that defines the shape of a logarithmic spiral; t Represents a random number between [0, 1]; B represents the spiral function.

[0092] According to another embodiment of the present invention, Figure 2 As shown, a judgment system based on AI automatic editing of video materials is also provided, and the judgment system includes: The data acquisition module 1 is used to acquire the video material data to be edited and pre-process the video material data to obtain multi-modal video data; The abnormality recognition module 2 is used to analyze the obtained multimodal video data based on the abnormality recognition algorithm, mark the abnormal material data, and generate a set of clip abnormality segments; The quality comparison module 3 is used to compare the generated abnormal editing segment set with the known high-quality video material standard to evaluate the influence of the abnormal editing segment set on the quality of the finished editing; Evaluation and judgment module 4, used to judge whether the set of abnormal clips meets the video editing standard based on the evaluation result, screen out the materials that meet the video editing standard, and generate a high-quality material library for editing for storage; The data acquisition module 1 is connected to the quality comparison module 3 via the abnormality recognition module 2 , and the quality comparison module 3 is connected to the evaluation and judgment module 4 .

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for determining video material based on AI automatic editing, characterized in that: The determination method comprises the following steps: S1. Obtaining video material data to be edited, and preprocessing the video material data to obtain multimodal video data; S2. Analyze the obtained multimodal video data based on an abnormality recognition algorithm, mark abnormal material data, and generate a set of abnormal clips; S3, comparing the generated abnormal editing segment set with known high-quality video material standards to evaluate the impact of the abnormal editing segment set on the quality of the finished editing; S4. Based on the evaluation results, determine whether the set of abnormal clips meets the video editing standard, select materials that meet the video editing standard, and generate a high-quality material library for editing for storage; The method of analyzing the obtained multimodal video data based on the abnormality recognition algorithm, marking the abnormal material data, and generating a set of clip abnormality segments includes the following steps: S21, quantizing the obtained multimodal video data, dividing the quantized multimodal video data into a training set and a test set, and setting a maximum number of iterations and a candidate parameter set size of the anomaly recognition algorithm; S22. Generate an initial parameter combination of the anomaly recognition algorithm using a random initialization strategy combined with the distribution characteristics of the feature space, and configure key parameters of the anomaly detection model; S23, dynamically adjusting the weight of the objective function of the anomaly detection model according to the changing rules of the abnormal features of the video clips, combining the temporal characteristics of the abnormal clips, and setting the feature priority; S24, using the anomaly detection model objective function, evaluating the recognition accuracy of the anomaly detection model for video data anomalies under each set of parameter combinations, obtaining a fitness value, and sorting the fitness values ​​of all parameter combinations to select the current optimal anomaly detection model parameters; S25. Based on the distribution of abnormal features, the nonlinear deviation of the abnormal detection model is calculated. If the nonlinear deviation is less than the threshold, a local optimization strategy is adopted to adjust the parameter combination of the abnormal detection model. Otherwise, a full local optimization strategy is adopted to adjust the direction of all parameter combinations of the abnormal detection model and adjust the weights of the auxiliary features. S26. Based on the parameter selection algorithm, compare the optimization result of the current anomaly detection model with the known optimal parameter combination, and select the anomaly detection model parameters with the best performance; S27, randomly perturbing the current optimal anomaly detection model parameters to generate new candidate anomaly detection model parameter combinations, optimizing the perturbed anomaly detection model parameters using a model optimization algorithm, and obtaining local optimal anomaly detection model parameters in the perturbed area; S28. If the local optimal anomaly detection model parameters are better than the current optimal anomaly detection model parameters, the current optimal anomaly detection model is recorded and the anomaly segment set is updated. Otherwise, return to step S22 to continue optimization. S29: output the optimal anomaly detection model parameters and the anomaly marking results of the anomaly detection model, and generate a set of clip anomaly segments according to the anomaly marking results of the anomaly detection model.

2. The method for determining the video material based on AI automatic editing according to claim 1, characterized in that: The method of comparing the optimization result of the current anomaly detection model with the known optimal parameter combination based on the parameter selection algorithm and selecting the anomaly detection model parameters with the best performance includes the following steps: S261. Define the key parameter search range and feature selection range of the anomaly detection model; S262, randomly generate a number of key parameter configurations and feature subsets, each combination of the key parameter configuration and the feature subset represents a candidate parameter configuration, and all combinations constitute a candidate parameter configuration group; S263, using each candidate parameter configuration to train the anomaly detection model, calculating its fitness value, and recording the performance result of the current candidate parameter configuration; S264, performing a neighborhood search near the key parameter configuration with the highest fitness in the candidate parameter configuration set, updating and recording the key parameter configuration and feature subset with the highest fitness; S265, distributing the remaining candidate parameter configurations to random key parameter configurations and feature subsets in the global scope, calculating their fitness values, and updating the global optimal key parameter configuration; S266. Repeat steps S263 to S265 until the maximum number of iterations is reached, and compare the current optimal key parameter configuration with the known optimal parameter configuration. If the current optimal key parameter configuration is better than the known optimal parameter configuration, replace it with the new optimal parameter configuration as the anomaly detection model parameter with the best performance.

3. The method for determining the video material based on AI automatic editing according to claim 2, characterized in that: The performing of neighborhood search near the key parameter configuration with the highest current fitness in the candidate parameter configuration set, and updating and recording the key parameter configuration and feature subset with the highest fitness comprises the following steps: S2641. In the current candidate parameter configuration set, identify the key parameter configuration with the highest fitness value as the center of the neighborhood search; S2642, defining a neighborhood range according to a preset rule, taking the current optimal parameter configuration as the center, setting a disturbance amplitude for each dimension of the key parameter, and calculating a neighborhood separately for each key parameter configuration to form a multi-dimensional search space; S2643. Randomly generate a number of new key parameter configurations within the neighborhood, retrain the anomaly detection model using the new parameter configurations, and calculate the fitness value of each new key parameter configuration; S2644. Compare the fitness values ​​of all new key parameter configurations within the domain with the fitness value of the current optimal parameter configuration. If the fitness value of the new key parameter configuration is greater than the fitness value of the current optimal parameter configuration, the new key parameter configuration is used as the key parameter configuration with the highest fitness, and the feature subset corresponding to the key parameter configuration with the highest fitness is updated and recorded. Otherwise, the current optimal parameter configuration is kept unchanged.

4. The method for determining the video material based on AI automatic editing according to claim 1, characterized in that: The method of randomly perturbing the current optimal anomaly detection model parameters to generate a new candidate anomaly detection model parameter combination, optimizing the perturbed anomaly detection model parameters using a model optimization algorithm, and obtaining the local optimal anomaly detection model parameters in the perturbed area comprises the following steps: S271, initializing the parameters of the model optimization algorithm, and randomly selecting a number of candidate anomaly detection model parameter configurations in the disturbance area; S272, inputting the candidate anomaly detection model parameter configuration into the anomaly detection model, and calculating the fitness value of each candidate anomaly detection model parameter configuration through the objective function; S273, selecting the anomaly detection model parameter configuration with the highest fitness value from all candidate anomaly detection model parameter configurations as the current optimal anomaly detection model parameter solution; S274, updating the position of each anomaly detection model parameter configuration according to the position update formula; S275. If the fitness value of any candidate parameter is greater than the fitness value of the current optimal anomaly detection model parameter, the optimal anomaly detection model parameter scheme is updated, the anomaly detection model parameter configuration with the lowest fitness value is removed from the current candidate anomaly detection model parameter set, and a new anomaly detection model parameter configuration is randomly generated within the disturbance region of the current optimal anomaly detection model parameter. S276. Check whether the maximum number of iterations is reached. If reached, output the current optimal anomaly detection model parameter configuration as the local optimal anomaly detection model parameter of the disturbance area.

5. The method for determining video material based on AI automatic editing according to claim 4, characterized in that: The position update formula is: ； In the formula, Y i ( s +1) indicates the i The anomaly detection model parameters are configured in s +1 position at iteration; rand Represents a random number; Y Pc represents the location of the optimal anomaly detection model parameter scheme within the disturbance region; Pc Represents the optimal anomaly detection model parameter scheme.

6. The method for determining video material based on AI automatic editing according to claim 1, characterized in that: The step of comparing the generated abnormal editing segment set with a known high-quality video material standard to evaluate the influence of the abnormal editing segment set on the quality of the finished editing includes the following steps: S31, importing the generated abnormal editing segment set and known high-quality video material standards, setting evaluation parameters, and preprocessing the abnormal editing segment set and the high-quality video material; S32, screening out the editing features that have the greatest impact on the quality of the finished editing through a feature screening algorithm; S33, dividing the abnormal clip set into a training set and a test set; S34, randomly initializing the parameter configuration of the clip evaluation model, defining the parameter search space as a multi-dimensional space, and using each group of parameter configurations as a combination of evaluation criteria; S35, using a cross-validation method to evaluate the fitness value of each parameter configuration in the training set, recording the fitness value of the current parameter configuration, and marking the best fitness value among all parameter configurations; S36, using a heuristic algorithm to adjust the evaluation model parameter configuration and generate a new parameter configuration; S37, evaluating the training set using the new parameter configuration, calculating the fitness value of each clip abnormal segment set under the new parameter configuration; and updating the optimal fitness value and its corresponding parameter configuration among all parameter configurations; S38, check whether the maximum number of iterations has been reached, if so, proceed to step S39, otherwise, return to step S36, continue to adjust the evaluation model parameter configuration; output the evaluation model parameter configuration with the highest fitness value; S39. Use the optimal parameter configuration to evaluate the abnormal editing segments in the test set, quantify the matching degree between the abnormal editing segments and the high-quality standards, calculate the impact score of the editing abnormality segment set on the quality of the finished editing, and output the evaluation results.

7. The method for determining the video material based on AI automatic editing according to claim 6, characterized in that: The method of screening out the editing features that have the greatest impact on the quality of the finished editing by using the feature screening algorithm comprises the following steps: S321, initializing all parameters in the feature screening algorithm; S322, randomly selecting a feature combination as an initial feature candidate set, and initializing the feature weights using chaos thinking; S323, evaluating the influence of each candidate feature combination on the editing quality, calculating the fitness value of the candidate feature combination, and finding the feature combination with the best performance at present; S324, updating the weight of the current feature combination to generate a new candidate feature combination; S325, checking whether the updated feature combination meets the feature quantity threshold, if so, updating the feature combination according to the update rule, otherwise, keeping the current feature combination unchanged; S326, recalculate the fitness value of the updated feature combination, and compare it with the previous feature combination to find the feature combination with the highest fitness value; S327, comparing the fitness values ​​of the current feature combination and the historical optimal feature combination, and recording the feature combination with the highest fitness as the new optimal feature combination; S328, check whether the maximum number of iterations has been reached. If so, output the optimal feature combination as the editing feature that has the greatest impact on the quality of the finished editing. Otherwise, return to step S324 and continue iterative optimization.

8. The method for determining video material based on AI automatic editing according to claim 7, characterized in that: The method of adjusting the evaluation model parameter configuration by using a heuristic algorithm to generate a new parameter configuration includes the following steps: S361, setting the parameters of the heuristic algorithm, randomly generating a parameter configuration population matrix, each row represents a candidate parameter configuration, and each column represents an evaluation model parameter; S362, evaluate the fitness value of each parameter configuration in turn, check whether the parameter configuration exceeds the upper limit and the lower limit, if so, adjust the parameter configuration to the boundary value, otherwise, retain the current parameter configuration, and calculate the fitness value of each parameter configuration through the fitness function; S363, when the initial iteration is performed, the parameter configurations are sorted from small to large according to the fitness value, the sorted parameter configuration population is the optimal parameter configuration population, and the first parameter configuration in the optimal parameter configuration population is selected as the optimal parameter configuration. When the subsequent iteration is performed, the current optimal parameter configuration population is merged with the previous optimal parameter configuration population, and they are re-sorted according to the fitness value to select a new optimal parameter configuration population, and the first parameter configuration in the new optimal parameter configuration population is selected as the optimal parameter configuration; S364, updating the parameter configurations in the parameter configuration population one by one according to the updating rule of the heuristic algorithm; S365. Check whether the maximum number of iterations has been reached. If so, output the optimal parameter configuration and the corresponding fitness value as the new parameter configuration. Otherwise, return to step S362 and continue to execute until the maximum number of iterations has been reached.

9. The method for determining video material based on AI automatic editing according to claim 8, characterized in that: The formula of the update rule is: ； In the formula, A e Indicates e Parameter configuration; M n Indicates n Target parameter configuration; G e Indicates e The parameter configuration and n The difference between the target parameter configurations; b represents the constant that defines the shape of the logarithmic spiral; t Represents a random number; B represents the spiral function.

10. A judgment system based on AI automatic editing of video materials, used to implement the judgment method based on AI automatic editing of video materials according to any one of claims 1 to 9, characterized in that: The judgment system includes: A data acquisition module is used to acquire video material data to be edited and pre-process the video material data to obtain multimodal video data; An anomaly recognition module is used to analyze the obtained multimodal video data based on an anomaly recognition algorithm, mark the abnormal material data, and generate a set of clip abnormality segments; A quality comparison module is used to compare the generated abnormal editing segment set with the known high-quality video material standard to evaluate the impact of the abnormal editing segment set on the quality of the finished editing; An evaluation and judgment module is used to judge whether the set of abnormal clips meets the video editing standard based on the evaluation results, screen out the materials that meet the video editing standard, and generate a high-quality material library for editing for storage; Among them, the data acquisition module is connected with the quality comparison module through the abnormality recognition module, and the quality comparison module is connected with the evaluation and judgment module.

Images