Martial art instrument confrontation training analysis management system and method based on visual analysis

Abstract

The invention belongs to the technical field of training management, and discloses a martial art instrument confrontation training analysis management system based on visual analysis, which monitors the martial art instrument confrontation training process in real time through a motion capture module, obtains comprehensive and accurate training data, and sends the data to a data processing module. The data processing module removes noise and abnormal values in the data and extracts key training feature data, the data analysis module obtains the processed training feature data, and feature parameters of trainees and martial art instruments in the training process are analyzed through the action posture analysis unit and the instrument track analysis unit respectively; and then the training effect is evaluated by combining the characteristic parameter analysis results of the trainees and the martial art instruments, and the training management module provides personalized training suggestions for the trainees based on the evaluation result of the training effect, so that the quality and efficiency of martial art instrument confrontation training are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of training management, and in particular relates to a martial arts equipment confrontation training analysis management system and method based on visual analysis. Background Art

[0002] Martial arts are a form of combat technique inherited from ancient military warfare. Practicing martial arts strengthens the body and serves as a defense against enemy attacks. Martial artists, guided by the principle of "stopping aggression," guide practitioners through traditional methods of education (martialization) that foster an understanding of the objective laws governing the relationship between man, nature, and society. They guide and safeguard human material civilization and represent a contemporary expression of traditional martial arts.

[0003] Current methods for evaluating and analyzing martial arts weapon confrontation training often rely on the instructor's experience and judgment, lacking objective and accurate data. This approach often suffers from the following issues: It's difficult to accurately record key information, such as the trainee's movements and the trajectory of both players, hindering rapid improvement in training. With the continuous advancement of computer vision technology, this paper proposes a martial arts weapon confrontation training analysis and management system and method based on visual analysis. Summary of the Invention

[0004] The purpose of the present invention is to provide a martial arts equipment confrontation training analysis and management system based on visual analysis, which uses visual analysis technology to monitor and analyze the martial arts equipment confrontation training process in real time, obtain comprehensive and accurate training data, and effectively manage and analyze the data to provide students with personalized training suggestions, thereby improving the quality and efficiency of martial arts equipment confrontation training.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A martial arts equipment confrontation training analysis and management system based on visual analysis, including a motion capture module, a data processing module, a data analysis module and a training management module;

[0007] The motion capture module includes multiple high-definition cameras, which are arranged at different locations in the martial arts equipment confrontation training field to collect image and video data during the training process;

[0008] The data processing module is used to further process and integrate the data output by the motion capture module and extract key training feature parameters;

[0009] The data analysis module includes a posture analysis unit and an instrument trajectory analysis unit. The posture analysis unit is used to analyze the student's posture based on the motion parameters of the key points of the human body. The instrument trajectory analysis unit is used to analyze the motion trajectory of the martial arts equipment based on the motion parameters of the key positions of the martial arts equipment. The analysis results of the posture analysis unit and the instrument trajectory analysis unit are combined to evaluate the student's training process.

[0010] The training management module adjusts the trainee training plan based on the evaluation results of the data analysis module.

[0011] As a further description of the technical solution of the present invention, the working process of the motion capture module includes:

[0012] Arrange multiple high-definition cameras around the martial arts equipment confrontation training ground to ensure coverage of the entire training area, and configure the camera parameters accordingly;

[0013] Install the various module software of this system on the computer equipment at the training site, and perform system initialization settings, including entering student information and setting training standards.

[0014] Through the above technical solution, this embodiment provides a system deployment method that rationally arranges multiple high-definition cameras around a martial arts equipment confrontation training venue to ensure coverage of the entire training area. The cameras are connected to the system's image acquisition module and corresponding parameters such as resolution and frame rate are configured. Simultaneously, the various module software of this system are installed on the computer equipment at the training venue, and the system is initialized and configured, including entering student information and setting training standards.

[0015] As a further description of the technical solution of the present invention, the specific working process of the data processing module includes:

[0016] The confrontation training is decomposed into x training movements. According to the training standards, the human body is divided into n key points and the martial arts equipment is divided into m key positions of the equipment. The key feature parameters of the human body at the n key points in the x training movements and the key feature parameters of the equipment at the m key positions in the x training movements are extracted respectively.

[0017] The key characteristic parameters of the human body include: the consumption time of a single action, the linear velocity of key points and the linear acceleration of key points;

[0018] The key characteristic parameters of the instrument include: the linear velocity of the end of the key position of the instrument, the linear acceleration of the end of the key position of the instrument, and the movement amplitude of the key position of the instrument.

[0019] As a further description of the technical solution of the present invention, the specific working process of the action posture analysis unit includes:

[0020] Get the time T of the yth training action y , the linear velocity V of the i-th key point of the y-th training action yi and the linear acceleration a of the i-th key point of the y-th training action yi , where y belongs to x and i belongs to n;

[0021] Construct a mathematical calculation model for the first evaluation coefficient of training, the expression is:

[0022]

[0023] Where, T y0 V is the standard time consumption of the yth training action preset by the system, yi0 The standard linear speed of the i-th key point of the y-th training action preset by the system, a yi0 The standard linear acceleration of the i-th key point of the y-th training action preset by the system, k y is the difficulty coefficient of the yth training action, and μ1 is the first evaluation coefficient of the training.

[0024] As a further description of the technical solution of the present invention, the specific working process of the instrument trajectory analysis unit includes:

[0025] Get the range of motion F of the jth key position of the equipment in the yth training action yj , obtain the linear velocity V at the end of the key position of the jth equipment in the yth training action yj and the linear acceleration a at the end of the key position of the jth equipment in the yth training action yj , where y belongs to x and j belongs to m;

[0026] Construct a mathematical calculation model for training the second evaluation coefficient, the expression is:

[0027]

[0028] Where, F yj0 The standard range of motion for the jth key position of the equipment in the yth training action preset by the system, V yj0 The standard linear speed of the jth key position of the equipment for the yth training action preset by the system, a yj0 The standard linear acceleration of the jth equipment key position of the yth training action preset by the system, k y is the difficulty coefficient of the y-th training action, and μ2 is the second evaluation coefficient of the training.

[0029] As a further description of the technical solution of the present invention, the specific working process of the data analysis module includes:

[0030] Obtain the heart rate change rate R and the number of completed movements z after the trainee completes the entire confrontation training, and construct a mathematical calculation model for the comprehensive evaluation coefficient of the training. The expression is:

[0031]

[0032] Where μ is the comprehensive evaluation coefficient. The comprehensive evaluation coefficient is compared with the comprehensive evaluation coefficient threshold set by the system. If the comprehensive evaluation coefficient is greater than or equal to the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is considered qualified. If the comprehensive evaluation coefficient is less than the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is considered unqualified.

[0033] As a further description of the technical solution of the present invention, the working process of the training management module includes:

[0034] If the current adversarial training is qualified, keep the current training times. If the current adversarial training is unqualified, use the formula Calculate the completion degree of the yth training action. Similarly, calculate the completion degree of x training actions in sequence and compare them with the x action completion degree qualification value set by the system. For training actions that are lower than the x action completion degree qualification value, increase the number of training times.

[0035] A method for analyzing and managing martial arts equipment confrontation training based on visual analysis, the method comprising the following steps:

[0036] Step S1: Arrange multiple high-definition cameras reasonably around the martial arts equipment confrontation training ground and configure the corresponding parameters. At the same time, install the various module software of the system on the computer equipment of the training ground and perform system initialization settings, including entering student information and setting training standards;

[0037] Step S2: When students are performing martial arts weapon confrontation training, the motion capture module collects image and video data of the training process in real time through the camera and transmits the data to the data processing module;

[0038] Step S3: The data processing module processes and analyzes the data, performs motion posture analysis, and equipment trajectory tracking to extract key training feature parameters, and normalizes all training parameters to the same scale.

[0039] Step S4: The training analysis module evaluates the trainee's training performance based on the key training feature parameters provided by the data processing module;

[0040] Step S5: The training management module provides suggestions for subsequent training of the training action based on the evaluation results.

[0041] Beneficial effects of the present invention:

[0042] Improve the accuracy of training evaluation: Through visual analysis technology, students' training data can be obtained objectively and accurately, avoiding the subjectivity and errors of human judgment, and improving the accuracy and reliability of training evaluation.

[0043] Provide personalized training plans: Based on the trainees' training data and analysis results, the system can provide personalized training suggestions for each trainee to meet the trainees' personalized training needs and improve training results.

[0044] Promoting the scientific development of martial arts training: This invention applies computer vision technology to martial arts equipment confrontation training, providing new technical means and methods for martial arts training, and helping to promote the scientific and intelligent development of martial arts training.

[0045] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. 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 creative work.

[0047] Figure 1 This is a structural diagram of the martial arts equipment confrontation training analysis and management system based on visual analysis of the present invention. DETAILED DESCRIPTION

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0049] See also Figure 1 As shown, a martial arts equipment confrontation training analysis and management system based on visual analysis is disclosed, which includes a motion capture module, a data processing module, a data analysis module and a training management module;

[0050] The motion capture module includes multiple high-definition cameras, which are arranged at different locations in the martial arts equipment confrontation training field to collect image and video data during the training process;

[0051] The data processing module is used to further process and integrate the data output by the motion capture module and extract key training feature parameters;

[0052] The data analysis module includes a posture analysis unit and an instrument trajectory analysis unit. The posture analysis unit is used to analyze the student's posture based on the motion parameters of the key points of the human body. The instrument trajectory analysis unit is used to analyze the motion trajectory of the martial arts equipment based on the motion parameters of the key positions of the martial arts equipment. The analysis results of the posture analysis unit and the instrument trajectory analysis unit are combined to evaluate the student's training process.

[0053] The training management module adjusts the trainee training plan based on the evaluation results of the data analysis module.

[0054] As a further description of the technical solution of the present invention, the working process of the motion capture module includes:

[0055] Arrange multiple high-definition cameras around the martial arts equipment confrontation training ground to ensure coverage of the entire training area, and configure the camera parameters accordingly;

[0056] Install the various module software of this system on the computer equipment at the training site, and perform system initialization settings, including entering student information and setting training standards.

[0057] Through the above technical solution, the present invention provides a martial arts equipment confrontation training analysis and management system based on visual analysis, which monitors the martial arts equipment confrontation training process in real time through a motion capture module, obtains comprehensive and accurate training data, and sends the data to a data processing module. The data processing module removes noise and outliers in the data and extracts key training feature data. The data analysis module obtains the processed training feature data, analyzes the characteristic parameters of the students and martial arts equipment during the training process through a motion posture analysis unit and an equipment trajectory analysis unit respectively, and then evaluates the training effect based on the analysis results of the characteristic parameters of the students and martial arts equipment. The training management module provides personalized training suggestions to the students based on the evaluation results of the training effect, thereby improving the quality and efficiency of martial arts equipment confrontation training.

[0058] As a further description of the technical solution of the present invention, the specific working process of the data processing module includes:

[0059] The confrontation training is decomposed into x training movements. According to the training standards, the human body is divided into n key points and the martial arts equipment is divided into m key positions of the equipment. The key feature parameters of the human body at the n key points in the x training movements and the key feature parameters of the equipment at the m key positions in the x training movements are extracted respectively.

[0060] The key characteristic parameters of the human body include: the consumption time of a single action, the linear velocity of key points and the linear acceleration of key points;

[0061] The key characteristic parameters of the instrument include: the linear velocity of the end of the key position of the instrument, the linear acceleration of the end of the key position of the instrument, and the movement amplitude of the key position of the instrument.

[0062] Through the above technical solution, this embodiment further processes and integrates the data output by the motion capture module, including data cleaning, noise reduction, feature extraction and other operations, removes noise and outliers in the data, and extracts key training feature data, such as: key feature parameters of the human body and key feature parameters of the instrument.

[0063] As a further description of the technical solution of the present invention, the specific working process of the action posture analysis unit includes:

[0064] Get the time T of the yth training action y , the linear velocity V of the i-th key point of the y-th training action yi and the linear acceleration a of the i-th key point of the y-th training action yi , where y belongs to x and i belongs to n;

[0065] Construct a mathematical calculation model for the first evaluation coefficient of training, the expression is:

[0066]

[0067] Where, T y0 V is the standard time consumption of the yth training action preset by the system, yi0 The standard linear speed of the i-th key point of the y-th training action preset by the system, a yi0 The standard linear acceleration of the i-th key point of the y-th training action preset by the system, k y is the difficulty coefficient of the yth training action, and μ1 is the first evaluation coefficient of the training.

[0068] Through the above technical solution, this embodiment provides a method for analyzing movement posture based on key characteristic parameters of the human body during training. The method decomposes the training into x training movements, divides the human body into n key points, obtains the linear velocity and linear acceleration of each key point during each training movement, and obtains the time consumed by each training movement. Then, the method is used to calculate the linear velocity and linear acceleration of each key point during each training movement. Calculate the first evaluation coefficient of training, where T y0 V is the standard time consumption of the yth training action preset by the system, yi0 The standard linear speed of the i-th key point of the y-th training action preset by the system, a yi0 The standard linear acceleration of the i-th key point of the y-th training action preset by the system, k y is the difficulty coefficient of the yth training action.

[0069] As a further description of the technical solution of the present invention, the specific working process of the instrument trajectory analysis unit includes:

[0070] Get the range of motion F of the jth key position of the equipment in the yth training action yj , obtain the linear velocity V at the end of the key position of the jth equipment in the yth training action yj and the linear acceleration a at the end of the key position of the jth equipment in the yth training action yj , where y belongs to x and j belongs to m;

[0071] Construct a mathematical calculation model for training the second evaluation coefficient, the expression is:

[0072]

[0073] Where, F yj0 The standard range of motion for the jth key position of the equipment in the yth training action preset by the system, V yj0 The standard linear speed of the jth key position of the equipment for the yth training action preset by the system, a yj0 The standard linear acceleration a of the key position of the jth equipment in the yth training action preset by the system yi , k y is the difficulty coefficient of the y-th training action, and μ2 is the second evaluation coefficient of the training.

[0074] Through the above technical solution, this embodiment provides a method for analyzing the trajectory of an instrument based on the key characteristic parameters of the instrument during training. By decomposing the training into x training movements, dividing the martial arts instruments into m instrument key positions, respectively obtaining the motion amplitude of the j-th instrument key position of the y-th training movement, and obtaining the linear velocity and acceleration of the end of the j-th instrument key position of the y-th training movement, the formula Calculate the second evaluation coefficient, where Fyj0 is the standard range of motion for the jth key position of the yth training action preset by the system, V yj0 The standard linear speed of the jth key position of the equipment for the yth training action preset by the system, a yj0 The standard linear acceleration of the jth equipment key position of the yth training action preset by the system, k y is the difficulty coefficient of the yth training action.

[0075] As a further description of the technical solution of the present invention, the specific working process of the data analysis module includes:

[0076] Obtain the heart rate change rate R and the number of completed movements z after the trainee completes the entire confrontation training, and construct a mathematical calculation model for the comprehensive evaluation coefficient of the training. The expression is:

[0077]

[0078] In the formula, μ is the comprehensive evaluation coefficient, α and β are the weight coefficients corresponding to the first evaluation coefficient and the second evaluation coefficient respectively. The comprehensive evaluation coefficient is compared with the comprehensive evaluation coefficient threshold set by the system. If the comprehensive evaluation coefficient is greater than or equal to the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is considered qualified. If the comprehensive evaluation coefficient is less than the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is considered unqualified.

[0079] As a further description of the technical solution of the present invention, the working process of the training management module includes:

[0080] If the current adversarial training is qualified, keep the current training times. If the current adversarial training is unqualified, use the formula Calculate the completion degree of the yth training action. Similarly, calculate the completion degree of x training actions in sequence and compare them with the x action completion degree qualification value set by the system. For training actions that are lower than the x action completion degree qualification value, increase the number of training times.

[0081] Through the above technical solution, this embodiment conducts a comprehensive evaluation of the training based on the analysis results of the key characteristic parameters of the human body and the key characteristic parameters of the equipment, and combines the number of movements completed during the training process and the physical condition of the trainees. Calculate the comprehensive evaluation coefficient, and then compare the comprehensive evaluation coefficient with the comprehensive evaluation coefficient threshold set by the system. If the comprehensive evaluation coefficient is greater than or equal to the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is qualified and the current training times are maintained. If the comprehensive evaluation coefficient is less than the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is unqualified. If the current adversarial training is unqualified, the formula Calculate the completion degree of the yth training action. Similarly, calculate the completion degree of x training actions in sequence and compare them with the x action completion degree qualification value set by the system. For training actions that are lower than the x action completion degree qualification value, increase the number of training times.

[0082] A method for analyzing and managing martial arts equipment confrontation training based on visual analysis, the method comprising the following steps:

[0083] Step S1: Arrange multiple high-definition cameras reasonably around the martial arts equipment confrontation training ground and configure the corresponding parameters. At the same time, install the various module software of the system on the computer equipment of the training ground and perform system initialization settings, including entering student information and setting training standards;

[0084] Step S2: When students are performing martial arts weapon confrontation training, the motion capture module collects image and video data of the training process in real time through the camera and transmits the data to the data processing module;

[0085] Step S3: The data processing module processes and analyzes the data, performs motion posture analysis, and equipment trajectory tracking to extract key training feature parameters, and normalizes all training parameters to the same scale.

[0086] Step S4: The training analysis module evaluates the trainee's training performance based on the key training feature parameters provided by the data processing module;

[0087] Step S5: The training management module provides suggestions for subsequent training of the training action based on the evaluation results.

[0088] It should be noted that the formulas in this application are all dimensionless and calculated numerically. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The thresholds involved in this application and the system-set coefficients in the formulas are all empirical values, which are selected and set by technical personnel in this field according to actual conditions.

[0089] The above content is merely an example and explanation of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the claims, they should all fall within the scope of protection of the present invention.

Claims

1. A martial arts equipment confrontation training analysis and management system based on visual analysis, characterized in that: Including motion capture module, data processing module, data analysis module and training management module; The motion capture module includes multiple high-definition cameras, which are arranged at different locations in the martial arts equipment confrontation training field to collect image and video data during the training process; The data processing module is used to further process and integrate the data output by the motion capture module and extract key training feature parameters; The data analysis module includes a posture analysis unit and an instrument trajectory analysis unit. The posture analysis unit is used to analyze the student's posture based on the motion parameters of the key points of the human body. The instrument trajectory analysis unit is used to analyze the motion trajectory of the martial arts equipment based on the motion parameters of the key positions of the martial arts equipment. The analysis results of the posture analysis unit and the instrument trajectory analysis unit are combined to evaluate the student's training process. The training management module adjusts the trainee training plan based on the evaluation results of the data analysis module.

2. The martial arts equipment confrontation training analysis and management system based on visual analysis according to claim 1, characterized in that: The working process of the motion capture module includes: Arrange multiple high-definition cameras around the martial arts equipment confrontation training ground to ensure coverage of the entire training area, and configure the camera parameters accordingly; Install the various module software of this system on the computer equipment at the training site, and perform system initialization settings, including entering student information and setting training standards.

3. The martial arts equipment confrontation training analysis and management system based on visual analysis according to claim 2, characterized in that: The specific working process of the data processing module includes: The confrontation training is decomposed into x training movements. According to the training standards, the human body is divided into n key points and the martial arts equipment is divided into m key positions of the equipment. The key feature parameters of the human body at the n key points in the x training movements and the key feature parameters of the equipment at the m key positions in the x training movements are extracted respectively. The key characteristic parameters of the human body include: the consumption time of a single action, the linear velocity of key points and the linear acceleration of key points; The key characteristic parameters of the instrument include: the linear velocity of the end of the key position of the instrument, the linear acceleration of the end of the key position of the instrument, and the movement amplitude of the key position of the instrument.

4. The martial arts equipment confrontation training analysis and management system based on visual analysis according to claim 3, characterized in that: The specific working process of the action posture analysis unit includes: Get the time T of the yth training action y , the linear velocity V of the i-th key point of the y-th training action yi and the linear acceleration a of the i-th key point of the y-th training action yi , where y belongs to x and i belongs to n; Construct a mathematical calculation model for the first evaluation coefficient of training, the expression is: Where, T y0 V is the standard time consumption of the yth training action preset by the system, yi0 The standard linear speed of the i-th key point of the y-th training action preset by the system, a yi0 The standard linear acceleration of the i-th key point of the y-th training action preset by the system, k y is the difficulty coefficient of the yth training action, and μ1 is the first evaluation coefficient of the training.

5. The martial arts equipment confrontation training analysis and management system based on visual analysis according to claim 3, characterized in that: The specific working process of the instrument trajectory analysis unit includes: Get the range of motion F of the jth key position of the equipment in the yth training action yj , obtain the linear velocity V at the end of the key position of the jth equipment in the yth training action yj and the linear acceleration a at the end of the key position of the jth equipment in the yth training action yj , where y belongs to x and j belongs to m; Construct a mathematical calculation model for training the second evaluation coefficient, the expression is: Where, F yj0 The standard range of motion for the jth key position of the equipment in the yth training action preset by the system, V yj0 The standard linear speed of the jth key position of the equipment for the yth training action preset by the system, a yj0 The standard linear acceleration of the jth equipment key position of the yth training action preset by the system, k y is the difficulty coefficient of the y-th training action, and μ2 is the second evaluation coefficient of the training.

6. The martial arts equipment confrontation training analysis and management system based on visual analysis according to claim 3, characterized in that: The specific working process of the data analysis module includes: Obtain the heart rate change rate R and the number of completed movements z after the trainee completes the entire confrontation training, and construct a mathematical calculation model for the comprehensive evaluation coefficient of the training. The expression is: Where μ is the comprehensive evaluation coefficient. The comprehensive evaluation coefficient is compared with the comprehensive evaluation coefficient threshold set by the system. If the comprehensive evaluation coefficient is greater than or equal to the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is considered qualified. If the comprehensive evaluation coefficient is less than the comprehensive evaluation coefficient threshold set by the system, the current adversarial training is considered unqualified.

7. The martial arts equipment confrontation training analysis and management system based on visual analysis according to claim 1, characterized in that: The working process of the training management module includes: If the current adversarial training is qualified, keep the current training times. If the current adversarial training is unqualified, use the formula Calculate the completion degree of the yth training action. Similarly, calculate the completion degree of x training actions in sequence and compare them with the x action completion degree qualification value set by the system. For training actions that are lower than the x action completion degree qualification value, increase the number of training times.

8. A method for analyzing and managing martial arts equipment confrontation training based on visual analysis, the method being applicable to the martial arts equipment confrontation training analysis and management system based on visual analysis according to any one of claims 1 to 7, characterized in that: The method comprises the following steps: Step S1: Arrange multiple high-definition cameras reasonably around the martial arts equipment confrontation training ground and configure the corresponding parameters. At the same time, install the various module software of the system on the computer equipment of the training ground and perform system initialization settings, including entering student information and setting training standards; Step S2: When students are performing martial arts weapon confrontation training, the motion capture module collects image and video data of the training process in real time through the camera and transmits the data to the data processing module; Step S3: The data processing module processes and analyzes the data, performs motion posture analysis, and equipment trajectory tracking to extract key training feature parameters, and normalizes all training parameters to the same scale. Step S4: The training analysis module evaluates the trainee's training performance based on the key training feature parameters provided by the data processing module; Step S5: The training management module provides suggestions for subsequent training of the training action based on the evaluation results.

Images