Learning machine interactive teaching auxiliary system
By collecting multi-dimensional data and generating indices, the haptic interaction model is dynamically adjusted, which solves the problem of insufficient adaptability to individual differences in interactive teaching systems for learning machines, and improves learning effectiveness and security.
Patent Information
- Application Number
- CN202511504255.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-21
- Publication Date
- 2026-02-03
AI Technical Summary
Traditional interactive teaching aids for learning machines suffer from insufficient adaptability to individual differences, inadequate standardization of actions, and a lack of a unified evaluation mechanism, resulting in poor learning outcomes and the risk of user injury.
By collecting user data, practice videos, and course data through a multi-dimensional acquisition module, a difference index, a proficiency index, and a conversion index are generated. The haptic interaction model is dynamically adjusted to provide individualized differential interaction and auxiliary teaching suggestions.
It enables individualized interaction based on differences, dynamically adjusts exercise intensity, avoids sports injuries caused by differences in movement rhythm, and improves learning effectiveness and safety.
Smart Images
Figure CN121455327A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of somatosensory interaction teaching assistance, in particular to a learning machine interactive teaching assistance system. BACKGROUND
[0002] The somatosensory interaction learning machine is an innovative education tool that combines somatosensory technology, artificial intelligence and interactive teaching. It captures the user's body movements, postures and motion trajectories in real time, seamlessly connects physical behavior with virtual teaching content, and creates an immersive and intuitive interactive experience for learners. The core advantage of somatosensory interaction technology lies in natural interaction and multi-modal feedback. Users do not need to rely on traditional controllers or buttons, but can complete operations with body movements alone, significantly reducing the learning threshold and improving participation enthusiasm. In dance, fitness, martial arts and other teaching, somatosensory technology not only needs to analyze whether the user's movement amplitude meets the teaching standard, but also needs to ensure the standardization and safety of the movement. The core goal of teaching assistance is to improve movement accuracy, strengthen rhythm perception and achieve personalized advancement. In the future, it is expected to play a more important role in education, fitness, rehabilitation and other fields.
[0003] At present, the somatosensory interaction technology used by the traditional learning machine interactive teaching assistance system is mostly a general model. In the teaching assistance process, the general model has poor adaptability to individual differences of users, and there is a problem of insufficient standardization of somatosensory interaction movements, which affects the learning effect. In addition, there is a lack of unified movement proficiency evaluation mechanism, making it difficult to accurately reflect the real learning achievements. When multiple courses are linked, due to the difference in movement rhythm, it is necessary to control artificially, which is easy to cause a high risk of user injury. SUMMARY
[0004] In view of the shortcomings of the prior art, the present application provides a learning machine interactive teaching assistance system, which has the advantages of emphasizing individualized difference interaction and good dynamic assistance teaching effect, and solves the problems of insufficient movement standardization and lack of unified evaluation mechanism of the traditional learning machine interactive teaching assistance system.
[0005] To achieve the above purpose, the present application provides the following technical scheme: a learning machine interactive teaching assistance system, comprising a multi-dimensional acquisition module and a teaching assistance module; The multi-dimensional acquisition module is composed of a user management unit, a learning management unit and a course management unit. The user management unit acquires user data set through network connection learning machine, the user data set includes management data of all users, the learning management unit acquires practice data set through network connection high-definition camera, the practice data set includes practice videos of all users, and the course management unit acquires course data set through network connection big data platform, the course data set includes management data of all teaching courses; The teaching auxiliary module is composed of a user evaluation unit, a learning evaluation unit, an interaction evaluation unit and an auxiliary management unit, the user evaluation unit analyzes the difference of each user from the somatosensory interaction model according to the user data set, and generates a corresponding difference index , the learning evaluation unit identifies the action card point and the positioning point of the user in the exercise video according to the exercise data set and the course data set, and analyzes the proficiency of each user to the teaching course, and generates a corresponding proficiency index , the interaction evaluation unit analyzes the difference degree of action rhythm between different courses according to the course data set, and generates a corresponding conversion index , the auxiliary management unit is provided with fixed value difference threshold , proficiency threshold and conversion threshold , combined with difference index , proficiency index and conversion index , output corresponding auxiliary teaching suggestions.
[0006] Preferably, the user data set includes the age, height, sitting height, leg length and arm span of each user.
[0007] Preferably, the exercise data set includes the exercise video of each user, wherein each exercise video contains several action screenshots.
[0008] Preferably, the course data set includes the name, action card point, positioning point and length of each teaching course.
[0009] Preferably, the difference index The calculation process is as follows: S11, according to the user data set, the management data of the first user is extracted, and the height of the first user is marked as , the sitting height of the first user is marked as , the leg length of the first user is marked as , and the arm span of the first user is marked as ; S12, the relative length of the trunk of the first user is calculated ; S13, the relative length of the leg of the first user is calculated ; S14, the arm height ratio of the first user is calculated ; S15. Collect standard values for measuring the relative length of the torso from a motion-sensing interaction model suitable for users under 60 years old via a network-connected learning machine. Standard values used to measure the relative length of the legs Standard values used to measure the arm span to height ratio And in a motion-sensing interaction model applicable to users aged 60 and above, the standard value used to measure the relative length of the torso. Standard values used to measure the relative length of the legs Standard values used to measure the arm span to height ratio Then, according to S11-S14, in accordance with the first... The age of the user is calculated using different weighting methods. Individual user difference index .
[0010] Preferably, the proficiency index The calculation process is as follows: S21. Based on the practice dataset, extract the first... Each user owns teaching courses. The video demonstrates imitation exercises, and the first... Individual User Imitation Teaching Course The total number of practice sessions is marked as ; S22. Extract teaching courses based on the course dataset. All action points in the game, and marked them in chronological order from earliest to latest. Then use video editing software to time the action precisely. to Extract teaching courses in sequence Screenshots of the motion-sensing interaction model; S23. Using video processing software, follow the action beats. to , and extract the first one in sequence The user Screenshots of the movements from the practice video. ; S24, Follow the action timings to By using desaturation and multiply blending modes, the results are compared sequentially. The user Screenshots of the actions and instructional lessons in the practice video. Screenshots of the motion-sensing interaction model are used to generate corresponding checkpoint scores. The judgment process is as follows: The first The user The practice video includes screenshots of individual movements and instructional lessons. In the process, the screenshot of a single action of the motion-sensing interaction model is desaturated to obtain two black and white images, and then the two black and white images are overlaid using the multiply blending mode. If no color change occurs after overlaying two black and white images, it indicates that the first image is the first black and white image. The user In this practice video, screenshots of individual movements and instructional lessons. In the middle, the single action of the motion-sensing interaction model is captured. Figure One To, the The magnitude of a single user's action has met the standard, and the [number]th [action] will be [action]. The user The timing score for the practice video Add 1 point; If a color change occurs after two black and white images are superimposed, it indicates that the first image is the first black and white image. The user In this practice video, screenshots of individual movements and instructional lessons. In the process, the screenshots of a single action in the motion-sensing interaction model are inconsistent. If a user's single action does not meet the standard, the [number]th [action] will be [action]. The user The timing score for the practice video Add 0 points; S25, the first The user The timing score for this practice video is marked as follows: , Then calculate the first Standard deviation of all score points for each user ; S26. Extract teaching courses based on the course dataset. All the positioning points in the data are marked sequentially from earliest to latest according to their appearance time. , The timestamp indicating the location point is then used in video processing software to analyze the data according to the location point. to Obtain the teaching courses in sequence In the context of the motion-sensing interaction model, the coordinates of the positioning point are shown. S27. Based on the practice dataset, extract the first... The first user's This is a practice video. Using video processing software, the video points are captured sequentially from earliest to latest, according to their appearance time. The user The coordinates of the positioning points in the practice video, and the teaching course. The coordinates of the positioning point in the motion-sensing interaction model and the first The user The coordinates of the positioning points in this practice video all originate from the same three-dimensional coordinate system; S28. According to the positioning point to Compare the first one in turn The user The positioning points and teaching courses in this practice video The positioning points of the motion-sensing interaction model are used to generate corresponding positioning scores. The judgment process is as follows: If the first The user The coordinates of all positioning points in this practice video and the teaching course If all the coordinates of the positioning points in the motion-sensing interaction model are consistent, and the timestamps of the positioning points appearing in the time dimension are consistent, then it indicates that the [number]th [position] is [position]. The user The completeness of the movements in the second practice session has reached the standard. The user Positioning score for the practice video It is recorded as 1 point; If the first The user The coordinates of any point in the practice video are related to the teaching course. If all the coordinates of the positioning points in the motion-sensing interaction model are consistent, or if the timestamps of the positioning points appearing are inconsistent in the time dimension, then it indicates that the first... The user The completeness of the movement in the first practice session did not meet the standard. The user Positioning score for the practice video Recorded as -1 point; S29. Based on S21, S26-S28, the cumulative number of... Location rating of all practice videos for each user and will the Individual user location score The sum is marked as ; S30. Collect teaching courses by connecting to a big data platform via the network. Recommended number of practice sessions and the standard value used to measure the standard deviation of user checkpoint scores Then, based on S21-S29, the weighted average is used to calculate the... Individual users on teaching courses Proficiency index .
[0011] Preferably, the conversion index The calculation process is as follows: S31. Extract teaching courses based on the course dataset. and teaching courses Management data, and teaching courses The duration is marked as The teaching courses The total number of action checkpoints is marked as The teaching courses The duration is marked as The teaching courses The total number of action checkpoints is marked as ; S32, Computation Teaching Course frequency of card points and teaching courses frequency of card points ; S33, Computation Teaching Course With teaching courses Time difference ; S34. Based on S31-S33, calculate the practice teaching course. User adaptation tutorial courses Conversion index at time .
[0012] Preferably, the difference index ≥ Difference Threshold If the difference between the user and the motion-sensing interaction model is significant, the motion-sensing interaction model should be reconstructed based on the user's age, height, sitting height, leg length, and arm span.
[0013] Preferably, the proficiency index <Proficiency Threshold If the user has not yet mastered the movements in the course, the teaching speed should be reduced, the user should be assisted in optimizing the range of motion, and the user should be advised to increase the frequency of practice.
[0014] Preferably, the conversion index >Conversion threshold If this occurs, it indicates that there is a significant difference in the rhythm of the movements between the user's current practice course and the adapted course, and the user should be re-adapted to a course with a gentler rhythm.
[0015] Compared with the prior art, the present invention provides an interactive teaching assistance system for learning machines, which has the following beneficial effects: 1. The application obtains the management data of all users, the practice videos and the management data of all teaching courses through the multi-dimensional acquisition module network connection learning machine, high-definition camera and big data platform, and classifies and forms user data set, practice data set and course data set, the teaching auxiliary module analyzes the difference of each user and the somatosensory interaction model according to the user data set, and generates the corresponding difference index , focuses on individualized difference interaction, the teaching auxiliary module identifies the action card point and positioning point of the user in the practice video according to the practice data set and the course data set, and analyzes the proficiency of each user to the teaching course, and generates the corresponding proficiency index , helps the user to master the course content gradually, the teaching auxiliary module analyzes the difference degree of action rhythm between different courses according to the course data set, and generates the corresponding conversion index , dynamically adjusts the exercise intensity to accurately adapt to the user's rhythm, and avoids the exercise injury caused by the too large difference of action rhythm.
[0016] 2. The application sets the fixed value of the difference threshold , the proficiency threshold and the conversion threshold , the difference index ≥ difference threshold , which indicates that the difference between the user and the somatosensory interaction model is significant, and the somatosensory interaction model should be reconstructed according to the user's age, height, sitting height, leg length and arm span, and the somatosensory interaction model is dynamically adjusted, the proficiency index < proficiency threshold , which indicates that the user has not mastered the action in the teaching course, the course teaching speed should be reduced, the action amplitude of the user should be optimized, and the user is suggested to increase the practice frequency, the error card point in the continuous action is accurately identified, the user is helped to master the course content gradually, and the conversion index > conversion threshold , which indicates that there is a significant difference in action rhythm between the teaching course currently practiced by the user and the adaptive teaching course, the teaching course with slow action rhythm change should be re-adapted, the user is avoided to produce frustration due to the sudden change of rhythm, and the dynamic auxiliary teaching effect is good. BRIEF DESCRIPTION OF DRAWINGS
[0017] Figure 1 The system flowchart of the application. DETAILED DESCRIPTION
[0018] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.
[0019] Embodiment 1 Please refer to Figure 1 The present application provides a learning machine interactive teaching auxiliary system, comprising a multi-dimensional acquisition module and a teaching auxiliary module. The multi-dimensional acquisition module is composed of a user management unit, a learning management unit and a course management unit. The user management unit acquires a user data set by connecting a learning machine through a network. The user data set includes the management data of all users. The user data set includes the age, height, sitting height, leg length and arm span of each user. The user data set intuitively reflects the body shape characteristics of the user, providing data support for subsequent construction of a user-specific body shape interaction model. The learning management unit acquires an exercise data set by connecting a high-definition camera through a network. The exercise data set includes the exercise videos of all users. The exercise data set includes the exercise videos of each user, wherein each exercise video contains a plurality of action screenshots. The exercise data set records in detail the learning process of the user, providing data support for subsequent extraction of action screenshot sequences. The course management unit acquires a course data set by connecting a big data platform through a network. The course data set includes the management data of all teaching courses. The course data set includes the name, action key points, positioning points and duration of each teaching course. The course data set is used to structureally manage the course content, which is helpful for subsequent user practice assistance. The teaching auxiliary module is composed of a user evaluation unit, a learning evaluation unit, an interaction evaluation unit and an auxiliary management unit. The user evaluation unit analyzes the difference between each user and the body shape interaction model according to the user data set, and generates a corresponding difference index The learning evaluation unit identifies the action key points and positioning points of the user in the exercise video according to the exercise data set and the course data set, and analyzes the proficiency of each user in the teaching course, and generates a corresponding proficiency index The interaction evaluation unit analyzes the difference degree of action rhythm between different courses according to the course data set, and generates a corresponding conversion index The auxiliary management unit sets fixed numerical difference threshold proficiency threshold and conversion threshold , and combines the difference index proficiency index and conversion index , output the corresponding auxiliary teaching suggestion, and the dynamic auxiliary teaching effect is good; Specifically, the difference threshold , the proficiency threshold and the conversion threshold are all based on the massive user practice data accumulated by the learning machine, and can be further combined with the joint range of motion characteristics of people of different ages to achieve precise adaptation by setting differentiated values, effectively solving the learning difficulties caused by physiological function differences; The difference index is greater than the difference threshold , which indicates that the difference between the user and the somatosensory interaction model is significant, and the somatosensory interaction model should be reconstructed according to the user's age, height, sitting height, leg length and arm span, and the somatosensory interaction model should be dynamically adjusted, and individual differences should be focused on; The proficiency index is less than the proficiency threshold , which indicates that the user has not yet mastered the action in the teaching course, and the course teaching speed should be reduced, the action amplitude of the user should be optimized, and the user should be suggested to increase the practice frequency, accurately identify the failure points in the continuous action, and help the user master the course content gradually; The conversion index is greater than the conversion threshold , which indicates that there is a significant difference in the action rhythm between the teaching course currently practiced by the user and the adapted teaching course, and the teaching course with gentle action rhythm change should be re-adapted to avoid the user's frustration due to the sudden change in rhythm.
[0020] In this embodiment, the user body feature data is collected by the multi-dimensional collection module, the learning behavior is recorded, and the course content is structured and managed, the difference index of the user body type and the standard model is calculated according to the user's age by the teaching assistance module , individual difference guidance is implemented, the proficiency index is calculated by analyzing the action compliance rate, the user is helped to master the course content gradually, the conversion index is calculated based on the change amount of the card point frequency, the exercise intensity is dynamically adjusted to accurately adapt to the user's rhythm, and exercise injury caused by too large action rhythm difference is avoided.
[0021] Embodiment 2 Please refer to Table 1 for difference index experimental data. This embodiment is based on the explanation and description of embodiment 1. Specifically, the difference index calculation process is as follows: S11, according to the user data set, the management data of the first user is extracted, and the height of the first user is marked as , the height of the first The user's sitting height is marked as , will the The user's leg length is marked as , will the The user's arm span is marked as ; S12, Calculate the first Relative length of the torso of each user Its expression is as follows:
[0022] S13, Calculate the... Relative length of each user's legs Its expression is as follows:
[0023] S14, Calculate the... The user's arm span to height ratio Its expression is as follows:
[0024] S15, according to the... The age of the user is calculated. Individual user difference index Its expression is as follows: If the first The age of each user is less than 60 years old.
[0025] In the formula, This represents the standard value used to measure the relative length of the torso in a motion-sensing interaction model applicable to users under 60 years old. This indicates the weighting of the absolute difference between the standard value and the relative length of the torso. This refers to the standard value used to measure the relative length of legs in a motion-sensing interaction model applicable to users under 60 years old. This indicates the weighting of the absolute difference between the standard value and the relative leg length. This represents the standard value used to measure the arm span to height ratio in a motion-sensing interaction model applicable to users under 60 years old. This indicates the weighting of the absolute difference between the standard value and the arm span-to-height ratio. , and All are constants. , Indicates according to , and Weights, calculated to obtain the first Individual user difference index ; If the age of the first user is ≥ 60 years old,
[0026] In the formula, represents a standard value for measuring the relative length of the torso in the body-sensation interaction model applicable to users aged 60 years old and above, represents a weight for the absolute difference between the standard value and the relative length of the torso, represents a standard value for measuring the relative length of the leg in the body-sensation interaction model applicable to users aged 60 years old and above, represents a weight for the absolute difference between the standard value and the relative length of the leg, represents a standard value for measuring the arm span to height ratio in the body-sensation interaction model applicable to users aged 60 years old and above, represents a weight for the absolute difference between the standard value and the arm span to height ratio, , and are constants, , represents a difference index of the first user calculated according to the weights , and . .
[0027]
[0028] Table 1 Difference Index Experimental Data In the table, user 1, user 2, and user 3 are selected as experimental objects, and the standard value for measuring the relative length of the torso in the body-sensation interaction model applicable to users under 60 years old is the standard value for measuring the relative length of the leg is the standard value for measuring the arm span to height ratio is , , , the standard value for measuring the relative length of the torso in the body-sensation interaction model applicable to users aged 60 years old and above is the standard value for measuring the relative length of the leg is the standard value for measuring the arm span to height ratio is , , , . The difference threshold is set to 0.07, and it is determined that the difference indexes of user 1 and user 2 have both exceeded the difference threshold This indicates that User 1 and User 2 differ significantly from the motion-sensing interaction model for users under 60 years old, and the motion-sensing interaction model should be reconstructed based on the user's age, height, sitting height, leg length, and arm span.
[0029] In this embodiment, due to age differences, there are significant differences in the body characteristics of adults and the elderly, which directly affects the accuracy of the motion-sensing interaction model. By analyzing the differences between each user and the motion-sensing interaction model, the system can automatically update the model, thereby solving the problem that the general model cannot meet individual assistance needs.
[0030] Example 3 Please refer to Table 2 for the proficiency index experimental data. This example is based on the explanation of Example 1. Specifically, the proficiency index... The calculation process is as follows: S21. Based on the practice dataset, extract the first... Each user owns teaching courses. The video demonstrates imitation exercises, and the first... Individual User Imitation Teaching Course The total number of practice sessions is marked as ; S22. Extract teaching courses based on the course dataset. All action points in the game, and marked them in chronological order from earliest to latest. Then use video editing software to time the action precisely. to Extract teaching courses in sequence Screenshots of the motion-sensing interaction model; S23. Using video processing software, follow the action beats. to , and extract the first one in sequence The user Screenshots of the movements from the practice video. ; S24, Follow the action timings to By using desaturation and multiply blending modes, the results are compared sequentially. The user Screenshots of the movements and instructional lessons from the practice video. Screenshots of the motion-sensing interaction model are used to generate corresponding checkpoint scores. The judgment process is as follows: The first The user The practice video includes screenshots of individual movements and instructional lessons. In the process, the screenshot of a single action of the motion-sensing interaction model is desaturated to obtain two black and white images, and then the two black and white images are overlaid using the multiply blending mode. If no color change occurs after overlaying two black and white images, it indicates that the first image is the first black and white image. The user In this practice video, screenshots of individual movements and instructional lessons. In the middle, the single action of the motion-sensing interaction model is captured. Figure One To, the If the range of motion of a single user meets the standard, the first user will be... The user The timing score for the practice video Add 1 point; If a color change occurs after two black and white images are superimposed, it indicates that the first image is the first black and white image. The user In this practice video, screenshots of individual movements and instructional lessons. In the process, the screenshots of a single action in the motion-sensing interaction model are inconsistent. If a user's single action does not meet the standard, the [number]th [action] will be [action]. The user The timing score for the practice video Add 0 points; Specifically, after superimposing two black and white images, the difference areas will change color. For example, the original white parts may turn gray or black, which can intuitively show the difference between the two images. Furthermore, through an objective quantitative scoring mechanism, the bias of subjective evaluation is effectively avoided. S25, Calculate the first Standard deviation of all user score points Its expression is as follows:
[0031]
[0032] In the formula, Indicates the first The average score of all checkpoints for each user. Indicates the first The user The scoring for hitting the beats in the practice video. ; S26. Extract teaching courses based on the course dataset. All the positioning points in the data are marked sequentially from earliest to latest according to their appearance time. , The timestamp indicating the location point is then used in video processing software to analyze the data according to the location point. to Obtain the teaching courses in sequence In the context of the motion-sensing interaction model, the coordinates of the positioning point are shown. S27. Based on the practice dataset, extract the first... The first user's This is a practice video. Using video processing software, the video points are captured sequentially from earliest to latest, according to their appearance time. The user The coordinates of the positioning points in the practice video, and the teaching course. The coordinates of the positioning point in the motion-sensing interaction model and the first The user The coordinates of the positioning points in this practice video all originate from the same three-dimensional coordinate system; S28. According to the positioning point to Compare the first one in turn The user The positioning points and teaching courses in this practice video The positioning points of the motion-sensing interaction model are used to generate corresponding positioning scores. The judgment process is as follows: If the first The user The coordinates of all positioning points in this practice video and the teaching course If all the coordinates of the positioning points in the motion-sensing interaction model are consistent, and the timestamps of the positioning points appearing in the time dimension are consistent, then it indicates that the [number]th [position] is [position]. The user The completeness of the movements in the second practice session has reached the standard. The user Positioning score for the practice video It is recorded as 1 point; If the first The user The coordinates of any point in the practice video are related to the teaching course. If all the coordinates of the positioning points in the motion-sensing interaction model are consistent, or if the timestamps of the positioning points appearing are inconsistent in the time dimension, then it indicates that the first... The user The completeness of the movement in the first practice session did not meet the standard. The user Positioning score for the practice video Recorded as -1 point; S29. Based on S21, S26-S28, the cumulative number of... Location rating of all practice videos for each user and will the Individual user location score The sum is marked as ; S30、calculating the proficiency index of the first user for the teaching course , which is expressed as follows:
[0033] In the formula, represents the recommended number of exercises for the teaching course, represents the weight for the ratio of the total number of exercises to the recommended number of exercises, represents the weight for the average score of the checkpoints, represents the standard value for measuring the standard deviation of the scores of the checkpoints of the user, represents the weight for the ratio of the standard value to the standard deviation of the scores of the checkpoints of the user, represents the average value of the positioning scores of the first user, represents the weight for the average value of the positioning scores, , , and are constants, and , , and are weights, , represents the proficiency index of the first user for the teaching course , , and calculated according to the weights.
[0034]
[0035] Table 2: Proficiency index experimental data In the table, users 4, 5 and 6 who practice the teaching course at the same time are selected as experimental objects, and the total number of exercises of users 4, 5 and 6 is 5. In the exercise video of user 4, the scores of the 5 checkpoints are 5, 5, 6, 8 and 10, respectively, and the positioning scores are -1, -1, -1, -1 and 1, respectively. In the exercise video of user 5, the scores of the 5 checkpoints are 6, 6, 6, 7 and 8, respectively, and the positioning scores are -1, 1, 1, 1 and 1, respectively. In the exercise video of user 6, the scores of the 5 checkpoints are 5, 5, 6, 7 and 8, respectively, and the positioning scores are -1, -1, -1, -1 and 1, respectively. 3, 1, 3, 6 and 4, respectively, 5 positioning scores -1, -1, -1, -1 and -1, respectively, collecting teaching courses through a network connection big data platform recommended exercise times standard value for measuring the standard deviation of user card point scoring criteria , , , , ; proficiency threshold is set to 1.5, and it is determined that the proficiency indices of user 4 and user 6 are both lower than the proficiency threshold , indicating that user 4 and user 6 have not yet mastered the movements in the teaching course, and the teaching speed of the course should be reduced, the user's movement amplitude should be optimized, and the user should be advised to increase the exercise frequency. In this embodiment, the proficiency is automatically scored by taking screenshots to segment the motion card points, frame-by-frame comparison with the standard motion, which not only realizes the objective quantification of the scoring results, but also accurately identifies the error points in continuous movements, thereby providing effective assistance to the user when practicing the teaching course and ensuring the standardization of the operation.
[0036] Embodiment 4
[0037] Please refer to Table 3 for conversion index experimental data. This embodiment is an explanation and description based on embodiment 1. Specifically, the conversion index calculation process is as follows: S31, according to the course data set, extract the teaching course and the management data of the teaching course , and mark the duration of the teaching course as , mark the total number of motion card points of the teaching course as , mark the duration of the teaching course as , and mark the total number of motion card points of the teaching course as ; S32, calculate the card point frequency of the teaching course and the card point frequency of the teaching course , which are expressed as follows:
[0038]
[0039] S33, calculating the teaching course with the teaching course length difference , the expression is as follows:
[0040] S34, calculating the user-adapted teaching course of the practice teaching course The conversion index when , the expression is as follows:
[0041] In the formula, represents the teaching course card frequency and the absolute difference of the teaching course card frequency, represents the amount of change in card frequency per unit time, that is, the conversion index of the user-adapted teaching course of the practice teaching course .
[0042]
[0043] Table 3 Conversion index experimental data In the table, users 7, 8 and 9 with the same age are selected as experimental objects; The conversion threshold is set to 0.04, and it is judged that the conversion indexes of users 8 and 9 have exceeded the conversion threshold , indicating that the teaching courses currently practiced by users 8 and 9 and the adapted teaching courses have significant differences in action rhythm, and teaching courses with gentle changes in action rhythm should be re-adapted.
[0044] In this embodiment, by quantifying the course action rhythm, it is ensured that the user realizes gradual progress in the multi-course connection practice process, and avoids movement injuries caused by excessive action intensity.
[0045] The size of the threshold is set for easy comparison, and the size of the threshold depends on the amount of sample data and the base number set by the person skilled in the art for each group of sample data; as long as it does not affect the proportional relationship between the parameters and the quantized values.
[0046] The above formulas are obtained by collecting a large amount of data for software simulation and selecting a formula close to the true value, and the coefficients in the formula are set by the person skilled in the art according to the actual situation. The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can make equivalent replacement or change according to the technical scheme and the inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.
Claims
1. A learning machine interactive teaching aid system, characterized in that: Includes a multi-dimensional data acquisition module and a teaching support module; The multi-dimensional acquisition module consists of a user management unit, a learning management unit, and a course management unit. The user management unit collects user datasets by connecting to the learning machine via a network. The user datasets include management data for all users. The learning management unit collects practice datasets by connecting to a high-definition camera via a network. The practice datasets include practice videos for all users. The course management unit collects course datasets by connecting to a big data platform via a network. The course datasets include management data for all teaching courses. The teaching support module consists of a user assessment unit, a learning assessment unit, an interaction assessment unit, and an auxiliary management unit. The user assessment unit analyzes the differences between each user and the motion-sensing interaction model based on the user dataset and generates a corresponding difference index. The learning assessment unit identifies the user's action points and positioning points in the practice videos based on the practice dataset and course dataset, and then analyzes each user's proficiency with the teaching course to generate a corresponding proficiency index. The interactive evaluation unit analyzes the differences in movement rhythm between different courses based on the course dataset and generates corresponding conversion indices. The auxiliary management unit is set with a fixed difference threshold value. Proficiency threshold and conversion threshold Combined with the difference index Proficiency Index and conversion index Provide corresponding supplementary teaching suggestions.
2. The interactive teaching aid system for learning machines according to claim 1, characterized in that: The user dataset includes each user's age, height, sitting height, leg length, and arm span.
3. The interactive teaching aid system for learning machines according to claim 2, characterized in that: The practice dataset includes practice videos for each user, where each practice video contains several screenshots of the actions.
4. The interactive teaching aid system for learning machines according to claim 3, characterized in that: The course dataset includes the name, action points, location points, and duration of each teaching course.
5. The interactive teaching aid system for learning machines according to claim 4, characterized in that: The difference index The calculation process is as follows: S11. Based on the user dataset, extract the first... The management data of the first user, and the first user's management data ... The user's height is marked as , will the The user's sitting height is marked as , will the The user's leg length is marked as , will the The user's arm span is marked as ; S12, Calculate the first Relative length of the torso of each user ; S13, Calculate the... Relative length of each user's legs ; S14, Calculate the... The user's arm span to height ratio ; S15. Collect standard values for measuring the relative length of the torso from a motion-sensing interaction model suitable for users under 60 years old via a network-connected learning machine. Standard values used to measure the relative length of the legs Standard values used to measure the arm span to height ratio And in a motion-sensing interaction model applicable to users aged 60 and above, the standard value used to measure the relative length of the torso. Standard values used to measure the relative length of the legs Standard values used to measure the arm span to height ratio Then, according to S11-S14, in accordance with the first... The age of the user is calculated using different weighting methods. Individual user difference index .
6. The interactive teaching aid system for learning machines according to claim 5, characterized in that: The proficiency index The calculation process is as follows: S21. Based on the practice dataset, extract the first... Each user owns teaching courses. The video demonstrates imitation exercises, and the first... Individual User Imitation Teaching Course The total number of practice sessions is marked as ; S22. Extract teaching courses based on the course dataset. All action points in the game, and marked them in chronological order from earliest to latest. Then use video editing software to time the action precisely. to Extract teaching courses in sequence Screenshots of the motion-sensing interaction model; S23. Using video processing software, follow the action beats. to , and extract the first one in sequence The user Screenshots of the movements from the practice video. ; S24, Follow the action timings to By using desaturation and multiply blending modes, the results are compared sequentially. The user Screenshots of the actions and instructional lessons in the practice video. Screenshots of the motion-sensing interaction model are used to generate corresponding checkpoint scores. The judgment process is as follows: The first The user The practice video includes screenshots of individual movements and instructional lessons. In the process, the screenshot of a single action of the motion-sensing interaction model is desaturated to obtain two black and white images, and then the two black and white images are overlaid using the multiply blending mode. If no color change occurs after overlaying two black and white images, it indicates that the first image is the first black and white image. The user In this practice video, screenshots of individual movements and instructional lessons. In the middle, the screenshots of a single action of the motion-sensing interaction model are consistent, the first... The magnitude of a single user's action has met the standard, and the [number]th [action] will be [action]. The user The timing score for the practice video Add 1 point; If a color change occurs after two black and white images are superimposed, it indicates that the first image is the first black and white image. The user In this practice video, screenshots of individual movements and instructional lessons. In the process, the screenshots of a single action in the motion-sensing interaction model are inconsistent. If a user's single action does not meet the standard, the [number]th [action] will be [action]. The user The timing score for the practice video Add 0 points; S25, the first The user The timing score for this practice video is marked as follows: , Then calculate the first Standard deviation of all score points for each user ; S26. Extract teaching courses based on the course dataset. All the positioning points in the data are marked sequentially from earliest to latest according to their appearance time. , The timestamp indicating the location point is then used in video processing software to analyze the data according to the location point. to Obtain the teaching courses in sequence In the context of the motion-sensing interaction model, the coordinates of the positioning point are shown. S27. Based on the practice dataset, extract the first... The first user's This is a practice video. Using video processing software, the video points are captured sequentially from earliest to latest, according to their appearance time. The user The coordinates of the positioning points in the practice video, and the teaching course. The coordinates of the positioning point in the motion-sensing interaction model and the first The user The coordinates of the positioning points in this practice video all originate from the same three-dimensional coordinate system; S28. According to the positioning point to Compare the first one in turn The user The positioning points and teaching courses in this practice video The positioning points of the motion-sensing interaction model are used to generate corresponding positioning scores. The judgment process is as follows: If the first The user The coordinates of all positioning points in this practice video and the teaching course If all the coordinates of the positioning points in the motion-sensing interaction model are consistent, and the timestamps of the positioning points appearing in the time dimension are consistent, then it indicates that the [number]th [position] is [position]. The user The completeness of the movements in the second practice session has reached the standard. The user Positioning score for the practice video It is recorded as 1 point; If the first The user The coordinates of any point in the practice video are related to the teaching course. If all the coordinates of the positioning points in the motion-sensing interaction model are consistent, or if the timestamps of the positioning points appearing are inconsistent in the time dimension, then it indicates that the first... The user The completeness of the movement in the first practice session did not meet the standard. The user Positioning score for the practice video Recorded as -1 point; S29. Based on S21, S26-S28, the cumulative number of... Location rating of all practice videos for each user and will the Individual user location score The sum is marked as ; S30. Collect teaching courses by connecting to a big data platform via the network. Recommended number of practice sessions and the standard value used to measure the standard deviation of user checkpoint scores Then, based on S21-S29, the weighted average is used to calculate the... Individual users on teaching courses Proficiency index .
7. The interactive teaching aid system for learning machines according to claim 6, characterized in that: The conversion index The calculation process is as follows: S31. Extract teaching courses based on the course dataset. and teaching courses Management data, and teaching courses The duration is marked as The teaching courses The total number of action checkpoints is marked as The teaching courses The duration is marked as The teaching courses The total number of action checkpoints is marked as ; S32, Computation Teaching Course frequency of card points and teaching courses frequency of card points ; S33, Computation Teaching Course With teaching courses Time difference ; S34. Based on S31-S33, calculate the practice teaching course. User adaptation tutorial courses Conversion index at time .
8. The interactive teaching aid system for learning machines according to claim 7, characterized in that: The difference index ≥ Difference Threshold If the difference between the user and the motion-sensing interaction model is significant, the motion-sensing interaction model should be reconstructed based on the user's age, height, sitting height, leg length, and arm span.
9. The interactive teaching aid system for learning machines according to claim 8, characterized in that: The proficiency index <Proficiency Threshold If the user has not yet mastered the movements in the course, the teaching speed should be reduced, the user should be assisted in optimizing the range of motion, and the user should be advised to increase the frequency of practice.
10. The interactive teaching aid system for learning machines according to claim 9, characterized in that: The conversion index >Conversion threshold If this occurs, it indicates that there is a significant difference in the rhythm of the movements between the user's current practice course and the adapted course, and the user should be re-adapted to a course with a gentler rhythm.