An online teaching interaction system based on a cloud platform

Abstract

The application relates to the technical field of intelligent interaction, in particular to an online teaching interaction system based on a cloud platform, wherein a data acquisition module is used to determine the action disturbance route complexity and action repetition degree of a teaching end; a data analysis module is used to calculate an action complexity representation parameter and divide an action abnormality tendency; an interaction processing module is used to compare the action start frame and the action end frame of each node of the teaching end video and the learning end video for weak abnormality tendency, determine the learning end action state, determine a plurality of extraction intervals for strong abnormality tendency actions, compare the corresponding predicted node action graph of the extraction interval with the actual node action graph in the learning end video frame in the extraction interval, and determine the learning end action state; and a state reminding module is used to determine whether a state correction reminding needs to be sent. The application compares and analyzes learning actions, and improves the interaction efficiency and analysis accuracy.

Description

Technical Field

[0001] This invention relates to the field of intelligent interactive technology, and in particular to an online teaching interactive system based on a cloud platform. Background Technology

[0002] Online teaching interaction systems utilize network, multimedia, and software technologies to break down the time and space limitations of traditional teaching, enabling students and teachers to conduct remote teaching and interaction via the internet. In particular, cloud computing architecture, with its powerful computing and storage capabilities, ensures stable system operation, supporting massive data storage and flexible resource allocation. Artificial intelligence algorithms, by mining learning data, provide intelligent assistance in teaching and Q&A, offering comprehensive and robust technical support for cloud-based online teaching interaction systems. Simultaneously, object detection and image classification technologies can identify specific targets in images, supporting teaching decisions. The development of online teaching interaction systems has not only improved teaching efficiency and quality but also provided students with a more personalized and flexible learning experience.

[0003] Chinese Patent Publication No. CN114863433A discloses a real-time handwriting interactive online auxiliary teaching system, comprising: an extraction module for extracting the hand skeleton from real-time live images captured by a camera using deep learning methods to obtain the coordinates of key hand points; a gesture recognition module for recognizing gestures using the coordinates of key hand points and switching modes based on the gesture recognition results, the modes including: character input, basic graphic drawing, and sketch drawing; a character recognition module for recognizing handwritten characters input by the user and obtaining the corresponding printed characters; a graphic drawing module for recognizing basic graphics drawn by the user and obtaining the basic graphics after trajectory smoothing; and an image generation module for generating and visualizing an image based on the sketch drawn by the user. This invention can solve the problem of missing whiteboard writing in online classes.

[0004] Chinese Patent Publication No. CN118656485A discloses an online paperless language teaching interaction method, medium, and device. It constructs a basic database for each student, and based on this database, generates a current stage learning database, enabling differentiated instruction for students at different learning levels. It constructs error-prone vocabulary and grammar groups based on incorrect answers, allowing for multi-dimensional exploration of these errors. A first set of questions is extracted from the error-prone vocabulary group, and a second set from the error-prone grammar group, and these are inserted into the question combinations for the next round. This strengthens the training of incorrect answers in the next round, and not only focuses on incorrect answers, but also includes questions of varying difficulty levels in the next round's question combinations. This considers the difficulty distribution of the question combinations in each round of training for the student, making the generation of question combinations dynamic and diverse, thus improving learning efficiency.

[0005] However, the following problems still exist in the existing technology.

[0006] In existing technologies, online courses such as dance and fitness instruction are mostly in the form of one teaching terminal corresponding to several learning terminals. When monitoring each learning terminal, a traversal analysis of video data is often used, which consumes a lot of computing power when there are many learning terminals. Summary of the Invention

[0007] To address this issue, the present invention provides an online teaching and interaction system based on a cloud platform, which solves the problem that in the existing technology, online courses such as dance and fitness teaching are mostly in the form of one teaching terminal corresponding to several learning terminals. When monitoring each learning terminal, the method of traversing and analyzing video data is often used, which consumes a lot of computing power when there are many learning terminals.

[0008] To achieve the above objectives, the present invention provides an online teaching interaction system based on a cloud platform, comprising:

[0009] The data acquisition module acquires authorized teaching videos and several learning videos to determine the complexity of motion perturbation routes and the repetition of motions based on the teaching videos.

[0010] The data analysis module, which is connected to the data acquisition module, is used to calculate motion complexity representation parameters based on the motion disturbance route complexity and the motion repetition, and to classify motion variation tendencies.

[0011] An interactive processing module, connected to both the data acquisition module and the data analysis module, analyzes the learning video in response to the classification results of action variation trends.

[0012] To address the tendency for weak mutations, the action state on the learning end is determined by comparing the key points of the corresponding action start and end frames in the teaching and learning end videos.

[0013] To address the strong tendency of variation, several reference intervals are determined based on the action start frame corresponding to the teaching end video, and the delay time of the reference interval relative to the time of the action start frame is determined. Several extraction intervals are determined based on the action start frame of the teaching end video and each corresponding delay time. The predicted joint point action map corresponding to the extraction interval is predicted and compared with the actual joint point action map in the learning end video frame within the extraction interval to determine the action state of the learning end.

[0014] The status reminder module, which is connected to the interaction processing module, is used to determine whether a status correction reminder needs to be issued based on the action status of the learning terminal.

[0015] Furthermore, the data acquisition module is used to determine the complexity of the motion disturbance route and the repetition rate of the motion, including,

[0016] Used to record the number of times the direction of each joint point changes, the number of times the action is repeated, and the duration of the teaching video;

[0017] The ratio of the number of times the direction of each joint point changes to the duration is used as the change frequency.

[0018] The average value of each ratio corresponding to each joint is used to determine the motion disturbance path complexity.

[0019] The ratio of the number of times the action is repeated to the duration of the action is the action repeatability.

[0020] The joints mentioned include head joints and limb joints.

[0021] Furthermore, the data analysis module is used to calculate the complex characterization parameters of the action, including,

[0022] The ratio of the complexity of the motion disturbance path to the complexity of the baseline motion disturbance path is used as the first influencing factor;

[0023] The ratio of baseline motion repeatability to motion repeatability is used to determine the second influencing factor;

[0024] The weighted sum of the first and second impact factors is used to determine the action complexity characterization parameter.

[0025] Furthermore, the data analysis module is used to classify the tendency of behavioral variation, wherein,

[0026] If the motion complexity representation parameter is greater than the motion complexity representation parameter threshold, then the motion variation tendency is classified as a strong variation tendency.

[0027] If the motion complexity representation parameter is less than or equal to the motion complexity representation parameter threshold, then the motion variation tendency is classified as a weak variation tendency.

[0028] Furthermore, the interactive processing module is used to compare the key points of the corresponding action start frame and action end frame in the teaching end video and the learning end video, including,

[0029] Used to extract joint point diagrams from the start and end frames of actions in the teaching video and the learning video;

[0030] This is used to place the joint point map corresponding to the starting frame of each action in the same coordinate system, and to scale the joint point map corresponding to the learning end video until it meets the stacking error standard, and record the current scaling ratio.

[0031] This is used to place the joint point diagrams corresponding to the end frames of each action in the same coordinate system, and to scale the joint point diagrams corresponding to the learning video based on the scaling ratio.

[0032] Used to compare the joint point maps corresponding to the end frames of each action to determine the average distance difference of each joint point.

[0033] Furthermore, the interactive processing module is used to compare the key points of the corresponding action start frames and action end frames in the teaching end video and the learning end video to determine the action state of the learning end.

[0034] If the average distance difference is greater than the preset average distance difference standard threshold, then the learning end action state is determined to be a non-standard action state.

[0035] If the average distance difference is less than or equal to the preset average distance difference standard threshold, then the learning end action state is determined to be the standard action state.

[0036] Furthermore, the interactive processing module is used to determine several reference intervals based on the starting frame of the corresponding action in the teaching video, including,

[0037] Used to determine the start time of the corresponding action in the teaching video;

[0038] This is used to delay the current time, determine several reference times, and determine several reference intervals starting from the reference times, so as to record the delay time corresponding to each reference interval.

[0039] The starting time of the reference interval is the reference time, and the interval length is the predetermined duration.

[0040] Furthermore, the interactive processing module is used to determine several extraction intervals based on the action start frame and corresponding delay duration of the teaching terminal video, and predict the motion map of the corresponding key points for each extraction interval, including...

[0041] Determine the start time of the action frame in the teaching video, delay the time by a corresponding delay duration, determine several extraction times, and determine several extraction intervals starting from the extraction times.

[0042] The correspondence between each extraction interval and the reference interval is determined based on the delay duration;

[0043] Determine the reference interval corresponding to the extraction interval, extract any teaching terminal video frame within the reference interval, identify the joint motion map, and obtain the predicted joint motion map corresponding to the extraction interval.

[0044] The starting time of the extraction interval is the extraction time, and the interval length is the predetermined duration.

[0045] Furthermore, the interactive processing module is used to compare the predicted keypoint motion map with the actual keypoint motion map in the learning end video frames within the extraction interval, including,

[0046] By selecting any video frame from the learning end within each extraction interval, the actual joint motion map is determined and compared with the corresponding predicted joint motion map to determine the degree of overlap of the joints corresponding to each extraction interval.

[0047] Determine whether each sampling interval meets the verification criteria;

[0048] The verification criterion is that the overlap of the joints corresponding to the extracted interval is greater than or equal to the preset joint overlap threshold.

[0049] Furthermore, the interaction processing module is used to determine the action state of the learning end, including,

[0050] Determine the proportion of sampled intervals that meet the verification criteria;

[0051] If the ratio is greater than or equal to a predetermined ratio threshold, then the learning end action state is determined to be the standard action state;

[0052] If the ratio is less than a predetermined ratio threshold, the learning terminal action state is determined to be a non-standard action state.

[0053] Compared with existing technologies, this invention includes a data acquisition module to determine the complexity of the motion disturbance path and the repetition rate of the motion on the teaching end; a data analysis module to calculate motion complexity representation parameters and classify motion variation tendencies; an interactive processing module to determine the motion state on the learning end by comparing the start and end frames of the motion in the teaching end video and the learning end video for weak variation tendencies; and for motions with strong variation tendencies, it identifies several extraction intervals, predicts the motion diagrams of the predicted joint points corresponding to the extraction intervals, and compares them with the actual joint point motion diagrams in the learning end video frames within the extraction intervals to determine the motion state on the learning end. A status reminder module is used to determine whether a status correction reminder needs to be issued. This invention, through comparative analysis of learning motions, ensures the accuracy of motion analysis while saving computational resources.

[0054] In particular, by calculating the motion complexity representation parameters of the teaching end, the motion variation tendency is classified, and the learning end video is analyzed in a targeted manner for different variation tendencies. In reality, the learning end is mostly multiple ports. If the motion variation tendency is not distinguished and the motion is analyzed indiscriminately, simple motions will consume too much computing power, while complex motions may not receive enough resource support, thus reducing the efficiency of motion analysis. For example, the upward straightening motion can be analyzed by a simple method, but if the same analysis method is used as for complex motions, it will cause computational overflow and waste computing power. Based on this, the present invention pre-calculates the motion complexity representation parameters of the teaching end and determines different analysis methods for strong and weak variation tendencies to improve the efficiency and accuracy of motion analysis.

[0055] In particular, for actions with a weak tendency to change, the action state of the learning end is determined by comparing the key points of the start and end frames of the actions on the teaching end and the learning end. In practice, simple actions have a weak tendency to change, especially for vertical actions. By determining the start and end points, the accuracy of the student's actions can be determined from the action data on the teaching end. This allows the system to decide whether to issue a status correction reminder to the learning end, thereby improving the efficiency and accuracy of action analysis.

[0056] In particular, for movements with a strong tendency to mutate, the predicted joint motion diagrams corresponding to the extracted intervals are compared with the actual joint motion diagrams in the learning end video frames within the extracted intervals to determine the movement state of the learning end. In reality, complex movements have a strong tendency to mutate, and the probability of non-standard movements increases during learning. If the same analysis method as for movements with a weak tendency to mutate is used, it may lead to analysis errors, thus affecting the judgment result of the movement state of the learning end. Based on this, the present invention combines the movement video segments of the teaching end to determine several extracted intervals, predicts the predicted joint motion diagrams corresponding to the extracted intervals, and compares them with the actual joint motion diagrams in the learning end video frames within the extracted intervals to improve the efficiency and accuracy of movement analysis. Attached Figure Description

[0057] Figure 1 A schematic diagram of the structure of an online teaching interaction system based on a cloud platform, as an embodiment of the invention;

[0058] Figure 2 A logic block diagram for classifying the tendency of action variation in an embodiment of the invention;

[0059] Figure 3 A logical block diagram for determining the action state of the learning terminal by comparing the key points of the corresponding action start frame and action end frame of the teaching terminal video and the learning terminal video in the embodiment of the invention.

[0060] Figure 4This is a logic block diagram for determining the action state of the learning terminal according to an embodiment of the invention.

[0061] Figure 5 This is a logic block diagram illustrating how a status correction reminder is needed based on the learning terminal's action state, according to an embodiment of the invention. Detailed Implementation

[0062] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0063] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0064] It should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0065] Please see Figure 1 , Figure 1 This is a schematic diagram of the structure of a cloud-based online teaching interaction system according to an embodiment of the invention. The cloud-based online teaching interaction system according to an embodiment of the invention includes:

[0066] The data acquisition module acquires authorized teaching videos and several learning videos to determine the complexity of motion perturbation routes and the repetition of motions based on the teaching videos.

[0067] The data analysis module, which is connected to the data acquisition module, is used to calculate motion complexity representation parameters based on the motion disturbance route complexity and the motion repetition, and to classify motion variation tendencies.

[0068] An interactive processing module, connected to the data acquisition module and the data analysis module, analyzes the learning video in response to the classification results of action variation trends.

[0069] To address the tendency for weak mutations, the action state on the learning end is determined by comparing the key points of the corresponding action start and end frames in the teaching and learning end videos.

[0070] To address the strong tendency of variation, several reference intervals are determined based on the action start frame corresponding to the teaching end video, and the delay time of the reference interval relative to the time of the action start frame is determined. Several extraction intervals are determined based on the action start frame of the teaching end video and each corresponding delay time. The predicted joint point action map corresponding to the extraction interval is predicted and compared with the actual joint point action map in the learning end video frame within the extraction interval to determine the action state of the learning end.

[0071] The status reminder module, which is connected to the interaction processing module, is used to determine whether a status correction reminder needs to be issued based on the action status of the learning terminal.

[0072] Specifically, the teaching end is the terminal used by the teacher, and the learning end is the terminal used by the learner. There are no restrictions on the method of obtaining videos on the teaching end and the learning end. For example, it can be obtaining authorized video data uploaded by the teaching end or the learning end, which will not be elaborated further.

[0073] Understandably, motion perturbation path complexity characterizes the complexity of the joint motion path. The higher the motion perturbation path complexity, the greater the motion complexity. Motion repeatability characterizes the probability of repeated actions in the whole set of actions. The lower the motion repeatability, the greater the motion complexity.

[0074] Specifically, the data acquisition module is used to determine the complexity of the motion disturbance path and the repetition rate of the motion, including,

[0075] Used to record the number of times the direction of each joint point changes, the number of times the action is repeated, and the duration of the teaching video;

[0076] The ratio of the number of times the direction of each joint point changes to the duration is used as the change frequency.

[0077] The average value used to determine the change frequency corresponding to each key point is the motion disturbance path complexity.

[0078] The ratio of the number of times the action is repeated to the duration of the action is the action repeatability.

[0079] The joints mentioned include head joints and limb joints.

[0080] Specifically, the steps for determining the number of direction changes are as follows.

[0081] Calculate the positional changes of the key points at each time point relative to the previous time point to obtain several direction vectors;

[0082] If it is determined that there is an angle between adjacent direction vectors that is greater than a predetermined angle threshold, it is recorded as a direction change.

[0083] The predetermined included angle threshold is selected within the range [15°, 30°].

[0084] Specifically, the data analysis module is used to calculate parameters representing the complexity of actions, including,

[0085] The ratio of the complexity of the motion disturbance path to the complexity of the baseline motion disturbance path is used as the first influencing factor;

[0086] The ratio of baseline motion repeatability to motion repeatability is used to determine the second influencing factor;

[0087] The weighted sum of the first and second impact factors is used to determine the action complexity characterization parameter.

[0088] Specifically, the baseline motion disturbance route complexity is calculated in advance. Several teaching videos are obtained in advance to determine the complexity of several motion disturbance routes, and the average of the complexity of each motion disturbance route is determined as the baseline motion disturbance route complexity.

[0089] Specifically, the baseline repeatability is calculated in advance by acquiring several teaching videos to determine several repeatability values, and the average of the repeatability values ​​is determined as the baseline repeatability.

[0090] Specifically, the sum of the weighting coefficients of the first impact factor and the second impact factor is 1, the weighting coefficient of the first impact factor is 0.65, and the weighting coefficient of the second impact factor is 0.35.

[0091] Please see Figure 2 , Figure 2 This is a logic block diagram illustrating the classification of action variation tendencies according to an embodiment of the invention. Specifically, the data analysis module is used to classify action variation tendencies, wherein...

[0092] If the motion complexity representation parameter is greater than the motion complexity representation parameter threshold, then the motion variation tendency is classified as a strong variation tendency.

[0093] If the motion complexity representation parameter is less than or equal to the motion complexity representation parameter threshold, then the motion variation tendency is classified as a weak variation tendency.

[0094] Specifically, the threshold for the complex representation parameter of a movement is calculated using historical data. Several teaching videos are pre-acquired to determine several complex representation parameters of a movement, and the average value of each complex representation parameter is determined as the threshold for the complex representation parameter of a movement.

[0095] Understandably, the threshold value for the complex representation parameter of the movement is a periodically changing value, and it needs to be recalculated at each change cycle to ensure the accuracy of the classification of movement variation tendency. In this implementation, the change cycle is determined to be 15 videos from the teaching end.

[0096] Specifically, by calculating the motion complexity representation parameters of the teaching end, motion variation tendencies are classified, and the learning end videos are analyzed in a targeted manner for different variation tendencies. In reality, the learning end is mostly multiple ports. If motion variation tendencies are not distinguished and motions are analyzed indiscriminately, simple motions will consume too much computing power, while complex motions may not receive sufficient resource support, thus reducing motion analysis efficiency. For example, a motion of stretching upwards can be analyzed using a simple method, but if the same analysis method is used for complex motions, it will cause computational overflow and waste computing power. Based on this, the present invention pre-calculates the motion complexity representation parameters of the teaching end and determines different analysis methods for strong and weak variation tendencies to improve motion analysis efficiency and accuracy.

[0097] Specifically, the interaction processing module is used to compare the key points of the corresponding action start frames and action end frames in the teaching video and the learning video, including,

[0098] Used to extract joint point diagrams from the start and end frames of actions in the teaching video and the learning video;

[0099] This is used to place the joint point map corresponding to the starting frame of each action in the same coordinate system, and to scale the joint point map corresponding to the learning end video until it meets the stacking error standard, and record the current scaling ratio.

[0100] This is used to place the joint point diagrams corresponding to the end frames of each action in the same coordinate system, and to scale the joint point diagrams corresponding to the learning video based on the scaling ratio.

[0101] Used to compare the joint point diagrams corresponding to the end frames of each action to determine the average distance difference between each joint point.

[0102] It is understandable that there will be a difference in the scale of the joint maps between the teaching video and the learning video. If the scale is not adjusted, it may lead to errors in the judgment of the action on the learning end. Therefore, it is necessary to adjust the scale of the joint maps in the start frame and end frame of the action in the teaching video and the learning video to be consistent, and then place them on the same coordinate system for calculation.

[0103] Specifically, the stacking error standard is that the average distance difference between each joint point corresponding to the start frame of the action is less than the distance difference threshold. The distance difference threshold is determined based on the key point map corresponding to the video on the learning end. The maximum distance between each joint point in the key point map is determined, and the distance difference threshold is determined between 0.1 times and 0.15 times the maximum distance.

[0104] Please see Figure 3 , Figure 3This is a logical block diagram illustrating how the learning end's action state is determined by comparing the key points of the corresponding action start and end frames in the teaching and learning end videos, as described in this embodiment of the invention. Specifically, the interaction processing module is used to compare the key points of the corresponding action start and end frames in the teaching and learning end videos to determine the learning end's action state.

[0105] If the average distance difference is greater than the preset average distance difference standard threshold, then the learning end action state is determined to be a non-standard action state.

[0106] If the average distance difference is less than or equal to the preset average distance difference standard threshold, then the learning end action state is determined to be the standard action state.

[0107] Specifically, the average distance difference standard threshold is determined based on the joint point graph corresponding to any video frame on the teaching end. The maximum distance between each joint point in the joint point graph is determined, and the average distance difference standard threshold is set to between 0.15 times and 0.2 times the maximum distance.

[0108] Specifically, for actions with a weak tendency to change, the action state of the learning end is determined by comparing the key points of the start and end frames of the actions on the teaching end and the learning end. In practice, simple actions have a weak tendency to change, especially for vertical actions. By determining the start and end points, the accuracy of the student's actions can be determined using the action data from the teaching end. This allows the system to decide whether to issue a status correction reminder to the learning end, thereby improving the efficiency and accuracy of action analysis.

[0109] Specifically, the interactive processing module is used to determine several reference intervals based on the start frame of the corresponding action in the teaching video, including,

[0110] Used to determine the start time of the corresponding action in the teaching video;

[0111] This is used to delay the current time, determine several reference times, and determine several reference intervals starting from the reference times, so as to record the delay time corresponding to each reference interval.

[0112] The starting time of the reference interval is the reference time, and the interval length is the predetermined duration.

[0113] Specifically, no specific value is set for the delay duration. For example, in practice, the delay duration is determined to be 0.1 times the duration of the teaching video. Of course, those skilled in the art can also determine it according to the actual situation, which will not be elaborated here.

[0114] Specifically, in practice, the predetermined duration is selected within the range of [0.5s, 2s].

[0115] Specifically, the interactive processing module is used to determine several extraction intervals based on the action start frame and corresponding delay duration of the teaching video, and to predict the corresponding keypoint motion graph for each extraction interval, including...

[0116] Determine the start time of the action frame in the teaching video, delay the time by a corresponding delay duration, determine several extraction times, and determine several extraction intervals starting from the extraction times.

[0117] The correspondence between each extraction interval and the reference interval is determined based on the delay duration;

[0118] Determine the reference interval corresponding to the extraction interval, extract any teaching terminal video frame within the reference interval, identify the joint motion map, and obtain the predicted joint motion map corresponding to the extraction interval.

[0119] The starting time of the extraction interval is the extraction time, and the interval length is the predetermined duration.

[0120] It is understandable that, due to the consistency of the delay duration, there is a correspondence between the extraction interval and the reference interval, that is, the extraction interval with the same delay duration corresponds to the reference interval, which will not be elaborated further.

[0121] Specifically, the interactive processing module is used to compare the predicted joint motion map with the actual joint motion map in the learning end video frames within the extraction interval, including,

[0122] By selecting any video frame from the learning end within each extraction interval, the actual joint motion map is determined and compared with the corresponding predicted joint motion map to determine the degree of overlap of the joints corresponding to each extraction interval.

[0123] Determine whether each sampling interval meets the verification criteria;

[0124] The verification criterion is that the overlap of the joints corresponding to the extracted interval is greater than or equal to the preset joint overlap threshold.

[0125] Specifically, when solving for the joint overlap, you can first determine the distance between each joint in the two joint diagrams, calculate the average distance, and then determine the joint overlap by taking the reciprocal of the average distance.

[0126] The preset joint overlap threshold is predetermined. Specifically, joint maps of several individuals performing a single action are obtained in advance, the average joint overlap between the joint maps is calculated, and the product of the average joint overlap and the error coefficient is determined as the preset joint overlap threshold.

[0127] Please see Figure 4 , Figure 4This is a logic block diagram illustrating how to determine the action state of the learning terminal, as per an embodiment of the invention. Specifically, the interaction processing module is used to determine the action state of the learning terminal, including:

[0128] Determining the proportion of sampled intervals that meet the verification criteria can be understood as calculating the ratio of the number of sampled intervals that meet the verification criteria to the total number of sampled intervals.

[0129] If the ratio is greater than or equal to a predetermined ratio threshold, then the learning end action state is determined to be the standard action state;

[0130] If the ratio is less than a predetermined ratio threshold, the learning terminal action state is determined to be a non-standard action state.

[0131] Specifically, the purpose of setting the proportion threshold is to characterize the situation where most of the sampling intervals meet the verification criteria, so the proportion threshold is selected within the interval [0.5, 0.6].

[0132] Specifically, for movements with a strong tendency to mutate, the predicted joint motion diagrams corresponding to the extracted intervals are compared with the actual joint motion diagrams in the learning end video frames within the extracted intervals to determine the movement state of the learning end. In reality, complex movements have a strong tendency to mutate, and the probability of non-standard movements increases during learning. If the same analysis method as for movements with a weak tendency to mutate is used, it may lead to analysis errors, thus affecting the judgment result of the movement state of the learning end. Based on this, the present invention combines the movement video segments of the teaching end to determine several extraction intervals, predicts the predicted joint motion diagrams corresponding to the extracted intervals, and compares them with the actual joint motion diagrams in the learning end video frames within the extracted intervals to improve the efficiency and accuracy of movement analysis.

[0133] Please participate Figure 5 , Figure 5 This is a logic block diagram illustrating how a status correction reminder is needed based on the learning device's action state, according to an embodiment of the invention. Specifically, the status reminder module determines whether a status correction reminder needs to be issued based on the learning device's action state.

[0134] If the learning terminal's action status is a non-standard action status, the system determines that it needs to issue a status correction reminder.

[0135] If the learning terminal's action state is a standard action state, then the system determines that it does not need to issue a status correction reminder.

[0136] Status correction reminders can be sent to the learning terminal, which will not be elaborated further.

[0137] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0138] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A cloud-based online teaching interaction system, characterized in that, include: The data acquisition module acquires authorized teaching videos and several learning videos to determine the complexity of motion perturbation routes and the repetition of motions based on the teaching videos. The data analysis module, which is connected to the data acquisition module, is used to calculate motion complexity representation parameters based on the motion disturbance route complexity and the motion repetition, and to classify motion variation tendencies. An interactive processing module, connected to both the data acquisition module and the data analysis module, analyzes the learning video in response to the classification results of action variation trends. To address the tendency for weak mutations, the action state on the learning end is determined by comparing the key points of the corresponding action start and end frames in the teaching and learning end videos. To address the strong tendency of variation, several reference intervals are determined based on the action start frame corresponding to the teaching end video, and the delay time of the reference interval relative to the time of the action start frame is determined. Several extraction intervals are determined based on the action start frame of the teaching end video and each corresponding delay time. The predicted joint point action map corresponding to the extraction interval is predicted and compared with the actual joint point action map in the learning end video frame within the extraction interval to determine the action state of the learning end. A status reminder module, which is connected to the interaction processing module, is used to determine whether a status correction reminder needs to be issued based on the action status of the learning terminal. The data acquisition module is used to determine the complexity of the motion disturbance route and the repetition rate of the motion, including, Used to record the number of times the direction of each joint point changes, the number of times the action is repeated, and the duration of the teaching video; The ratio of the number of times the direction of each joint point changes to the duration is used as the change frequency. The average value used to determine the change frequency corresponding to each key point is the motion disturbance path complexity. The ratio of the number of times the action is repeated to the duration of the action is the action repeatability.

2. The online teaching interaction system based on a cloud platform according to claim 1, characterized in that, The data analysis module is used to calculate parameters representing the complexity of actions, including: The ratio of the complexity of the motion disturbance path to the complexity of the baseline motion disturbance path is used as the first influencing factor; The ratio of baseline motion repeatability to motion repeatability is used to determine the second influencing factor; The weighted sum of the first and second impact factors is used to determine the action complexity characterization parameter.

3. The online teaching interaction system based on a cloud platform according to claim 1, characterized in that, The data analysis module is used to classify the tendency of behavioral variation, wherein, If the motion complexity representation parameter is greater than the motion complexity representation parameter threshold, then the motion variation tendency is classified as a strong variation tendency. If the motion complexity representation parameter is less than or equal to the motion complexity representation parameter threshold, then the motion variation tendency is classified as a weak variation tendency.

4. The online teaching interaction system based on a cloud platform according to claim 1, characterized in that, The interactive processing module is used to compare the key points of the corresponding action start frame and action end frame in the teaching end video and the learning end video, including, Used to extract joint point diagrams from the start and end frames of actions in the teaching video and the learning video; This is used to place the joint point map corresponding to the starting frame of each action in the same coordinate system, and to scale the joint point map corresponding to the learning end video until it meets the stacking error standard, and record the current scaling ratio. This is used to place the joint point diagrams corresponding to the end frames of each action in the same coordinate system, and to scale the joint point diagrams corresponding to the learning video based on the scaling ratio. Used to compare the joint point diagrams corresponding to the end frames of each action to determine the average distance difference between each joint point.

5. The online teaching interaction system based on a cloud platform according to claim 4, characterized in that, The interactive processing module is used to compare the key points of the corresponding action start frames and action end frames in the teaching end video and the learning end video to determine the action state of the learning end. If the average distance difference is greater than the preset average distance difference standard threshold, then the learning end action state is determined to be a non-standard action state. If the average distance difference is less than or equal to the preset average distance difference standard threshold, then the learning end action state is determined to be the standard action state.

6. The online teaching interaction system based on a cloud platform according to claim 1, characterized in that, The interactive processing module is used to determine several reference intervals based on the starting frame of the corresponding action in the teaching video, including: Used to determine the start time of the corresponding action in the teaching video; This is used to delay the current time, determine several reference times, and determine several reference intervals starting from the reference times, so as to record the delay time corresponding to each reference interval. The starting time of the reference interval is the reference time, and the interval length is the predetermined duration.

7. The online teaching interaction system based on a cloud platform according to claim 6, characterized in that, The interactive processing module is used to determine several extraction intervals based on the action start frame and corresponding delay duration of the teaching terminal video, and to predict the motion map of the corresponding key points for each extraction interval, including... Determine the start time of the action frame in the teaching video, delay the time by a corresponding delay duration, determine several extraction times, and determine several extraction intervals starting from the extraction times. The correspondence between each extraction interval and the reference interval is determined based on the delay duration; Determine the reference interval corresponding to the extraction interval, extract any teaching terminal video frame within the reference interval, identify the joint motion map, and obtain the predicted joint motion map corresponding to the extraction interval. The starting time of the extraction interval is the extraction time, and the interval length is the predetermined duration.

8. The online teaching interaction system based on a cloud platform according to claim 7, characterized in that, The interactive processing module is used to compare the predicted joint motion map with the actual joint motion map in the learning end video frames within the extraction interval, including, By selecting any video frame from the learning end within each extraction interval, the actual joint motion map is determined and compared with the corresponding predicted joint motion map to determine the degree of overlap of the joints corresponding to each extraction interval. Determine whether each sampling interval meets the verification criteria; The verification criterion is that the overlap of the joints corresponding to the extracted interval is greater than or equal to the preset joint overlap threshold.

9. The online teaching interaction system based on a cloud platform according to claim 8, characterized in that, The interaction processing module is used to determine the action state of the learning terminal, including, Determine the proportion of sampled intervals that meet the verification criteria; If the ratio is greater than or equal to a predetermined ratio threshold, then the learning end action state is determined to be the standard action state; If the ratio is less than a predetermined ratio threshold, the learning terminal action state is determined to be a non-standard action state.

Images