Artificial intelligence education and learning method, system, device and storage medium

Abstract

The present application provides an artificial intelligence education and learning method, system, device and storage medium, which relates to the field of artificial intelligence technology. The method includes: obtaining the teacher's location information and the teacher's first behavior state information; combining the location information and the first behavior state information to generate the student's standard behavior state information; obtaining the target student's second behavior state information, comparing the standard behavior state information and the second behavior state information, generating a comparison result, and generating the target student's concentration score based on the comparison result; obtaining the third behavior state information of other students within a preset range of the target student, generating an adjustment index based on the third behavior state information; adjusting the concentration score by the adjustment index to generate a target concentration score, and generating a reminder message when the target concentration score is lower than the preset score. The technical effect of the present application is to improve the accuracy of the concentration assessment results.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and specifically to an artificial intelligence education and learning method, system, device, and storage medium. Background Art

[0002] With the rapid development of artificial intelligence (AI) technology and its increasing application in education, intelligent teaching management systems have become a crucial tool for improving teaching quality. In classroom teaching scenarios, student focus is a crucial factor influencing learning outcomes. Accurate focus assessment not only helps teachers identify learning issues but also provides a valuable basis for improving teaching methods. Therefore, accurately and in real time assessing student focus has become a critical technical challenge in education.

[0003] Currently, some approaches use computer vision to analyze students' facial expressions or gestures to assess their concentration. While these methods provide a basic assessment of student concentration, they fail to consider the impact of external factors on students' facial expressions and gestures within the classroom environment, resulting in inaccurate concentration assessment results. Summary of the Invention

[0004] This application provides an artificial intelligence education and learning method, system, device and storage medium for improving the accuracy of concentration assessment results.

[0005] In the first aspect, the present application provides an artificial intelligence education and learning method, which includes: when a teacher is in a teaching state in the classroom, obtaining the teacher's location information and the teacher's first behavior status information; combining the location information and the first behavior status information to generate standard behavior status information of the target student; obtaining the second behavior status information of the target student, comparing the standard behavior status information and the second behavior status information to generate a comparison result, and generating a concentration score of the target student based on the comparison result; obtaining the third behavior status information of other students within a preset range of the target student, and generating an adjustment index based on the third behavior status information; adjusting the concentration score by the adjustment index to generate a target concentration score, and generating a reminder message when the target concentration score is lower than the preset score.

[0006] By adopting this technical solution, standard behavior status information is dynamically generated by obtaining the teacher's location information and primary behavior status information, avoiding the evaluation bias caused by using fixed standards. Simultaneously, an adjustment index is generated by obtaining the third behavior status information of other students to adjust the concentration score, effectively eliminating the interference of environmental factors. This assessment method based on teacher behavior and group behavior can accurately identify normal behavioral changes caused by students following teaching activities, improving the accuracy of concentration assessment results.

[0007] Optionally, the standard behavior state information includes a standard head orientation, a standard gaze direction and a standard sitting posture, and the combination of the position information and the first behavior state information to generate the standard behavior state information of the target student includes: determining the area where the teacher is located according to the position information, and the area includes the podium area and the aisle area; if the teacher is in the podium area, identifying the first behavior state information through a first behavior model to generate the standard head orientation, the standard gaze direction and the standard sitting posture of the target student, and the first behavior model is obtained by training the first behavior data of the teacher when teaching in the podium area and the first standard behavior data of the student; if the teacher is in the aisle area, identifying the first behavior state information through a second behavior model to generate the standard head orientation, the standard gaze direction and the standard sitting posture of the target student, and the second behavior model is obtained by training the second behavior data of the teacher when teaching in the aisle area and the second standard behavior data of the student.

[0008] By adopting the above technical solution, by distinguishing the different teaching scenarios of teachers in the podium area and the aisle area, the first behavior model and the second behavior model with targeted training are respectively used to generate standard behavior state information. This scenario-based model selection mechanism enables the system to accurately identify the standard head orientation, standard gaze direction and standard sitting posture that students should have in different teaching scenarios. For example, when the teacher is teaching at the podium, the standard behavior of the students is mainly positive gaze; when the teacher is walking in the aisle to explain, the students need to make appropriate head turns and follow their eyes. Through this scenario-based behavior modeling method, the system can more accurately evaluate the students' concentration state in different teaching scenarios.

[0009] Optionally, the second behavior state information includes actual head orientation, actual gaze direction and actual sitting posture. The comparison of the standard behavior state information and the second behavior state information generates a comparison result, including: comparing the standard head orientation and the actual head orientation to obtain a head orientation deviation value; comparing the standard gaze direction and the actual gaze direction to obtain a gaze direction deviation value; comparing the standard sitting posture state and the actual sitting posture state to obtain a sitting posture deviation value; and using the head orientation deviation value, the gaze direction deviation value and the sitting posture deviation value as the comparison result.

[0010] By employing this technical solution, students' behavior is broken down into three dimensions: head orientation, gaze direction, and sitting posture. Deviations from the standard behavior are calculated for each, generating a multi-dimensional comparison. This refined deviation calculation method accurately identifies the specific differences between student behavior and the standard. This not only helps the system more accurately assess a student's overall focus, but also identifies which behavioral dimensions show significant deviations, providing a basis for subsequent targeted guidance.

[0011] Optionally, generating the concentration score of the target student based on the comparison result includes: generating a first score based on the head orientation deviation value, and the head orientation deviation value is negatively correlated with the first score; generating a second score based on the gaze direction deviation value, and the gaze direction deviation value is negatively correlated with the second score; generating a third score based on the sitting posture deviation value, and the sitting posture deviation value is negatively correlated with the third score; and performing weighted summation of the first score, the second score, and the third score to generate the concentration score of the target student.

[0012] By adopting the above technical solution and establishing a negative correlation between deviation values ​​and scores, scores for the three dimensions of head orientation, gaze direction, and sitting posture are calculated separately, and a weighted summation is used to generate the final concentration score. This dimensional scoring and weighted fusion method enables the system to assign different weights based on the importance of different behavioral characteristics. This not only preserves the independent contribution of each dimension's behavioral characteristics, but also enables a comprehensive assessment of the student's concentration state, improving the accuracy and rationality of the concentration score. Furthermore, the negative correlation scoring mechanism ensures that the greater the behavioral deviation, the lower the score of the corresponding dimension, meeting the actual needs of concentration assessment.

[0013] Optionally, generating an adjustment index based on the third behavioral state information includes: calculating the average concentration score of the other students based on the third behavioral state information; calculating the score difference between the concentration score and the average concentration score, and generating an adjustment index based on the score difference, wherein the score difference is inversely proportional to the adjustment index.

[0014] By employing this technical solution, the average concentration score of other students is calculated and compared with the target student's concentration score. Based on the difference in scores, an inversely proportional adjustment index is generated. This dynamic adjustment mechanism, based on group behavior, can effectively identify and compensate for changes in group behavior caused by environmental factors. When the target student's concentration score is close to the average, a larger adjustment index is assigned to compensate for environmental influences; when the difference is significant, a smaller adjustment index is assigned to maintain individual behavioral characteristics. This improves the concentration assessment's adaptability to environmental interference while maintaining sensitivity to individual distractions.

[0015] Optionally, adjusting the concentration score by using the adjustment index to generate a target concentration score includes: arithmetically adding the adjustment index to the concentration score to generate a target concentration score.

[0016] By employing the above technical solution, the adjustment index and the focus score are combined through arithmetic addition to generate the final target focus score. This simple and direct adjustment method not only maintains the basic characteristics of the focus score, but also effectively compensates for environmental factors through the adjustment index. When group behavior changes, a larger adjustment index can appropriately improve the score level; when individual students show distraction, a smaller adjustment index can maintain score differentiation, thus ensuring accuracy while providing clear and easy-to-interpret assessment results.

[0017] Optionally, after generating the target concentration score, it also includes: obtaining the target concentration score of the target student within a preset time period, generating a concentration trend curve based on the target concentration score of the target student within the preset time period; and generating learning behavior recommendations for the target student based on the concentration trend curve.

[0018] By adopting the above technical solution, by obtaining the target concentration score within a preset time period, a concentration trend curve is generated, and learning behavior recommendations are generated based on the curve characteristics. This dynamic tracking and analysis mechanism can reflect the changing patterns of students' concentration status from a macro perspective and help discover problems that may be overlooked by a single score. By analyzing the periodic fluctuations, sudden drops, or sustained lows of the trend curve, the system can generate targeted learning behavior recommendations, including both immediate behavioral adjustment recommendations and long-term learning habit improvement recommendations, thereby providing practical guidance for improving students' learning outcomes.

[0019] In a second aspect, the present application provides an artificial intelligence education and learning system, the system comprising: a first acquisition module, a combination module, a comparison module, a second acquisition module and an adjustment module; wherein,

[0020] The first acquisition module is used to obtain the teacher's location information and the teacher's first behavior status information when the teacher is in the teaching state in the classroom; the combination module is used to combine the location information and the first behavior status information to generate the standard behavior status information of the target student; the comparison module is used to obtain the second behavior status information of the target student, compare the standard behavior status information with the second behavior status information, generate a comparison result, and generate the target student's concentration score based on the comparison result; the second acquisition module is used to obtain the third behavior status information of other students within the preset range of the target student, and generate an adjustment index based on the third behavior status information; the adjustment module is used to adjust the concentration score by the adjustment index to generate a target concentration score, and generate a reminder message when the target concentration score is lower than the preset score.

[0021] In the third aspect, the present application provides an electronic device that adopts the following technical solution: it includes a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes a computer program such as any of the above-mentioned artificial intelligence education and learning methods.

[0022] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution: storing a computer program that can be loaded by a processor and execute any of the above-mentioned artificial intelligence education and learning methods.

[0023] In summary, this application includes at least one of the following beneficial technical effects:

[0024] By dynamically generating standard behavior status information based on the teacher's location and primary behavior status information, this approach avoids the evaluation bias caused by using fixed standards. Furthermore, by generating an adjustment index based on the third behavior status information of other students, the focus score is adjusted, effectively eliminating the interference of environmental factors. This assessment method based on both teacher and group behavior accurately identifies normal behavioral changes in students as they follow teaching activities, improving the accuracy of focus assessment results. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a flowchart of an artificial intelligence education and learning method provided by an embodiment of the present application;

[0026] Figure 2 This is a structural diagram of an artificial intelligence education and learning system provided by an embodiment of the present application;

[0027] Figure 3This is a structural diagram of an electronic device provided in an embodiment of the present application.

[0028] Description of reference numerals: 1000, electronic device; 1001, processor; 1002, communication bus; 1003, user interface; 1004, network interface; 1005, memory. DETAILED DESCRIPTION

[0029] In order to enable people skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0030] In the description of the embodiments of this application, words such as "exemplary," "for example," or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary," "for example," or "for example" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary," "for example," or "for example" is intended to present the relevant concepts in a concrete manner.

[0031] Figure 1 This is a flow chart of an artificial intelligence education and learning method provided by the embodiment of this application. Figure 1 As shown, the method includes S101-S105:

[0032] S101, when a teacher is in a teaching state in a classroom, obtain the teacher's location information and the teacher's first behavior state information.

[0033] In the classroom teaching process, the teacher, as the leader of knowledge transfer, has a direct impact on the students' attention allocation due to their position changes and behavioral status. In order to accurately assess the students' concentration, it is first necessary to obtain the teacher's spatial position information and behavioral status information in the classroom. Specifically, a plurality of cameras are set up in the classroom, and the cameras include but are not limited to wide-angle cameras installed on the front and back walls and the top of the classroom. When it is detected that the teacher is in a teaching state, the classroom screen is collected by the above-mentioned camera equipment and transmitted to the processing server in real time. Among them, the teaching state refers to the state in which the teacher is conducting teaching activities, which can be determined by a pre-trained teacher behavior recognition model. The model determines whether the teacher is currently in a teaching state by analyzing whether the teacher is performing teaching behaviors such as explaining, writing on the blackboard, and asking questions.

[0034] The processing server uses computer vision technology to analyze and process the image and extract the teacher's location information. The location information includes the teacher's two-dimensional coordinate values ​​in the classroom plane coordinate system and the height value relative to the ground, thereby determining the teacher's specific position in three-dimensional space. At the same time, the human body posture estimation algorithm is used to analyze the teacher's first behavioral state information. The first behavioral state information includes but is not limited to characteristic parameters such as the teacher's standing posture, arm movements, head orientation, and gaze direction. These characteristic parameters are obtained by identifying and tracking key points of the human body and recorded and stored in a standardized numerical form.

[0035] By acquiring the teacher's location information and first behavior status information in real time, the system can establish a mapping relationship between the teacher's teaching behavior and the student's standard attention state. For example, when the teacher is in the podium area explaining key knowledge, the student's attention should usually be focused on the teacher; when the teacher walks in the aisle area and instructs a student, the attention of nearby students may be reasonably shifted. This dynamic evaluation standard based on teacher behavior is more in line with the actual teaching scenario than a fixed evaluation standard, and can provide a more accurate reference for subsequent student concentration evaluation. At the same time, the collection process of these data does not require any additional operations by the teacher, will not interfere with normal teaching activities, and ensures the naturalness and continuity of the teaching process.

[0036] S102: Generate standard behavior status information of the target student by combining the position information and the first behavior status information.

[0037] In classroom teaching scenarios, students' standard behavior should be dynamically adjusted according to the teacher's position and teaching behavior. This step aims to establish a dynamic student behavior assessment standard to make the concentration assessment more consistent with the actual teaching situation.

[0038] Specifically, the system first determines the teacher's location based on the teacher's location information and divides the classroom space into a podium area and an aisle area. The podium area refers to the specific area at the front of the classroom where the teacher conducts concentrated lectures; the aisle area refers to the area outside the podium where the teacher moves around the classroom, patrols, or provides individual guidance.

[0039] When the system determines that the teacher is in the podium area, it calls the first behavioral model to process the teacher's first behavioral status information. This first behavioral model is trained using machine learning methods. Its training data includes a large amount of behavioral data from teachers teaching in the podium area, as well as standard student behavior data identified by a team of professional education experts. This standard behavior data is based on multidisciplinary theories such as educational psychology and ergonomics, combined with extensive classroom teaching experience. It was jointly researched and developed by education experts, psychologists, and experienced teachers. For example, from an ergonomic perspective, the optimal head tilt angle range for protecting eyesight and spinal health was determined; from an educational psychology perspective, the optimal attention allocation method for knowledge absorption was standardized. This scientific behavioral standard data is correlated with teachers' teaching behavior, forming a more convincing and scientific evaluation benchmark. Through correlation analysis of this data, a mapping relationship is established between teachers' teaching behavior at the podium and standard student behavior.

[0040] For example, when a teacher is explaining key points at the podium, the system generates standard student behavior status information, including: the standard head orientation should be towards the podium, the standard gaze direction should be focused on the teacher or blackboard area, and the standard sitting posture should be kept upright and forward.

[0041] When the teacher is in the aisle area, the system switches to the second behavior model for processing. The training data of the second behavior model comes from the scene when the teacher is conducting teaching activities in the aisle area, including the behavior data of the teacher walking around, individual guidance, etc., as well as the standard performance of the students at this time. The standard behavior state information generated in this case is more flexible, allowing the student's head direction and gaze direction to deviate within a reasonable range. In particular, when the teacher is providing individual guidance in the nearby area, the attention of surrounding students may be temporarily shifted to the interactive area. This change in behavior state is also included in the standard range.

[0042] Based on the above embodiment, as an optional implementation, in S102, the standard behavior status information includes the standard head orientation, the standard gaze direction, and the standard sitting posture. Combining the position information and the first behavior status information, generating the student's standard behavior status information specifically includes S21-S23:

[0043] S21, determining the area where the teacher is located based on the location information, which includes the podium area and the aisle area.

[0044] During actual classroom instruction, changes in the teacher's position often correspond to different teaching behaviors and objectives, which directly impacts the standard behavior expected of students. To this end, the system first needs to accurately identify and zonal the teacher's location information. A positioning system deployed within the classroom can obtain the teacher's spatial coordinates in real time. The system divides the classroom space into two main areas: the podium area and the aisle area. The podium area refers to the specific space at the front of the classroom used for concentrated teaching, while the aisle area includes all areas within the classroom other than the podium, primarily used for teacher movement and individual guidance.

[0045] S22, if the teacher is in the podium area, the first behavior state information is identified through the first behavior model to generate the target student's standard head orientation, standard gaze direction and standard sitting posture. The first behavior model is obtained by training the first behavior data of the teacher when teaching in the podium area and the first standard behavior data of the students.

[0046] When the system determines that the teacher is in the podium area, the first behavior model will be enabled to process the teacher's first behavior state information. The first behavior state information contains the teacher's specific behavioral characteristics in the podium area, such as pointing to the blackboard, explaining with gestures, emphasizing key points, and other teaching actions. The first behavior model generates corresponding student standard behavior state information by analyzing these behavioral characteristics and combining them with the pre-imported first standard behavior data. Specifically, when the teacher is explaining knowledge at the podium, the standard head orientation generated by the system should be consistent with the direction of the podium, the standard gaze direction should be concentrated on the teacher or teaching media presentation area, and the standard sitting posture should be kept upright and facing the podium.

[0047] The training process for the first behavioral model consists of five stages: data collection, data preprocessing, feature extraction, model construction, and model optimization. During the data collection stage, the system uses multiple high-definition cameras installed in the classroom to capture video data of the teacher teaching at the podium. This video data contains detailed information about the teacher's behavior, such as posture, gestures, and body orientation. Simultaneously, the system collects standard student behavior data corresponding to the teacher's actions, including head orientation angle values, gaze direction vectors, and sitting posture parameters.

[0048] During the data preprocessing phase, the system first performs noise reduction and illumination normalization on the collected video data to improve image quality. It then uses an object detection algorithm to locate the instructor in the video and human key point detection technology to extract the instructor's skeletal node information. For standard student behavior data, the system performs outlier detection and data cleaning to eliminate unreasonable data samples.

[0049] During the feature extraction phase, the system constructs a multidimensional feature vector based on preprocessed teacher behavior data. These include: spatial position features, representing the teacher's specific coordinate position at the podium; action features, including spatial parameters of key nodes such as arm lift angle and body turn angle; and temporal features, describing the continuous evolution of the teacher's behavior. The corresponding standard student behavior data is also converted into feature vectors, including 3D Euler angles of the head, gaze direction unit vector, and torso posture parameters.

[0050] During the model construction phase, the system employs a deep learning network architecture, specifically a long short-term memory (LSTM) network as its foundational model structure. This network consists of an input layer, multiple LSTM layers, a fully connected layer, and an output layer. The input layer receives the teacher's behavior feature vector, the LSTM layer captures the temporal dependencies of the behavior sequence, the fully connected layer performs feature mapping, and the output layer generates predicted standard student behavior parameters.

[0051] During the model optimization phase, the system uses mean squared error as the loss function and optimizes model parameters via a backpropagation algorithm. Specifically, the training dataset is first split into training and validation sets in an 8:2 ratio. The model is then trained using batch gradient descent, calculating the training and validation losses for each training cycle. When the validation loss stops decreasing, training is terminated using an early stopping strategy. Furthermore, dropout is used to prevent overfitting, with a dropout rate of 0.5.

[0052] Through this training process, the first behavioral model establishes a mapping between teacher behavior and student behavior. When the model receives new teacher behavior data, it can predict the corresponding student behavior in real time. To ensure the model's practicality, the system also incorporates a regular update mechanism, continuously optimizing model performance through incremental learning to adapt to changing teaching scenarios.

[0053] S23. If the teacher is in the aisle area, the first behavior state information is identified through the second behavior model to generate the target student's standard head orientation, standard gaze direction and standard sitting posture. The second behavior model is obtained by training the second behavior data of the teacher when teaching in the aisle area and the second standard behavior data of the students.

[0054] When the teacher is in the aisle area, the system switches to the second behavior model for processing. Similar to the first behavior model, the second behavior model works based on the second behavior state information of the teacher in the aisle area, and the construction method is the same as the first behavior model. For details, please refer to the construction method of the first behavior model, and we will not go into details here. This information includes features such as the teacher's walking route, stop location, and interactive objects. By analyzing these features and combining them with the second standard behavior data, the system generates more flexible standard behavior state information. For example, when a teacher is giving individual guidance in a certain area, the standard head orientation and gaze direction of students in the surrounding area can deviate from the direction of the podium and turn towards the teacher's position, while the standard sitting posture needs to maintain basic correctness requirements.

[0055] S103, obtaining the second behavior state information of the target student, comparing the standard behavior state information with the second behavior state information, generating a comparison result, and generating a concentration score of the target student based on the comparison result.

[0056] In order to accurately assess students' concentration in class, the system needs to compare and analyze the students' actual behavior with the standard behavior generated above. First, the system collects real-time behavioral data of target students through high-definition cameras installed in the classroom. Target students refer to specific students whose concentration needs to be assessed. The system uses facial recognition technology to accurately locate and track them. By analyzing the collected image data through computer vision algorithms, the system extracts the target student's second behavioral state information, specifically including behavioral characteristics in three key dimensions: actual head orientation, actual gaze direction, and actual sitting posture.

[0057] Among them, the actual head orientation is obtained through facial feature point detection technology, and the Euler angle of the head in three-dimensional space is calculated to obtain the specific values ​​of the head pitch angle, yaw angle and roll angle; the actual gaze direction is determined by the eye tracking algorithm by analyzing the pupil position and eye rotation angle to determine the student's line of sight direction vector; the actual sitting posture is determined by using human skeleton key point detection technology to extract the spatial coordinates of each joint point of the student's upper body, and calculate the sitting posture feature parameters such as trunk inclination and shoulder horizontality.

[0058] After acquiring this actual behavioral data, the system accurately compares it with the previously generated standard behavioral state information. In the head orientation dimension, the Euclidean distance between the actual three-dimensional head angle and the standard head orientation angle is calculated to obtain the head orientation deviation value. In the gaze direction dimension, the angle between the actual gaze direction and the standard gaze direction is calculated to obtain the gaze direction deviation value. In the sitting posture dimension, the sitting posture deviation value is obtained by calculating the degree of difference between the actual sitting posture characteristic parameters and the standard sitting posture parameters.

[0059] Based on the deviations in these three dimensions, the system uses a weighted scoring mechanism to generate a concentration score. Specifically, the system calculates the first score based on the head orientation deviation, with larger deviations indicating lower scores. The second score is calculated based on the gaze direction deviation, also following the negative correlation between deviation and score. The third score is calculated based on the sitting posture deviation. Finally, the system takes a weighted sum of these three scores to obtain the target student's concentration score.

[0060] Based on the above embodiment, as an optional implementation, in S103, the second behavior state information includes the actual head orientation, the actual gaze direction, and the actual sitting posture. Comparing the standard behavior state information with the second behavior state information to generate a comparison result specifically includes S31-S34:

[0061] S31, comparing the standard head orientation and the actual head orientation to obtain a head orientation deviation value.

[0062] In order to accurately assess the students' actual state of concentration, the system needs to quantitatively compare the students' actual behavior with the standard behavior. This comparison process is based on three key dimensions: head orientation, gaze direction, and sitting posture. The system first processes the comparison of the head orientation dimension and obtains the head orientation deviation value by calculating the Euclidean distance between the standard head orientation and the actual head orientation. Specifically, since the head orientation data is represented by three-dimensional Euler angles, including pitch angle, yaw angle, and roll angle, the system subtracts these three angle values ​​and calculates the square root of the sum of the squares to obtain a scalar value representing the degree of head deviation.

[0063] S32, comparing the standard gaze direction and the actual gaze direction to obtain a gaze direction deviation value.

[0064] When comparing the gaze direction dimension, the system calculates the angle between the two three-dimensional unit vectors—the standard gaze direction and the actual gaze direction—to obtain the gaze direction deviation value. This calculation method uses the dot product of the two direction vectors to obtain the cosine value, and then uses the inverse cosine function to obtain the angle value in radians. This calculation method based on the vector angle can intuitively reflect the degree of deviation between the student's gaze direction and the expected gaze direction.

[0065] S33, comparing the standard sitting posture state and the actual sitting posture state to obtain a sitting posture state deviation value.

[0066] For posture comparison, the system calculates the Mahalanobis distance between the standard and actual posture parameters to determine the deviation. These parameters contain multiple characteristic components, such as torso tilt angle and shoulder levelness. The advantage of using the Mahalanobis distance is that it considers the correlation between different characteristic components, providing a more accurate measure of the overall deviation in posture.

[0067] S34, taking the head orientation deviation value, the gaze direction deviation value and the sitting posture deviation value as comparison results.

[0068] The vector formed by these three deviation values ​​is the final comparison result. Each component is normalized to a uniform range of [0, 1] to facilitate subsequent scoring calculations. Breaking down the behavioral comparison into three dimensions allows for more precise identification of specific issues with students' focus. Secondly, different distance metrics are used to calculate the characteristics of the data in each dimension, improving the accuracy of the comparison results. Finally, normalization makes the deviation values ​​across different dimensions comparable, providing standardized baseline data for subsequent scoring.

[0069] Based on the above embodiment, as an optional implementation, in S103, generating a concentration score of the target student according to the comparison result specifically includes S331-S334:

[0070] S331: Generate a first score based on the head orientation deviation value, where the head orientation deviation value is negatively correlated with the first score.

[0071] In order to convert quantitative behavioral deviations into intuitive concentration scores, the system adopts dimensional scoring and weighted fusion methods. In the head orientation scoring link, the system uses an exponential decay function to generate the first score. The function expression is: score1=100*exp(-k1*d1); where score1 represents the score of the head orientation dimension, and the value range is [0,100]; d1 represents the head orientation deviation value; k1 is the decay coefficient, which is used to control the rate of score decrease. In this embodiment, the value is 0.5; exp() is a natural exponential function. The reason for choosing the exponential decay function is that when the deviation value is close to 0, the score is close to the full score of 100, which reflects the importance of standard head orientation; as the deviation value increases, the score shows a gentle downward trend. This exponential decay characteristic can better simulate the change of concentration with the degree of head deviation.

[0072] S332: Generate a second score according to the gaze direction deviation value, where the gaze direction deviation value is negatively correlated with the second score.

[0073] For the gaze direction score, the system uses a similar exponential decay function to generate a second score: score2 = 100 * exp(-k2 * d2). Here, score2 represents the gaze direction score, ranging from 0 to 100; d2 represents the gaze direction deviation; and k2 is the decay coefficient, set to 0.8 (larger than k1 to reflect the stronger impact of gaze direction deviation on focus). This function is sensitive to subtle deviations in gaze direction, consistent with its central role as a core indicator of focus.

[0074] S333: Generate a third score based on the sitting posture deviation value, where the sitting posture deviation value is negatively correlated with the third score.

[0075] For the sitting posture score, the system uses a piecewise linear function to generate a third score: when d3 ≤ th, score3 = 100 * (1 - d3 / th); when d3 > th, score3 = 0. Here, score3 represents the score of the sitting posture dimension, with a value range of [0, 100]; d3 represents the sitting posture deviation; and th is a threshold parameter, which in this embodiment is 0.5. This segmented processing method is based on the following considerations: within the threshold range, the score decreases linearly with the deviation value, reflecting the gradual nature of normal sitting posture adjustment; beyond the threshold, the score is directly reset to zero, reflecting a negative evaluation of obviously inappropriate sitting posture.

[0076] S334: Perform a weighted summation of the first score, the second score, and the third score to generate a concentration score for the target student.

[0077] The final focus score is obtained by weighted summation:

[0078] Concentration score = w1score1 + w2score2 + w3score3; among them, w1, w2, and w3 are the weight coefficients of the head orientation score, gaze direction score, and sitting posture score, respectively, and satisfy w1+w2+w3=1. In this embodiment, considering that the gaze direction is a direct indicator reflecting the student's immediate attention level, its weight coefficient is set to the highest, w2=0.5; head orientation, as an overall representation of facial orientation, plays an important auxiliary role in concentration assessment, and its weight coefficient is w1=0.3; sitting posture, as a behavioral manifestation of long-term concentration, plays a supplementary verification role, and its weight coefficient is w3=0.2. This weight configuration forms a multi-dimensional evaluation system with gaze behavior as the main focus and head posture and sitting posture as the auxiliary. In actual situations, it can be set according to actual conditions.

[0079] S104: Acquire third behavior status information of other students within a preset range of the target student, and generate an adjustment index based on the third behavior status information.

[0080] In real-world classroom settings, students' concentration is often affected by surrounding factors. To make concentration assessment more objective and reasonable, this step introduces group environmental factors as adjustment parameters. The system first determines a preset range for the target student, which is centered around the target student and covers the eight seats closest to them.

[0081] The system uses deployed cameras to collect behavioral data from other students within a preset range. This third behavioral status information shares the same data structure as the second behavioral status information, including characteristic parameters such as the students' head orientation, gaze direction, and sitting posture. The system processes this data using the same computer vision algorithms and behavioral analysis models to generate a unique concentration score for each student. This consistent data collection and processing method ensures comparability of scoring results.

[0082] After obtaining the concentration scores of all students within a preset range, the system calculates the arithmetic mean of these scores to obtain the average concentration score. This average reflects the overall concentration level of the current local teaching environment. The system then calculates the difference between the target student's concentration score and the average concentration score, known as the score difference. The positive or negative sign and magnitude of the score difference reflect the target student's relative concentration level in the current group environment.

[0083] The reason for obtaining the average concentration score of other students is to avoid system misjudgments. For example, if there is a change in the external environment, such as excessive sunlight, the target student may have a large head tilt. At this time, the system may think that the target student is less focused and will remind him, but in fact the target student is in a focused state, which is a false alarm. Considering the average concentration score of other students within the preset range of the target student can solve this problem well. Still taking the above problem as an example, if other students in the target student's area also have similar head deviations due to sunlight problems, resulting in their concentration scores being low, the average concentration score calculated by the system will also be reduced accordingly. The adjustment index calculated by the score difference will be larger. In the subsequent score adjustment, this larger adjustment index will appropriately increase the target student's concentration score, thereby avoiding misjudgments due to the environment or other external factors. This dynamic adjustment mechanism based on group behavior can effectively filter out the influence of external interference factors and improve the accuracy and reliability of system evaluation.

[0084] Based on the score difference, the system generates an adjustment index. Specifically, when the score difference is larger, the generated adjustment index is relatively smaller, which is used to reduce the extreme difference between individual scores and group scores; when the score difference is smaller, the influence of the adjustment index is correspondingly reduced to maintain the independence of individual scores.

[0085] Based on the above embodiment, as an optional implementation, in S104, generating the adjustment index according to the third behavior state information specifically includes S41-S42:

[0086] S41: Calculate the average concentration scores of other students based on the third behavior status information.

[0087] To eliminate the influence of external environmental factors on concentration assessment, the system introduces a dynamic adjustment mechanism based on group behavior. First, the system processes the third-behavior status information of other students within the preset range of the target student, using the same scoring method as the target student to obtain a concentration score for each student. The arithmetic mean of these scores is then calculated to obtain the average concentration score.

[0088] S42, calculating the score difference between the concentration score and the average concentration score, and generating an adjustment index based on the score difference, wherein the score difference is inversely proportional to the adjustment index.

[0089] Next, the system calculates the difference between the target student's concentration score and the average concentration score, i.e., the score difference. Based on the score difference, the system generates an adjustment index through an inverse relationship, i.e., the larger the score difference, the smaller the generated adjustment index; the smaller the score difference, the larger the generated adjustment index. The reason for choosing this inverse relationship is that when the score difference is small, it means that the target student's behavior is more consistent with the group behavior. In this case, a larger adjustment index is given to compensate for possible environmental influences; when the score difference is large, it means that the target student's behavior is significantly different from the group behavior. In this case, a smaller adjustment index is given to maintain the discrimination of the assessment.

[0090] Adjusting the concentration score by adjusting the index to generate the target concentration score specifically includes: arithmetically adding the adjustment index to the concentration score to generate the target concentration score.

[0091] After obtaining the adjustment index, the system needs to rationally integrate it with the focus score to generate the final target focus score. The system uses arithmetic addition to perform the adjustment, directly adding the adjustment index to the focus score to obtain the target focus score. The choice of arithmetic addition, a simple and direct adjustment method, is based on the following considerations: First, the addition operation maintains the basic characteristics of the focus score, and the adjustment index serves as a correction to the original score; second, the result of the addition operation has a clear physical meaning, which is easy to understand and interpret; finally, this linear adjustment method can maintain the monotonicity of the score, that is, a higher focus score remains at a relatively high level after adjustment.

[0092] Through this adjustment mechanism, the system can effectively deal with environmental interference in various teaching scenarios. For example, when the lighting conditions in the classroom change, it may cause students in a certain area to have a large head deviation due to adjusting their sitting posture. At this time, the concentration scores of these students will generally decrease. The system calculates a larger adjustment index, and after adding and adjusting it, it can raise the target concentration scores of these students to a reasonable level. At the same time, for individual students who do have distracted attention, since the generated adjustment index is small, the target concentration score after adding and adjusting can still reflect their actual state of concentration.

[0093] S105 , adjusting the concentration score by adjusting the index to generate a target concentration score. When the target concentration score is lower than a preset score, generating a reminder message.

[0094] To obtain a more accurate assessment of student focus, the system adjusts the previously obtained focus score for environmental factors. Specifically, the system arithmetically adds the adjustment index to the focus score to obtain the target focus score. This simple arithmetic addition method, validated by extensive experimental data, more intuitively reflects the impact of environmental factors than other complex adjustment methods while maintaining interpretability of the calculated results.

[0095] The system pre-sets a preset scoring threshold for concentration. Taking into account the differences in different course types, teaching links and learning objectives, this threshold can be adjusted dynamically. For example, in the knowledge explanation link, the system sets a higher preset scoring threshold to ensure that students can fully receive and understand new knowledge; in the practical operation link, because students' attention needs to switch between the operation object and the teacher's guidance, the system lowers the preset scoring threshold accordingly; in the classroom discussion link, the system will adjust the threshold standard according to the discussion format and interaction needs. This dynamic threshold setting based on the teaching scenario makes the system's reminder mechanism more in line with actual teaching needs and avoids the evaluation bias that may be caused by fixed standards.

[0096] When the calculated target concentration score is lower than the preset score, it means that the student may be distracted, and the system will automatically generate a reminder message. This reminder message includes two levels: the first is a reminder for teachers, in which the system will remind teachers to pay attention to the student's learning status through signal light changes or brief text prompts in a specific area on the teacher's monitor in a way that does not affect teaching; the second is a reminder for students, in which the system will use a micro-vibration device at the student's seat or a visual prompt on the personal display terminal to subtly remind students to adjust their attention in a way that does not affect other students.

[0097] After generating the target focus score, it also includes:

[0098] Obtain the target concentration score of the target student within a preset time period, generate a concentration trend curve based on the target concentration score of the target student within the preset time period; and generate learning behavior suggestions for the target student based on the concentration trend curve.

[0099] To comprehensively assess changes in students' concentration and provide targeted guidance, the system continuously tracks and analyzes the target student's concentration. Specifically, the system obtains all target concentration scores for the target student within a preset time period, which can be the length of a class or a specific teaching session. The system plots this time-series score data into a concentration trend curve, which intuitively illustrates the dynamic changes in a student's concentration level.

[0100] Based on the morphological characteristics of the concentration trend curve, the system generates corresponding learning behavior suggestions. When the trend curve shows periodic fluctuations, it indicates that the student's attention is regularly dispersed. The suggestions generated by the system include arranging the learning rhythm reasonably and taking appropriate short breaks during the period of low attention. When the trend curve shows a sudden drop, the system will record the corresponding time points and teaching content, and recommend that students focus on reviewing the key points of knowledge during these periods. When the trend curve continues to be low at a certain level, the system will recommend adjusting the learning posture or changing the seat position to improve the learning environment. When the trend curve shows a gradual downward trend, the system will recommend appropriate adjustments to the work and rest time to ensure adequate rest.

[0101] Based on the above method, this application also discloses an artificial intelligence education and learning system, such as Figure 2 As shown, Figure 2 This is a structural diagram of an artificial intelligence education and learning system provided by an embodiment of the present application. The system includes: a first acquisition module, a combination module, a comparison module, a second acquisition module and an adjustment module; wherein,

[0102] The first acquisition module is used to obtain the teacher's location information and the teacher's first behavior status information when the teacher is in the teaching state in the classroom; the combination module is used to combine the location information and the first behavior status information to generate the standard behavior status information of the target student; the comparison module is used to obtain the second behavior status information of the target student, compare the standard behavior status information and the second behavior status information, generate a comparison result, and generate a concentration score of the target student based on the comparison result; the second acquisition module is used to obtain the third behavior status information of other students within the preset range of the target student, and generate an adjustment index based on the third behavior status information; the adjustment module is used to adjust the concentration score through the adjustment index to generate a target concentration score, and generate a reminder message when the target concentration score is lower than the preset score.

[0103] It should be noted that the above embodiments provide systems that implement their functions using only the division of the above functional modules as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0104] See Figure 3 , is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 3 As shown, the electronic device 1000 may include: at least one processor 1001 , at least one network interface 1004 , a user interface 1003 , a memory 1005 , and at least one communication bus 1002 .

[0105] The communication bus 1002 is used to implement the connection and communication between these components.

[0106] The user interface 1003 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 1003 may also include a standard wired interface and a wireless interface.

[0107] The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0108] Processor 1001 may include one or more processing cores. Using various interfaces and circuits, processor 1001 connects to various components within the server. It executes instructions, programs, code sets, or instruction sets stored in memory 1005, as well as accesses data stored in memory 1005, to perform various server functions and process data. Optionally, processor 1001 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). Processor 1001 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing content displayed on the display; and the modem handles wireless communications. It is understood that the modem may not be integrated into processor 1001 but implemented as a separate chip.

[0109] Among them, the memory 1005 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 1005 includes a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 1005 may optionally be at least one storage device located away from the aforementioned processor 1001. As Figure 3 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of an artificial intelligence education and learning method.

[0110] exist Figure 3In the electronic device 1000 shown, the user interface 1003 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 1001 can be used to call an application program storing an artificial intelligence education and learning method in the memory 1005. When executed by one or more processors, the electronic device executes one or more methods described in the above embodiments.

[0111] An electronic device readable storage medium stores instructions, which, when executed by one or more processors, enable the electronic device to execute one or more of the methods described in the above embodiments.

[0112] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for this application.

[0113] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0114] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of the devices or units can be electrical or other forms.

[0115] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0116] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0117] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of this application, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of this application. The aforementioned memory includes various media that can store program code, such as USB flash drives, mobile hard drives, magnetic disks, or optical disks.

[0118] The above is only an exemplary embodiment of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the specification and practicing the disclosure herein, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not recorded in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. An artificial intelligence education and learning method, characterized in that: The method comprises: When a teacher is in a teaching state in the classroom, obtaining the teacher's location information and the teacher's first behavior state information; Combining the position information and the first behavior state information, generating standard behavior state information of the target student; the standard behavior state information includes a standard head orientation, a standard gaze direction, and a standard sitting posture, and combining the position information and the first behavior state information to generate the standard behavior state information of the target student includes: determining the area where the teacher is located according to the position information, and the area includes a podium area and an aisle area; if the teacher is in the podium area, identifying the first behavior state information through a first behavior model, and generating the standard head orientation, the standard gaze direction, and the standard sitting posture of the target student, the first behavior model is obtained by training the first behavior data of the teacher when teaching in the podium area, and the first standard behavior data of the student; if the teacher is in the aisle area, identifying the first behavior state information through a second behavior model, and generating the standard head orientation, the standard gaze direction, and the standard sitting posture of the target student, the second behavior model is obtained by training the second behavior data of the teacher when teaching in the aisle area, and the second standard behavior data of the student; Acquiring second behavior state information of the target student, comparing the standard behavior state information with the second behavior state information to generate a comparison result, and generating a concentration score of the target student based on the comparison result; Obtaining third behavior status information of other students within a preset range of the target student, and generating an adjustment index based on the third behavior status information; processing the third behavior status information of other students within the preset range of the target student, using the same scoring method as the target student to obtain a concentration score for each student, and calculating the arithmetic mean of these scores to obtain an average concentration score; generating an adjustment index based on the score difference through an inverse relationship, where the greater the score difference, the smaller the generated adjustment index; and the smaller the score difference, the larger the generated adjustment index; The concentration score is adjusted using the adjustment index to generate a target concentration score, and when the target concentration score is lower than a preset score, a reminder message is generated.

2. The artificial intelligence education and learning method according to claim 1, characterized in that: The second behavior state information includes actual head orientation, actual gaze direction and actual sitting posture. The comparison of the standard behavior state information and the second behavior state information generates a comparison result, including: comparing the standard head orientation and the actual head orientation to obtain a head orientation deviation value; comparing the standard gaze direction and the actual gaze direction to obtain a gaze direction deviation value; comparing the standard sitting posture state and the actual sitting posture state to obtain a sitting posture deviation value; and taking the head orientation deviation value, the gaze direction deviation value and the sitting posture deviation value as the comparison result.

3. The artificial intelligence education and learning method according to claim 2, characterized in that: Generating the concentration score of the target student based on the comparison result includes: generating a first score based on the head orientation deviation value, and the head orientation deviation value is negatively correlated with the first score; generating a second score based on the gaze direction deviation value, and the gaze direction deviation value is negatively correlated with the second score; generating a third score based on the sitting posture deviation value, and the sitting posture deviation value is negatively correlated with the third score; and performing a weighted summation of the first score, the second score, and the third score to generate the concentration score of the target student.

4. The artificial intelligence education and learning method according to claim 1, characterized in that: Generating the adjustment index based on the third behavioral state information includes: calculating the average concentration score of the other students based on the third behavioral state information; calculating the score difference between the concentration score and the average concentration score, and generating the adjustment index based on the score difference, wherein the score difference is inversely proportional to the adjustment index.

5. The artificial intelligence education and learning method according to claim 1, characterized in that: The adjusting the concentration score by using the adjustment index to generate a target concentration score includes: arithmetically adding the adjustment index to the concentration score to generate a target concentration score.

6. The artificial intelligence education and learning method according to claim 1, characterized in that: After generating the target concentration score, it also includes: obtaining the target concentration score of the target student within a preset time period, and generating a concentration trend curve based on the target concentration score of the target student within the preset time period; generating learning behavior recommendations for the target student based on the concentration trend curve.

7. An artificial intelligence education and learning system, characterized in that: The system includes: a first acquisition module, a combination module, a comparison module, a second acquisition module and an adjustment module; wherein the first acquisition module is used to acquire the teacher's location information and the teacher's first behavior status information when the teacher is in a teaching state in the classroom; The combining module is used to combine the position information and the first behavior state information to generate standard behavior state information of the target student; the standard behavior state information includes a standard head orientation, a standard gaze direction and a standard sitting posture, and the combining module generates the standard behavior state information of the target student, including: determining the area where the teacher is located according to the position information, where the area includes a podium area and an aisle area; if the teacher is in the podium area, identifying the first behavior state information through a first behavior model to generate the standard head orientation, the standard gaze direction and the standard sitting posture of the target student, where the first behavior model is obtained by training the first behavior data of the teacher when teaching in the podium area and the first standard behavior data of the student; if the teacher is in the aisle area, identifying the first behavior state information through a second behavior model to generate the standard head orientation, the standard gaze direction and the standard sitting posture of the target student, where the second behavior model is obtained by training the second behavior data of the teacher when teaching in the aisle area and the second standard behavior data of the student; The comparison module is used to obtain the second behavior state information of the target student, compare the standard behavior state information with the second behavior state information, generate a comparison result, and generate a concentration score for the target student based on the comparison result; the second acquisition module is used to obtain the third behavior state information of other students within the preset range of the target student, and generate an adjustment index based on the third behavior state information; the third behavior state information of other students within the preset range of the target student is processed, and the same scoring method as that of the target student is used to obtain the concentration score of each student, and the arithmetic mean of these scores is calculated to obtain the average concentration score, and the adjustment index is generated through an inverse relationship according to the score difference. The larger the score difference, the smaller the generated adjustment index; the smaller the score difference, the larger the generated adjustment index; the adjustment module is used to adjust the concentration score through the adjustment index to generate a target concentration score, and generate a reminder message when the target concentration score is lower than the preset score.

8. An electronic device, characterized in that: It includes a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that A computer program is stored which can be loaded by a processor and execute the method according to any one of claims 1 to 6.

Images