Interactive teaching methods, devices, electronic equipment and storage media

Abstract

This invention relates to the field of educational information technology, and more particularly to a classroom teaching interaction method, device, electronic device, and storage medium. The method first obtains multiple first-state queues; then, for each first-state queue, it performs dimensionality reduction using a corresponding dimensionality reduction operator to obtain a second-state queue; next, it selects the closest queue from multiple historical second-state queues as a reference queue, and determines a focus index based on the reference queue; finally, it determines an overall classroom activity index based on multiple focus indices, and when the activity index is lower than a first index threshold, it selects a target proposal from multiple prepared proposals and sends it to a first terminal for question-and-answer sessions. This invention reflects the overall classroom activity through the focus of multiple students, allowing for more in-depth explanations of key content based on the overall classroom activity, thus improving the utilization of class time.

Description

Technical Field

[0001] This invention relates to the field of educational information technology, and in particular to a classroom teaching interaction method, device, electronic device, and storage medium. Background Technology

[0002] The essence of classroom interaction is "two-way information flow between teachers and students". When this flow is interrupted, teachers cannot obtain timely learning feedback from students and can only proceed with teaching based on pre-set lesson plans, thus falling into the teaching misconception of "feeling good about themselves", which ultimately leads to an imbalance in time allocation and waste of classroom time.

[0003] Some teachers prioritize "completing the lesson plan" over "ensuring students grasp the key points." For example, in math classes, in order to "smoothly" finish the chapter content, they spend a lot of time on the "knowledge point introduction" section (such as playing a 10-minute video about everyday life that is not strongly related to the knowledge point), but because they are worried that interaction will take up time, they compress the key content of "function graph transformation" into a 5-minute list of formulas, causing the key content to be ignored in the "quick process".

[0004] Through the above process, we have discovered that the crucial role of interaction is to "adjust the pace of teaching in a timely manner." Without interaction, teachers cannot assess students' level of understanding and can only proceed at a fixed pace. The imbalance in content caused by the lack of interaction is not simply a matter of "improper time allocation," but rather it will have a chain of negative impacts on students' learning outcomes and interest, as well as on teachers' ability to improve their teaching. In the long run, this will seriously restrict the quality of classroom teaching.

[0005] Currently, some visual image technologies can provide estimates of students' facial expressions or inner states based on their facial expressions. These are called estimates because the results obtained through these technologies have significant uncertainty (individual instantaneous expressions are easily affected by non-learning factors, such as external noise and personal emotional fluctuations, and cannot consistently reflect learning focus). At present, the effectiveness of using these technologies to assist classroom teaching is minimal.

[0006] How to improve the teaching interaction process and increase the utilization rate of teaching time allocation based on the student group's situation is an urgent problem to be solved.

[0007] Therefore, it is necessary to develop and design an interactive classroom teaching method. Summary of the Invention

[0008] The present invention provides a classroom teaching interaction method, device, electronic device and storage medium to solve the problem of unreasonable time allocation for teaching content caused by insufficient classroom interaction in the prior art.

[0009] In a first aspect, embodiments of the present invention provide a classroom teaching interaction method, including:

[0010] Multiple first state queues are obtained, wherein each first state queue is obtained based on a corresponding first terminal, and the first state queue includes multiple first states representing facial expressions arranged in order according to time nodes;

[0011] For each first state queue, the second state queue is obtained by performing dimensionality reduction using the dimensionality reduction operator corresponding to the first state queue.

[0012] The closest queue to each second-state queue from multiple historical second-state queues will be used as the reference queue, and the focus index will be determined based on the reference queue.

[0013] The overall classroom activity index is determined based on multiple attention indices. When the activity index is lower than the first index threshold, a target proposal is selected from multiple prepared proposals and sent to the first terminal for questioning and answering.

[0014] In one possible implementation, the dimensionality reduction operator is constructed using multiple historical first-state queues, including:

[0015] Obtain multiple historical first state queues, wherein the historical first state queues have the same dimension as the first state queues, and the multiple historical first state queues and the current first state queue correspond to the same first terminal;

[0016] Data is extracted from the multiple historical first state queues one dimension at a time, and the multiple extracted data are constructed into a first dimension array;

[0017] A first correlation matrix representing the relationship between dimensions is constructed using the first dimension array;

[0018] Calculate multiple eigenvalues ​​of the first correlation matrix, and determine a first eigenvector based on each eigenvalue;

[0019] Delete the minimum values ​​among the plurality of first feature values;

[0020] The remaining first eigenvectors corresponding to the multiple first eigenvalues ​​are used to construct a dimensionality reduction operator.

[0021] In one possible implementation, constructing a first correlation matrix representing the correlation between dimensions using the first dimension array includes:

[0022] For each array in the first dimension, calculate the mean and standard deviation;

[0023] Based on the first formula, the mean of the array, and the standard deviation of the array, each first-dimensional array is preprocessed to obtain the second-dimensional array, where the first formula is:

[0024]

[0025] In the formula, For the second dimension array One data point, For the first dimension array One data point, The mean of the first dimension array. The standard deviation of the first dimension array;

[0026] Extract dimension arrays from multiple second-dimensional arrays to serve as the arrays to be processed;

[0027] Calculate the dot product of the array to be processed with each of the other second-dimensional arrays, and add the dot product results to the first correlation matrix according to the dimension of the array to be processed and the dimensions of the other second dimensions for calculating the dot product;

[0028] If the traversal of the plurality of second-dimensional arrays is not completed, the process jumps to the step of extracting the dimension array from the plurality of second-dimensional arrays as the array to be processed.

[0029] In one possible implementation, the step of finding the closest queue for each second-state queue from multiple historical second-state queues as a reference queue, and determining the focus index based on the reference queue, includes:

[0030] For each second state queue, perform the following steps:

[0031] Obtain multiple historical second state queues, wherein the historical second state queues and the second state queues correspond to the same first terminal;

[0032] From the plurality of historical second state queues, find the queue that is closest to the nearest one and use it as the reference queue;

[0033] If the distance between the second state queue and the reference queue is greater than the first distance threshold, the focus index will be adjusted to the lowest level.

[0034] Otherwise, the focus index corresponding to the reference queue is used as the focus index corresponding to the second state queue.

[0035] In one possible implementation, the process of obtaining the focus index corresponding to the reference queue includes:

[0036] The first quantity is determined based on the dimension of the historical second state queue;

[0037] For each historical second state queue, find the first number of nearest neighbor queues and use the maximum distance to the nearest neighbor queue as the reference distance;

[0038] Sort the multiple reference distances according to their values ​​to obtain a reference distance queue;

[0039] Take two adjacent reference distances sequentially from the reference distance queue and perform a difference operation to obtain multiple difference distances;

[0040] The value with the largest absolute value is selected from the plurality of difference distances as the target difference distance;

[0041] The maximum value of the two source reference distances of the target differential distance is used as the first distance threshold;

[0042] For each historical second state queue, find the friendly neighbor queues that are less than the first distance threshold from multiple historical second state queues, and take the number of friendly neighbor queues as the number of friendly neighbors of the historical second state queue;

[0043] Sort the number of multiple friendly neighbors to obtain a queue of friendly neighbor counts;

[0044] The ratio of the rank of the number of friendly neighbors corresponding to the historical second state queue in the friendly neighbor count queue to the total number of data bits in the friendly neighbor count queue is used as the focus index of the historical second state queue.

[0045] In one possible implementation, determining the overall classroom activity index based on multiple focus indices includes:

[0046] Obtain multiple threshold pairs, where each threshold pair includes a focus threshold and a percentage threshold;

[0047] From the plurality of threshold pairs, threshold pairs are iteratively extracted, and after each extraction, the following steps are performed:

[0048] The percentage of the multiple focus indices that are lower than the extraction threshold and the focus threshold is counted as the statistical percentage, and the ratio of the statistical percentage to the percentage threshold is added to the activity index as the activity index.

[0049] After traversing the multiple threshold pairs, the activity index with the largest value is taken from the activity array as the overall classroom activity index.

[0050] In one possible implementation, the plurality of thresholds are determined based on a plurality of historical attention indices, including:

[0051] Obtain multiple historical focus indices;

[0052] Calculate the mean and standard deviation of the multiple historical focus indices as the focus mean and focus standard deviation;

[0053] Multiply the standard deviation of focus by multiple coefficients to obtain multiple first standard deviations;

[0054] Calculate the difference between the mean of the index and each first standard deviation, and use the result as the focus threshold;

[0055] Based on multiple focus thresholds and the second formula, a percentage threshold is determined, and the corresponding focus thresholds and percentage thresholds are constructed as threshold pairs. The second formula is:

[0056]

[0057] In the formula, This is the percentage threshold. Attention threshold For the standard deviation of attention, Pi It is a natural constant. The average level of focus. This is a focus index.

[0058] Secondly, embodiments of the present invention provide a classroom teaching interaction device for implementing the classroom teaching interaction method as described in the first aspect or any possible implementation thereof, the classroom teaching interaction device comprising:

[0059] The state queue acquisition module is used to acquire multiple first state queues, wherein each first state queue is acquired based on a corresponding first terminal, and the first state queue includes multiple first states representing facial expressions arranged in order according to time nodes.

[0060] The dimensionality reduction module is used to reduce the dimensionality of each first state queue by using the dimensionality reduction operator corresponding to the first state queue, so as to obtain the second state queue.

[0061] The focus index analysis module is used to find the closest queue for each second-state queue from multiple historical second-state queues as a reference queue, and determine the focus index based on the reference queue.

[0062] as well as,

[0063] The classroom activity determination module is used to determine the overall classroom activity index based on multiple attention indices. When the activity index is lower than the first index threshold, a target proposal is selected from multiple prepared proposals and sent to the first terminal for question and answer.

[0064] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement the steps of the method as described in the first aspect or any possible implementation of the first aspect.

[0065] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in the first aspect or any possible implementation thereof.

[0066] The beneficial effects of the embodiments of the present invention compared with the prior art are:

[0067] The classroom teaching interaction method of this invention first acquires multiple first state queues, each acquired based on a corresponding first terminal. Each first state queue includes multiple first states representing facial expressions arranged sequentially according to time nodes. Then, for each first state queue, dimensionality reduction is performed using a dimensionality reduction operator corresponding to that queue to obtain a second state queue. Next, the closest queue found for each second state queue from multiple historical second state queues is used as a reference queue, and a focus index is determined based on this reference queue. Finally, an overall classroom activity index is determined based on multiple focus indices. When the activity index is lower than a first index threshold, a target proposal is selected from multiple prepared proposals and sent to the first terminal for question-and-answer sessions. This invention determines student focus through dimensionality transformation and historical queue reference, reflecting the overall classroom activity through the focus of multiple students. Classroom teaching, based on the overall classroom activity, provides in-depth explanations of key content and quick explanations of general content, thus improving the utilization of classroom time. Attached Figure Description

[0068] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0069] Figure 1 This is a flowchart of the classroom teaching interaction method provided by an embodiment of the present invention;

[0070] Figure 2 This is a schematic diagram of the association matrix construction process provided by the embodiments of the present invention;

[0071] Figure 3This is a functional block diagram of the classroom teaching interactive device provided in the embodiments of the present invention;

[0072] Figure 4 This is a functional block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0073] In the following description, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0074] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0075] The embodiments of the present invention will be described in detail below. This example is implemented based on the technical solution of the present invention, and provides detailed implementation methods and specific operation processes. However, the protection scope of the present invention is not limited to the following embodiments.

[0076] Figure 1 A flowchart of a classroom teaching interaction method provided for an embodiment of the present invention.

[0077] like Figure 1 As shown, a flowchart illustrating the implementation of the classroom teaching interaction method provided by an embodiment of the present invention is presented, and is described in detail below:

[0078] In step 101, multiple first state queues are obtained, wherein each first state queue is obtained based on a corresponding first terminal, and the first state queue includes multiple first states representing facial expressions arranged in order according to time nodes.

[0079] For example, the present invention aims to provide a method for analyzing the overall classroom activity level by using facial expression states obtained from video analysis devices. In some current technologies, classroom application terminal devices, such as tablet computers, can capture students' facial expressions in real time and identify and classify the current state based on the facial expressions. These states may include: happy, sad, angry, fearful, surprised, disgusted, or neutral.

[0080] Classroom activity can reflect the teaching situation; however, since facial expressions reflect an individual's instantaneous state, using them as activity levels leads to uncertainties in the analysis results. Therefore, this invention proposes to statistically analyze multiple states for each individual, arranging these states into a queue according to time sequence: the first state queue. The focus of students during class is analyzed using this first state queue. Finally, based on the focus of multiple students, an overall classroom activity index is reflected. When activity levels are low, teaching methods can be adjusted; for example, sending questions to students to inquire about their current points of confusion, thus improving teaching efficiency.

[0081] To achieve the above objectives, this invention starts by performing dimensional transformation and focusing on the first state queue to analyze students' focus. The process described above will be discussed in detail below.

[0082] In step 102, for each first state queue, the dimensionality is reduced by the dimensionality reduction operator corresponding to the first state queue to obtain the second state queue.

[0083] In some implementations, the dimensionality reduction operator is constructed using multiple historical first-state queues, including:

[0084] Obtain multiple historical first state queues, wherein the historical first state queues have the same dimension as the first state queues, and the multiple historical first state queues and the current first state queue correspond to the same first terminal;

[0085] Data is extracted from the multiple historical first state queues one dimension at a time, and the multiple extracted data are constructed into a first dimension array;

[0086] A first correlation matrix representing the relationship between dimensions is constructed using the first dimension array;

[0087] Calculate multiple eigenvalues ​​of the first correlation matrix, and determine a first eigenvector based on each eigenvalue;

[0088] Delete the minimum values ​​among the plurality of first feature values;

[0089] The remaining first eigenvectors corresponding to the multiple first eigenvalues ​​are used to construct a dimensionality reduction operator.

[0090] In some implementations, constructing a first correlation matrix representing the correlation between dimensions using the first dimension array includes:

[0091] For each array in the first dimension, calculate the mean and standard deviation;

[0092] Based on the first formula, the mean of the array, and the standard deviation of the array, each first-dimensional array is preprocessed to obtain the second-dimensional array, where the first formula is:

[0093]

[0094] In the formula, For the second dimension array One data point, For the first dimension array One data point, The mean of the first dimension array. The standard deviation of the first dimension array;

[0095] Extract dimension arrays from multiple second-dimensional arrays to serve as the arrays to be processed;

[0096] Calculate the dot product of the array to be processed with each of the other second-dimensional arrays, and add the dot product results to the first correlation matrix according to the dimension of the array to be processed and the dimensions of the other second dimensions for calculating the dot product;

[0097] If the traversal of the plurality of second-dimensional arrays is not completed, the process jumps to the step of extracting the dimension array from the plurality of second-dimensional arrays as the array to be processed.

[0098] For example, in order to better reflect a student's level of concentration from facial expressions, the number of first states in the first state queue is usually increased. However, this increases the difficulty of determining the level of concentration. This invention proposes to use a dimensionality reduction operator to reduce the dimensionality of the first state queue.

[0099] In fact, the dimensionality reduction operator is a matrix constructed from multiple eigenvectors, and the dimensionality-reduced data queue is obtained using the following formula:

[0100]

[0101] In the formula, This is the dimensionality-reduced vector, which is directly transformed into the second state queue obtained after dimensionality reduction. The row vector constructed based on the first state queue. For dimensionality reduction operators;

[0102] For example, if the dimension of a first state queue is R, and the dimension reduction operator is a matrix with R rows and L columns, then the dimension of the second state queue obtained by the above operator is L. It can be known that L is less than R.

[0103] As can be seen, the aforementioned dimensionality reduction operator is a crucial data block for extracting data features. In fact, this data block is obtained through the historical first state queue, which is actually the first state queue obtained before the current moment.

[0104] When obtaining the operator through the historical first state queue, data is extracted from each historical first state queue dimension by dimension, and the data of the same dimension are constructed into a first dimension array. In other words, if the number of dimensions of the historical first state queue is R, then R first dimension arrays will be obtained, and the number of data in each first dimension array is the number of historical first state queues.

[0105] These first-dimensional arrays are subjected to mean and standard deviation calculations one by one. That is, each first-dimensional array yields a mean and a standard deviation. Using these two data points and the first formula, the first-dimensional arrays are preprocessed, and the preprocessed arrays are used as the second-dimensional arrays.

[0106]

[0107] In the formula, For the second dimension array One data point, For the first dimension array One data point, The mean of the first dimension array. is the standard deviation of the first dimension array.

[0108] Using these second-dimensional arrays, an association matrix is ​​constructed, specifically, as follows: Figure 2 As shown, the second-dimensional array 201 is sorted according to the previous dimensions. Then, for a given second-dimensional array 202, it is subjected to a dot product operation with the other second-dimensional arrays 201. The dot product result is then placed into the corresponding two dimensions of the association matrix 202. Figure 2 In the second second-dimensional array 201, the dot product operation is performed with the third and fourth second-dimensional arrays 201 respectively, and the results are stored in the third and fourth columns of the second row of the correlation matrix 202 respectively. In this way, a square matrix with the same number of dimensions is obtained: the correlation matrix 202.

[0109] By calculating the eigenvalues ​​and eigenvectors of the correlation matrix, we can see that there is a one-to-one correspondence between the eigenvalues ​​and eigenvectors. After deleting the smallest eigenvalues, the eigenvectors corresponding to the remaining eigenvalues ​​are the vectors that need to be retained. These vectors, as row vectors, constitute the operator.

[0110] In step 103, the closest queue found for each second state queue from multiple historical second state queues is used as a reference queue, and the focus index is determined based on the reference queue.

[0111] In some implementations, the step of finding the closest queue for each second-state queue from multiple historical second-state queues as a reference queue and determining the focus index based on the reference queue includes:

[0112] For each second state queue, perform the following steps:

[0113] Obtain multiple historical second state queues, wherein the historical second state queues and the second state queues correspond to the same first terminal;

[0114] From the plurality of historical second state queues, find the queue that is closest to the nearest one and use it as the reference queue;

[0115] If the distance between the second state queue and the reference queue is greater than the first distance threshold, the focus index will be adjusted to the lowest level.

[0116] Otherwise, the focus index corresponding to the reference queue is used as the focus index corresponding to the second state queue.

[0117] In some implementations, the process of obtaining the focus index corresponding to the reference queue includes:

[0118] The first quantity is determined based on the dimension of the historical second state queue;

[0119] For each historical second state queue, find the first number of nearest neighbor queues and use the maximum distance to the nearest neighbor queue as the reference distance;

[0120] Sort the multiple reference distances according to their values ​​to obtain a reference distance queue;

[0121] Take two adjacent reference distances sequentially from the reference distance queue and perform a difference operation to obtain multiple difference distances;

[0122] The value with the largest absolute value is selected from the plurality of difference distances as the target difference distance;

[0123] The maximum value of the two source reference distances of the target differential distance is used as the first distance threshold;

[0124] For each historical second state queue, find the friendly neighbor queues that are less than the first distance threshold from multiple historical second state queues, and take the number of friendly neighbor queues as the number of friendly neighbors of the historical second state queue;

[0125] Sort the number of multiple friendly neighbors to obtain a queue of friendly neighbor counts;

[0126] The ratio of the rank of the number of friendly neighbors corresponding to the historical second state queue in the friendly neighbor count queue to the total number of data bits in the friendly neighbor count queue is used as the focus index of the historical second state queue.

[0127] For example, in practice, each first-state queue obtained in each time period is reduced to a second-state queue in the manner described above. Attention analysis based on the second-state queue is actually performed by referring to historical second-state queues, which are queues from time periods preceding the second-state queue. Furthermore, attention analysis is conducted based on historical second-state queues corresponding to the same student. For instance, the aforementioned steps result in a second-state queue corresponding to the [student's] [time period]. A queue of students Among them, then it should be based on the student's previous second-state queue. , , …Analyzing the student's level of concentration, for the first… The same applies to each student.

[0128] In terms of focus analysis, this invention finds the nearest historical second-state queue from multiple historical second-state queues. If the distance to this queue is greater than a pre-set first distance threshold, it indicates poor focus, and the focus index is set to the lowest value. If the distance is less than the first distance threshold, the focus index of the nearest historical second-state queue is used as the focus index of that second-state queue.

[0129] Here, it's necessary to explain how to obtain the focus index for the historical second-state queue. First, based on the dimensions of the second-state queue, a primary quantity is determined, usually proportionally, for example, twice the number of dimensions. Then, for each historical second-state queue, the primary quantity with the nearest secondary quantity is found. The maximum distance between the primary quantity and its nearest secondary quantity is used as the reference distance for that historical second-state queue. Note that this distance must be calculated using the same method as the previously mentioned distances, such as Euclidean distance.

[0130] Through the aforementioned steps, each historical second-state queue corresponds to a reference distance, which is sorted by value. The sorted data is then subjected to a difference operation, which involves subtracting the smaller value from the larger one.

[0131] Next, the values ​​are sorted again according to their differences, and the maximum value is selected. The maximum value comes from two distance values, and we take the larger of the two distance values ​​from which the maximum value comes as the first distance threshold.

[0132] Next, for each historical second state queue, the number of historical second state queues that are less than the first distance threshold is counted. These numbers are marked as the number of friendly neighbors. These friendly neighbor numbers are then sorted again to obtain the friendly neighbor number queue.

[0133] Finally, the focus index of each historical second state queue is the ratio of the rank of its neighbor count in the neighbor count queue to the total number of neighbors in the neighbor count queue. For example, if the rank of a neighbor count in a historical second state queue is n1, and the total number of neighbors in the neighbor count queue is N, then the focus index of that historical second state queue is n1 / N.

[0134] In step 104, the overall classroom activity index is determined based on multiple attention indices. When the activity index is lower than the first index threshold, a target proposal is selected from multiple prepared proposals and sent to the first terminal for questioning and answering.

[0135] In some implementations, determining the overall classroom activity index based on multiple focus indices includes:

[0136] Obtain multiple threshold pairs, where each threshold pair includes a focus threshold and a percentage threshold;

[0137] From the plurality of threshold pairs, threshold pairs are iteratively extracted, and after each extraction, the following steps are performed:

[0138] The percentage of the multiple focus indices that are lower than the extraction threshold and the focus threshold is counted as the statistical percentage, and the ratio of the statistical percentage to the percentage threshold is added to the activity index as the activity index.

[0139] After traversing the multiple threshold pairs, the activity index with the largest value is taken from the activity array as the overall classroom activity index.

[0140] In some implementations, the plurality of thresholds are determined based on a plurality of historical attention indices, including:

[0141] Obtain multiple historical focus indices;

[0142] Calculate the mean and standard deviation of the multiple historical focus indices as the focus mean and focus standard deviation;

[0143] Multiply the standard deviation of focus by multiple coefficients to obtain multiple first standard deviations;

[0144] Calculate the difference between the mean of the index and each first standard deviation, and use the result as the focus threshold;

[0145] Based on multiple focus thresholds and the second formula, a percentage threshold is determined, and the corresponding focus thresholds and percentage thresholds are constructed as threshold pairs. The second formula is:

[0146]

[0147] In the formula, This is the percentage threshold. Attention threshold For the standard deviation of attention, Pi It is a natural constant. The average level of focus. This is a focus index.

[0148] For example, after each student's focus level is calculated through the aforementioned steps (usually the focus level is calculated using the terminals used by the students), the overall classroom activity index is determined based on the focus levels of multiple students.

[0149] In fact, this process is a value comparison process. This invention provides multiple threshold pairs, each of which includes a focus threshold and a percentage threshold.

[0150] Multiple focus indices are compared using threshold pairs. For example, for a certain threshold pair, based on the focus threshold in that threshold pair, the percentage of multiple focus indices that are lower than the focus threshold in that threshold pair is counted. The ratio of the counted percentage to the percentage threshold is the activity index given by that threshold pair.

[0151] For example, a threshold pair includes a focus threshold FT and a percentage threshold R. First, the percentage of multiple focus indices that are lower than the focus threshold FT is calculated, for example, r (statistical percentage). Then, the activity index given by the threshold pair is r / R.

[0152] As we can see, each threshold pair gives an activity index, and we take the maximum value as the overall classroom activity index.

[0153] Through the above process, we found that the focus threshold and the percentage threshold are two key values ​​for judging activity levels. In fact, these two values ​​are obtained through multiple historical focus indices.

[0154] First, the mean and standard deviation of multiple historical focus indices are calculated.

[0155] Then, multiply the standard deviation by a coefficient, such as 2 or 3, and subtract the mean from the standard deviation after multiplying by the coefficient (subtract 2σ or 3σ) to obtain multiple attention thresholds.

[0156] Finally, using the second formula based on the focus threshold, the percentage threshold is determined, and the two corresponding thresholds are combined into a threshold pair. The second formula is:

[0157]

[0158] In the formula, This is the percentage threshold. Attention threshold For the standard deviation of attention, Pi It is a natural constant. The average level of focus. This is a focus index.

[0159] One application scenario of this invention includes an educational robot, a large screen, and a tablet.

[0160] The educational robot, acting as the front-end sensing device, captures students' facial expressions every second using its built-in camera. Combined with AI recognition, it categorizes the students' states into four types: focused, interactive, distracted, and drowsy. A status queue is generated based on "timestamp + student ID + status," and transmitted in real-time to a large screen via the classroom's local area network. For example, if a robot monitors three students, it can generate a queue containing 20 data points within one minute, which is then immediately pushed to the large screen.

[0161] The large screen receives queues of multiple robots (such as three robots covering the front, middle and back rows of the classroom), integrates all data according to timestamps, and then determines the overall activity level of the classroom through the method provided by this invention. The results are then transmitted back to the Pad in real time.

[0162] The Pad receives activity data and automatically suggests initiating group quizzes when activity is low. After the teacher confirms, the questions are simultaneously projected onto the large screen. The robot then captures the students' hand-raising and interaction status, and the new queue is sent to the large screen to recalculate the activity level, forming an interactive closed loop of "collection-analysis-adjustment-re-monitoring" to dynamically adapt to classroom teaching.

[0163] The classroom teaching interaction method of this invention first acquires multiple first state queues, each acquired based on a corresponding first terminal. Each first state queue includes multiple first states representing facial expressions arranged sequentially according to time nodes. Then, for each first state queue, dimensionality reduction is performed using a dimensionality reduction operator corresponding to that queue to obtain a second state queue. Next, the closest queue found for each second state queue from multiple historical second state queues is used as a reference queue, and a focus index is determined based on this reference queue. Finally, an overall classroom activity index is determined based on multiple focus indices. When the activity index is lower than a first index threshold, a target proposal is selected from multiple prepared proposals and sent to the first terminal for question-and-answer sessions. This invention determines student focus through dimensionality transformation and historical queue reference, reflecting the overall classroom activity through the focus of multiple students. Classroom teaching, based on the overall classroom activity, provides in-depth explanations of key content and quick explanations of general content, thus improving the utilization of classroom time.

[0164] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0165] The following are embodiments of the apparatus of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0166] Figure 3 This is a functional block diagram of the classroom teaching interactive device provided in the embodiments of the present invention, with reference to... Figure 3 The classroom teaching interaction device includes: a state queue acquisition module 301, a dimensionality reduction module 302, a focus index analysis module 303, and a classroom activity determination module 304, wherein:

[0167] The state queue acquisition module 301 is used to acquire multiple first state queues, wherein each first state queue is acquired based on a corresponding first terminal, and the first state queue includes multiple first states representing facial expressions arranged in order according to time nodes.

[0168] The dimensionality reduction module 302 is used to reduce the dimensionality of each first state queue by using the dimensionality reduction operator corresponding to the first state queue to obtain the second state queue.

[0169] The focus index analysis module 303 is used to find the closest queue for each second state queue from multiple historical second state queues as a reference queue, and determine the focus index based on the reference queue.

[0170] The classroom activity determination module 304 is used to determine the overall classroom activity index based on multiple attention indices, and when the activity index is lower than the first index threshold, it selects a target proposal from multiple prepared proposals and sends it to the first terminal for question and answer.

[0171] Figure 4 This is a functional block diagram of the electronic device provided in an embodiment of the present invention. For example... Figure 4 As shown, the electronic device 4 of this embodiment includes a processor 400 and a memory 401, wherein the memory 401 stores a computer program 402 that can run on the processor 400. When the processor 400 executes the computer program 402, it implements the steps of the various classroom teaching interaction methods and embodiments described above, for example... Figure 1 Steps 101 to 104 are shown.

[0172] For example, the computer program 402 may be divided into one or more modules / units, which are stored in the memory 401 and executed by the processor 400 to complete the present invention.

[0173] The electronic device 4 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. The electronic device 4 may include, but is not limited to, a processor 400 and a memory 401. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 4 and does not constitute a limitation on electronic device 4. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 4 may also include input / output devices, network access devices, buses, etc.

[0174] The processor 400 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0175] The memory 401 can be an internal storage unit of the electronic device 4, such as a hard disk or memory. The memory 401 can also be an external storage device of the electronic device 4, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 401 can include both internal and external storage units of the electronic device 4. The memory 401 is used to store the computer program 402 and other programs and data required by the electronic device 4. The memory 401 can also be used to temporarily store data that has been output or will be output.

[0176] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the aforementioned method embodiments, and will not be repeated here.

[0177] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0178] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0179] In the embodiments provided by this invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0180] The units described as separate components may or may not be physically separate. 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 the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0181] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0182] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above-described embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various methods and apparatus embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0183] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A classroom teaching interaction method, characterized in that, include: Multiple first state queues are obtained, wherein each first state queue is obtained based on a corresponding first terminal, and the first state queue includes multiple first states representing facial expressions arranged in order according to time nodes; For each first state queue, the second state queue is obtained by performing dimensionality reduction using the dimensionality reduction operator corresponding to the first state queue. The closest queue found for each second-state queue from multiple historical second-state queues is used as a reference queue, and a focus index is determined based on the reference queue, including: For each second state queue, perform the following steps: Obtain multiple historical second state queues, wherein the historical second state queues and the second state queues correspond to the same first terminal; From the plurality of historical second state queues, find the queue that is closest to the nearest one and use it as the reference queue; If the distance between the second state queue and the reference queue is greater than the first distance threshold, the focus index will be adjusted to the lowest level. Otherwise, the focus index corresponding to the reference queue is used as the focus index corresponding to the second state queue; The overall classroom activity index is determined based on multiple attention indices. When the activity index is lower than the first index threshold, a target proposal is selected from multiple prepared proposals and sent to the first terminal for questioning and answering. The process of obtaining the focus index corresponding to the reference queue includes: The first quantity is determined based on the dimension of the historical second state queue; For each historical second state queue, find the first number of nearest neighbor queues and use the maximum distance to the nearest neighbor queue as the reference distance; Sort the multiple reference distances according to their values ​​to obtain a reference distance queue; Take two adjacent reference distances sequentially from the reference distance queue and perform a difference operation to obtain multiple difference distances; The value with the largest absolute value is selected from the plurality of difference distances as the target difference distance; The maximum value of the two source reference distances of the target differential distance is used as the first distance threshold; For each historical second state queue, find the friendly neighbor queues that are less than the first distance threshold from multiple historical second state queues, and take the number of friendly neighbor queues as the number of friendly neighbors of the historical second state queue; Sort the number of multiple friendly neighbors to obtain a queue of friendly neighbor counts; The ratio of the rank of the number of friendly neighbors corresponding to the historical second state queue in the friendly neighbor count queue to the total number of data bits in the friendly neighbor count queue is used as the focus index of the historical second state queue.

2. The classroom teaching interaction method according to claim 1, characterized in that, The dimensionality reduction operator is constructed using multiple historical first-state queues, including: Obtain multiple historical first state queues, wherein the historical first state queues have the same dimension as the first state queues, and the multiple historical first state queues and the current first state queue correspond to the same first terminal; Data is extracted from the multiple historical first state queues one dimension at a time, and the multiple extracted data are constructed into a first dimension array; A first correlation matrix representing the relationship between dimensions is constructed using the first dimension array; Calculate multiple eigenvalues ​​of the first correlation matrix, and determine a first eigenvector based on each eigenvalue; Delete the minimum values ​​among the multiple feature values; The first eigenvectors corresponding to the remaining multiple eigenvalues ​​are used to construct a dimensionality reduction operator.

3. The classroom teaching interaction method according to claim 2, characterized in that, The step of constructing a first correlation matrix representing the correlation between dimensions using the first dimension array includes: For each array in the first dimension, calculate the mean and standard deviation; Based on the first formula, the mean of the array, and the standard deviation of the array, each first-dimensional array is preprocessed to obtain the second-dimensional array, where the first formula is: In the formula, For the second dimension array One data point, For the first dimension array One data point, The mean of the first dimension array. The standard deviation of the first-dimensional array; Extract dimension arrays from multiple second-dimensional arrays to serve as the arrays to be processed; Calculate the dot product of the array to be processed with each of the other second-dimensional arrays, and add the dot product results to the first correlation matrix according to the dimension of the array to be processed and the dimensions of the other second dimensions for calculating the dot product; If the traversal of the plurality of second-dimensional arrays is not completed, the process jumps to the step of extracting the dimension array from the plurality of second-dimensional arrays as the array to be processed.

4. The classroom teaching interaction method according to any one of claims 1-3, characterized in that, The determination of the overall classroom activity index based on multiple focus indices includes: Obtain multiple threshold pairs, where each threshold pair includes a focus threshold and a percentage threshold; From the plurality of threshold pairs, threshold pairs are iteratively extracted, and after each extraction, the following steps are performed: The percentage of the multiple focus indices that are lower than the extraction threshold and the focus threshold is counted as the statistical percentage, and the ratio of the statistical percentage to the percentage threshold is added to the activity index as the activity index. After traversing the multiple threshold pairs, the activity index with the largest value is taken from the activity array as the overall classroom activity index.

5. The classroom teaching interaction method according to claim 4, characterized in that, The multiple thresholds are determined based on multiple historical attention indices, including: Obtain multiple historical focus indices; Calculate the mean and standard deviation of the multiple historical focus indices as the focus mean and focus standard deviation; Multiply the standard deviation of focus by multiple coefficients to obtain multiple first standard deviations; Calculate the difference between the mean of the multiple historical focus indices and each first standard deviation, and use the result as the focus threshold; Based on multiple focus thresholds and the second formula, a percentage threshold is determined, and the corresponding focus thresholds and percentage thresholds are constructed as threshold pairs. The second formula is: In the formula, This is the percentage threshold. Attention threshold For the standard deviation of attention, Pi It is a natural constant. The average level of focus. This is a focus index.

6. A classroom teaching interactive device, characterized in that, For implementing the classroom teaching interaction method as described in any one of claims 1-5, the classroom teaching interaction device includes: The state queue acquisition module is used to acquire multiple first state queues, wherein each first state queue is acquired based on a corresponding first terminal, and the first state queue includes multiple first states representing facial expressions arranged in order according to time nodes. The dimensionality reduction module is used to reduce the dimensionality of each first state queue by using the dimensionality reduction operator corresponding to the first state queue, so as to obtain the second state queue. The focus index analysis module is used to find the closest queue for each second-state queue from multiple historical second-state queues as a reference queue, and determine the focus index based on the reference queue. as well as, The classroom activity determination module is used to determine the overall classroom activity index based on multiple attention indices. When the activity index is lower than the first index threshold, a target proposal is selected from multiple prepared proposals and sent to the first terminal for question and answer.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images