Student exercise quality monitoring method and apparatus

Abstract

The application provides a student exercise quality monitoring method and device, and relates to the technical fields of artificial intelligence and monitoring. The method comprises the following steps: obtaining original exercise data of a student; training a student exercise quality monitoring model based on the original exercise data; inputting to-be-monitored exercise data of the student into the trained student exercise quality monitoring model; and judging whether the student has a situation of replacing exercise based on the output of the student exercise quality monitoring model. The student exercise quality monitoring method and device provided by the application can construct a suitable training set containing replacement exercise data from original non-replacement exercise data, so that the trained student exercise quality monitoring model can accurately identify whether the student has a situation of replacing exercise, without manual supervision, thereby saving labor costs and improving the efficiency of monitoring and identification.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and monitoring technology, and in particular to a method and device for monitoring the quality of student exercise. Background Technology

[0002] Physical fitness has always been an important indicator of talent cultivation in higher education institutions. Good physical fitness is crucial for students' mental and physical health, as well as their learning efficiency. Currently, many universities encourage students to run in groups to improve their physical fitness. However, some students may not take this seriously for various reasons, leading to instances of students hiring others to run for them. At present, this phenomenon can only be prevented through manual supervision, which undoubtedly increases manpower costs. Therefore, a robust system for detecting instances of students running for others can improve the management efficiency of relevant departments in universities, allowing them to promptly understand the specific situation of affected students and provide appropriate care and guidance. Summary of the Invention

[0003] In view of this, the present invention provides a method and apparatus for monitoring the quality of student exercise to solve at least one of the problems mentioned above.

[0004] To achieve the above objectives, the present invention adopts the following solution:

[0005] According to a first aspect of the present invention, a method for monitoring the quality of student exercise is provided. The method includes: acquiring raw exercise data of students; training a student exercise quality monitoring model based on the raw exercise data; inputting the student's exercise data to be monitored into the trained student exercise quality monitoring model; and determining whether the student has been using a substitute for another student based on the output of the student exercise quality monitoring model.

[0006] As an embodiment of the present invention, the student's original exercise data in the above method includes: the student's historical physiological characteristics data during exercise and the physiological characteristics data of the student's most recent exercise.

[0007] As an embodiment of the present invention, the method described above for training a student exercise quality monitoring model based on the original exercise data includes: dividing the original exercise data into normal data and abnormal data; the normal data includes the student's historical physiological characteristics data during exercise and the student's most recent exercise physiological characteristics data; the abnormal data includes the student's historical physiological characteristics data during exercise and the physiological characteristics data of other students during their most recent exercise; inputting the divided original exercise data into a first Bi-LSTM network and a second Bi-LSTM network respectively to extract a first feature and a second feature, wherein the first feature is a feature of the historical physiological characteristics data and the second feature is a feature of the most recent exercise physiological characteristics data; inputting the first feature and the second feature into an FNN network for comparison and calculation, wherein the two-dimensional vector output by the FNN network represents the probability of whether the student has been trained by someone else.

[0008] As an embodiment of the present invention, the above method further includes: dividing the acquired raw exercise data into a validation set, and verifying the trained student exercise quality monitoring model based on the validation set until a preset accuracy is met.

[0009] As an embodiment of the present invention, the above method further includes: acquiring students' physiological signs data based on wearable smart devices as exercise data to be monitored.

[0010] According to a second aspect of the present invention, a student exercise quality monitoring device is provided, the device comprising: a data acquisition unit for acquiring raw exercise data of students; a model training unit for training a student exercise quality monitoring model based on the raw exercise data; a data input unit for inputting the student's exercise data to be monitored into the trained student exercise quality monitoring model; and a monitoring unit for determining whether the student has been using a substitute trainer based on the output of the student exercise quality monitoring model.

[0011] As an embodiment of the present invention, the original exercise data of the student mentioned above includes: the student's historical physiological characteristics data during exercise and the physiological characteristics data of the student's most recent exercise.

[0012] As an embodiment of the present invention, the above-mentioned model training unit includes: a data partitioning module, used to partition the original exercise data into normal data and abnormal data, wherein the normal data includes the student's historical physiological characteristics data during exercise and the student's most recent exercise physiological characteristics data; the abnormal data includes the student's historical physiological characteristics data during exercise and the physiological characteristics data of other students during their most recent exercise; a feature extraction module, used to input the partitioned original exercise data into a first Bi-LSTM network and a second Bi-LSTM network respectively to extract a first feature and a second feature, wherein the first feature is a feature of the historical physiological characteristics data and the second feature is a feature of the most recent exercise physiological characteristics data; and a comparison calculation module, used to input the first feature and the second feature into an FNN network for comparison calculation, wherein the two-dimensional vector output by the FNN network represents the probability of whether the student is being trained by a substitute.

[0013] As an embodiment of the present invention, the above-mentioned device further includes: a model verification unit, used to divide the acquired raw exercise data into a verification set, and verify the trained student exercise quality monitoring model based on the verification set until a preset accuracy is met.

[0014] As an embodiment of the present invention, the above-mentioned device further includes: a data acquisition unit for acquiring students' physiological signs data based on wearable smart devices as exercise data to be monitored.

[0015] According to a third aspect of the present invention, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method.

[0016] According to a fourth aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0017] As can be seen from the above technical solution, the student exercise quality monitoring method and device provided by the present invention can construct a suitable training set containing training data from the original data without training substitutes, so that the trained student exercise quality monitoring model can accurately identify whether students have training substitutes, without the need for manual supervision, saving labor costs and improving the efficiency of monitoring and identification. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, 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. In the drawings:

[0019] Figure 1 This is a flowchart illustrating a method for monitoring student exercise quality provided in an embodiment of this application;

[0020] Figure 2 This is a flowchart illustrating a method for monitoring student exercise quality according to another embodiment of this application;

[0021] Figure 3 This is a schematic diagram of the structure of a student exercise quality monitoring device provided in an embodiment of this application;

[0022] Figure 4 This is a schematic diagram of the structure of the model training unit provided in the embodiments of this application;

[0023] Figure 5 This is a schematic diagram of the structure of a student exercise quality monitoring device provided in another embodiment of this application;

[0024] Figure 6 This is a schematic block diagram of the system configuration of the electronic device provided in the embodiments of this application. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0026] Currently, schools can only prevent students from running or training on behalf of others through manual supervision, which increases labor costs. Therefore, the purpose of this application is to provide a method and device for monitoring student exercise quality, which can accurately identify whether students are training on behalf of others without the need for manual supervision, thus saving labor costs.

[0027] The acquisition, storage, use, and processing of data in the technical solutions of the following embodiments of this application all comply with the relevant provisions of laws and regulations.

[0028] like Figure 1 The diagram shown is a flowchart of a student exercise quality monitoring method provided in an embodiment of this application. The method includes the following steps:

[0029] Step S101: Obtain the student's raw exercise data.

[0030] In this embodiment, the student's exercise data refers to the physiological characteristic data generated during exercise, such as pulse, heart rate, and blood pressure. This data can be collected by students wearing wearable devices specifically designed to extract physiological characteristic data during exercise, such as smart bracelets or smart rings.

[0031] Raw data can be accumulated in various ways. For example, when a new batch of students begins training with the smart bracelet system, their ECG data can be obtained under reliable conditions. These reliable conditions could include exercising with the smart bracelet under the supervision of a teacher or teaching assistant during physical education classes or group workouts. In this way, more and more raw exercise data can be accumulated. As the raw exercise data accumulates, the accuracy of the subsequently trained model will also increase.

[0032] Step S102: Train a student exercise quality monitoring model based on the original exercise data.

[0033] In this embodiment, the student exercise quality monitoring model is a deep learning binary classification model, which can output results for two scenarios: with or without substitute training.

[0034] Step S103: Input the student's exercise data to be monitored into the trained student exercise quality monitoring model.

[0035] Preferably, prior to this step, this embodiment may further include the following step: acquiring the student's physiological characteristics data based on a wearable smart device as the exercise data to be monitored. That is, before the student exercises, they must wear a smart device to acquire physiological characteristics data during the exercise, thereby obtaining the student's physiological characteristics data for this exercise.

[0036] Step S104: Determine whether a student is being trained by a substitute based on the output of the student exercise quality monitoring model.

[0037] As can be seen from the above technical solution, the student exercise quality monitoring method provided by the present invention identifies whether students have substitute training during exercise through a deep learning student exercise quality monitoring model. Therefore, no manual supervision is required, saving labor costs and improving the efficiency of monitoring and identification.

[0038] like Figure 2 The diagram shown is a flowchart of a student exercise quality monitoring method according to another embodiment of this application. The method includes the following steps:

[0039] Step S201: Obtain the student's historical physiological signs data during exercise and the physiological signs data of the student's most recent exercise.

[0040] In other words, each piece of data in the student's original exercise data obtained in this application contains two parts: one part is the student's historical physiological characteristics data during previous exercises, and the other part is the physiological characteristics data of the student's most recent exercise. This allows the subsequently trained model to identify whether the physiological characteristics data of the most recent exercise are the same as the student's previous physiological characteristics data, thereby determining whether there is any substitution training. It should be noted that these two parts of data must not overlap in the time dimension; that is, there should be no overlapping data.

[0041] Step S202: Divide the raw exercise data into normal data and abnormal data. The normal data includes the student's historical physiological characteristics data during exercise and the student's most recent exercise physiological characteristics data. The abnormal data includes the student's historical physiological characteristics data during exercise and the physiological characteristics data of other students during their most recent exercise.

[0042] As mentioned above, normal data is taken from the same student, while the historical physiological characteristic data and the physiological characteristic data during the most recent exercise in abnormal data are taken from two different students. For normal data, for example, a continuous ECG data segment of length m can be extracted from all the student's original data as historical physiological characteristic data, and then a data segment of length k with no overlap with the historical physiological characteristic data can be extracted as the most recent exercise data. It should be noted that the most recent exercise data can also precede the historical physiological characteristic data in the time dimension, but in order to better represent the student's recent physiological state, it is preferable that the time dimension of the most recent exercise data should be after the historical physiological characteristic data. For abnormal data, the data construction process is similar to that of normal data. The difference is that a continuous ECG data segment of length m can be extracted from student A's data as historical physiological characteristic data, and then a data segment of length k can be extracted from student B's data as the most recent exercise data. There is no restriction on the time dimension between the two parts of the abnormal data.

[0043] Step S203: Input the divided raw exercise data into the first Bi-LSTM network and the second Bi-LSTM network respectively to extract the first feature and the second feature. The first feature is the feature of historical physiological signs data, and the second feature is the feature of physiological signs data of the most recent exercise.

[0044] Step S204: Input the first feature and the second feature into an FNN network for comparison and calculation. The two-dimensional vector output by the FNN network represents the probability of whether the student has been cheated on.

[0045] Steps S202-S203 above describe the training process of a binary classification model in this application. The overall approach to selecting the binary classification model is to first extract features from past and recent exercise data, and then determine whether the two are similar based on these features. This model employs contrastive learning, treating the exercise data of different students as different categories. The goal is for the model to learn the features of different categories and then complete the classification. Although the number of student categories may seem large (equal to the number of students), the model does not need to distinguish each category. It only needs to learn how to compare two categories and output simple binary classification labels (i.e., for past and recent data, outputting a judgment result indicating whether they belong to the same or different categories). This simplifies the network structure, reduces network parameters and training overhead, and improves network performance.

[0046] In the model structure design, considering that the physiological characteristics data of each student are a continuous time series, this application uses two Bi-LSTM networks to extract their features: one to extract features from the student's historical physiological characteristics data, and the other to extract features from the physiological characteristics data of the most recent exercise (i.e., the features of the data to be judged in this instance). Specifically, the sequence is input into the Bi-LSTM, and the average of the outputs of each hidden layer is taken as its feature. After feature extraction, the features of the two sequences are input into an FNN network for comparison and calculation. The FNN network then outputs a two-dimensional vector (a0, a1) representing the probability of substitution. Specifically, when the model labels position 0 of the two-dimensional vector as the similar data label, if a0 > a1, it indicates no substitution phenomenon, and if a0 < a1, it indicates substitution phenomenon.

[0047] For example, assuming the ECG data at each time point is a vector of length n, then the past exercise data should be a tensor of length m×n (m is the length of the historical physiological data), and the recent exercise data should be a tensor of length k×n (k is the length of the most recent exercise data). Inputting these into two Bi-LSTMs respectively will yield two tensors of length m×h and k×h, whose means should both be vectors of length h. This represents the features extracted by the model from the two sets of data. Finally, these two vectors of length h are concatenated into a vector of length 2h, which is then input into a 2h-to-2 fully connected network. The final output should be a two-dimensional vector of length 2. Applying a softmax transformation to this vector yields the probability that the model classifies an element as belonging to one class or not. For example, if the final output is [2.3, 8.9], and the 0th position is set as the similar data label position when training the model, then since the value of the 0th position (2.3) is less than the value of the 1st position (8.9), it should be judged as dissimilar, which means that the student's most recent data is more likely to be the result of someone else training on their behalf.

[0048] Step S205: Divide the acquired raw exercise data into a validation set, and validate the trained student exercise quality monitoring model based on the validation set until the preset accuracy is met.

[0049] Step S206: Obtain students' physiological signs data based on wearable smart devices as exercise data to be monitored.

[0050] Step S207: Input the exercise data to be monitored into the trained student exercise quality monitoring model.

[0051] Step S208: Determine whether a student is being trained by a substitute based on the output of the student exercise quality monitoring model.

[0052] As can be seen from the above technical solution, the student exercise quality monitoring method provided by the present invention can construct a suitable training set containing training data from the original data without training substitutes, so that the trained student exercise quality monitoring model can accurately identify whether students have training substitutes, without the need for manual supervision, saving labor costs and improving the efficiency of monitoring and identification.

[0053] like Figure 3 The diagram shown is a structural schematic of a student exercise quality monitoring device provided in an embodiment of this application. The device includes: a data acquisition unit 310, a model training unit 320, a data input unit 330, and a monitoring unit 340, which are connected sequentially.

[0054] The data acquisition unit 310 is used to acquire the student's raw exercise data.

[0055] The model training unit 320 is used to train a student exercise quality monitoring model based on the original exercise data.

[0056] The data input unit 330 is used to input the student's exercise data to be monitored into the trained student exercise quality monitoring model.

[0057] The monitoring unit 340 is used to determine whether a student has been trained by a substitute based on the output of the student exercise quality monitoring model.

[0058] Preferably, the original exercise data of the student acquired by the data acquisition unit 310 includes: the student's historical physiological characteristics data during exercise and the physiological characteristics data of the student's most recent exercise.

[0059] Preferred, such as Figure 4 As shown, the model training unit 320 includes: a data partitioning module 321, a feature extraction module 322, and a comparison calculation module 323, which are connected sequentially. Wherein:

[0060] The data segmentation module 321 is used to divide the original exercise data into normal data and abnormal data. The normal data includes the student's historical physiological characteristics data during exercise, as well as the student's physiological characteristics data during the most recent exercise. The abnormal data includes the student's historical physiological characteristics data during exercise, as well as the physiological characteristics data of other students during their most recent exercise.

[0061] The feature extraction module 322 is used to input the segmented raw exercise data into the first Bi-LSTM network and the second Bi-LSTM network respectively to extract the first feature and the second feature. The first feature is the feature of historical physiological signs data, and the second feature is the feature of the physiological signs data of the most recent exercise.

[0062] The comparison calculation module 323 is used to input the first feature and the second feature into an FNN network for comparison calculation. The two-dimensional vector output by the FNN network represents the probability of whether the student has been cheated on.

[0063] Preferred, such as Figure 5 As shown, the above-mentioned device also includes: a model verification unit 350, which is connected to the data acquisition unit 310 and the model training unit 320 respectively, for dividing the acquired raw exercise data into a verification set, and verifying the trained student exercise quality monitoring model based on the verification set until the preset accuracy is met.

[0064] Preferred, such as Figure 5 As shown, the above-mentioned device also includes: a data acquisition unit 360, which is connected to the model training unit 320 and the data input unit 330 respectively, and is used to acquire students' physiological signs data based on wearable smart devices as exercise data to be monitored.

[0065] For a detailed description of each of the above units, please refer to the corresponding description in the foregoing method embodiments, which will not be repeated here.

[0066] As can be seen from the above technical solution, the student exercise quality monitoring device provided by the present invention can construct a suitable training set containing training data from the original data without training substitutes, so that the trained student exercise quality monitoring model can accurately identify whether students have training substitutes, without the need for manual supervision, saving labor costs and improving the efficiency of monitoring and identification.

[0067] This invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method.

[0068] This invention also provides a computer-readable storage medium storing a computer program for performing the above-described methods.

[0069] like Figure 6 As shown, the electronic device 600 may also include: a communication module 110, an input unit 120, an audio processor 130, a display 160, and a power supply 170. It is worth noting that the electronic device 600 does not necessarily need to include these components. Figure 6 All components shown; in addition, the electronic device 600 may also include Figure 6 For components not shown, please refer to existing technologies.

[0070] like Figure 6 As shown, the central processing unit 100, sometimes also referred to as a controller or operating control, may include a microprocessor or other processor device and / or logic device. The central processing unit 100 receives inputs and controls the operation of various components of the electronic device 600.

[0071] The memory 140 may be, for example, one or more of a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It may store the aforementioned failure-related information, and also store a program for executing that information. The central processing unit 100 may execute the program stored in the memory 140 to perform information storage or processing, etc.

[0072] Input unit 120 provides input to central processing unit 100. Input unit 120 may be, for example, a keypad or touch input device. Power supply 170 provides power to electronic device 600. Display 160 displays images and text. Display may be, for example, an LCD display, but is not limited thereto.

[0073] The memory 140 can be a solid-state memory, such as a read-only memory (ROM), random access memory (RAM), a SIM card, etc. It can also be a memory that retains information even when power is off, can be selectively erased, and contains more data; examples of this type of memory are sometimes referred to as EPROMs. The memory 140 can also be some other type of device. The memory 140 includes a buffer memory 141 (sometimes referred to as a buffer). The memory 140 may include an application / function storage unit 142 for storing application programs and function programs or processes for executing the operation of the electronic device 600 via the central processing unit 100.

[0074] The memory 140 may also include a data storage unit 143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 144 of the memory 140 may include various drivers for the electronic device's communication functions and / or for performing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0075] The communication module 110 is a transmitter / receiver 110 that transmits and receives signals via antenna 111. The communication module (transmitter / receiver) 110 is coupled to the central processing unit 100 to provide input signals and receive output signals, which can be the same as in a conventional mobile communication terminal.

[0076] Based on different communication technologies, multiple communication modules 110 can be configured in the same electronic device, such as cellular network modules, Bluetooth modules, and / or wireless LAN modules. The communication module (transmitter / receiver) 110 is also coupled to a speaker 131 and a microphone 132 via an audio processor 130 to provide audio output via the speaker 131 and receive audio input from the microphone 132, thereby enabling typical telecommunications functions. The audio processor 130 may include any suitable buffer, decoder, amplifier, etc. Additionally, the audio processor 130 is coupled to a central processing unit 100, enabling on-device recording via the microphone 132 and on-device playback of stored audio via the speaker 131.

[0077] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0078] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0079] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0080] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0081] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method of monitoring the quality of a student exercise, characterized by, The method comprises: obtaining original exercise data of a student, the original exercise data of the student comprising: historical physiological sign data of the student when exercising and physiological sign data of the student corresponding to the last exercise of the student; training a student exercise quality monitoring model based on the original exercise data; inputting to-be-monitored exercise data of the student into the trained student exercise quality monitoring model; judging whether the student has a situation of replacing exercise based on an output of the student exercise quality monitoring model; the training of the student exercise quality monitoring model based on the original exercise data comprises: dividing the original exercise data into normal data and abnormal data, the normal data comprising the historical physiological sign data of the student when exercising and the physiological sign data of the student corresponding to the last exercise of the student; and the abnormal data comprising the historical physiological sign data of the student when exercising and physiological sign data of other students corresponding to the last exercise of the other students; inputting the divided original exercise data into a first Bi-LSTM network and a second Bi-LSTM network respectively to extract first features and second features, the first features being features of the historical physiological sign data, and the second features being features of the physiological sign data corresponding to the last exercise of the student; inputting the first features and the second features into an FNN network for comparison calculation, a two-dimensional vector output by the FNN network representing a probability of whether the student replaces exercise.

2. The student exercise quality monitoring method of claim 1, wherein, The method further comprises: dividing the obtained original exercise data into a verification set, verifying the trained student exercise quality monitoring model based on the verification set, and stopping until a preset accuracy is met.

3. The student exercise quality monitoring method of claim 1, wherein, The method further comprises: obtaining physiological sign data of the student based on a wearable smart device as the to-be-monitored exercise data.

4. A student exercise quality monitoring device, characterized by, The device comprises: a data acquisition unit configured to obtain original exercise data of a student, the original exercise data of the student comprising: historical physiological sign data of the student when exercising and physiological sign data of the student corresponding to the last exercise of the student; a model training unit configured to train a student exercise quality monitoring model based on the original exercise data; a data input unit configured to input to-be-monitored exercise data of the student into the trained student exercise quality monitoring model; a monitoring unit configured to judge whether the student has a situation of replacing exercise based on an output of the student exercise quality monitoring model; the model training unit comprises: a data division module configured to divide the original exercise data into normal data and abnormal data, the normal data comprising the historical physiological sign data of the student when exercising and the physiological sign data of the student corresponding to the last exercise of the student; and the abnormal data comprising the historical physiological sign data of the student when exercising and physiological sign data of other students corresponding to the last exercise of the other students; a feature extraction module configured to input the divided original exercise data into a first Bi-LSTM network and a second Bi-LSTM network respectively to extract first features and second features, the first features being features of the historical physiological sign data, and the second features being features of the physiological sign data corresponding to the last exercise of the student; and The comparison calculation module is configured to input the first feature and the second feature into an FNN network for comparison calculation, and a two-dimensional vector output by the FNN network represents a probability of whether the student is practicing.

5. The student exercise quality monitoring apparatus of claim 4, wherein, The device further comprises a model verification unit configured to divide the obtained original exercise data into a verification set, and verify the trained student exercise quality monitoring model based on the verification set until a preset accuracy is met.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor, when executing the computer program, implements the steps of the method of any one of claims 1 to 3.

Images