Abnormal alert methods, devices, electronic equipment and storage media

Abstract

This application provides an anomaly alert method, device, electronic device, and storage medium. The method is applied to the server of an online examination system. After acquiring a preset correlation threshold, a real-time video of the target examinee's answer page, and a real-time monitoring video sent by monitoring equipment, the method determines the corresponding simulated answer action standard based on the real-time answer page video and the target examinee's real-time behavior based on the real-time monitoring video. Then, when the correlation between the real-time behavior and the simulated answer action standard is compared and the correlation does not meet the preset correlation threshold, an anomaly alert is generated and sent to the proctoring terminal in the system. This method comprehensively monitors the examinee's examination process through the answer page and real-time monitoring video, improving the monitoring effect. It also provides timely and effective alerts for anomalies during the examination based on the correlation, allowing proctors to promptly identify abnormal behavior by the examinee through the monitoring footage.

Description

Technical Field

[0001] This application relates to the field of online examination technology, and in particular to an anomaly alert method, device, electronic device, and storage medium. Background Technology

[0002] With the popularization of online examination systems, various new cheating methods have emerged. In order to effectively monitor the examination process, it is necessary to take measures.

[0003] Current technologies typically employ camera-based monitoring for exam proctoring, such as using computer webcams to capture facial images of examinees or using exam room cameras and mobile phone cameras to capture images of the examinees' surroundings. However, these methods primarily monitor examinees' actions and behaviors in the exam room, making the criteria for judging abnormal behavior somewhat one-sided and prone to misjudgment. Furthermore, these methods place a heavy workload on invigilators and cannot promptly detect abnormal behavior.

[0004] Therefore, the existing technology has the technical problem of being too one-sided in monitoring, and invigilators cannot promptly identify abnormal behavior of candidates from the monitoring footage, which needs to be improved. Summary of the Invention

[0005] This application provides an abnormality alert method, device, electronic device, and storage medium to solve the technical problem that existing monitoring is too one-sided, and invigilators cannot promptly identify abnormal behavior of examinees from the monitoring screen.

[0006] To address the aforementioned technical problems, this application provides the following technical solution:

[0007] This application provides an anomaly alert method applied to a server of an online examination system, the online examination system including monitoring equipment, proctoring terminals, examination terminals, and the server; the method includes:

[0008] The system acquires a preset correlation threshold, a real-time video of the target candidate's answer page, and a real-time monitoring video of the target candidate sent by the monitoring device.

[0009] Based on the real-time answer page video, determine the corresponding simulated answer action standards;

[0010] Based on the real-time monitoring video, determine the real-time behavior and actions of the target examinee;

[0011] The real-time behavioral actions are compared with the simulated answering action standards to obtain the correlation degree.

[0012] When the correlation degree does not meet the preset correlation degree threshold, an abnormal reminder message is generated and sent to the proctoring terminal.

[0013] Accordingly, this application also provides an anomaly alert device applied to the server of an online examination system, the online examination system including monitoring equipment, invigilation terminals, examination terminals, and the server; the device includes:

[0014] The first acquisition module is used to acquire a preset correlation threshold, a real-time answer page video of the target candidate, and a real-time monitoring video of the target candidate sent by the monitoring device;

[0015] The standard action determination module is used to determine the corresponding simulated answer action standard based on the real-time answer page video;

[0016] The real-time action determination module is used to determine the real-time behavior of the target examinee based on the real-time monitoring video.

[0017] The correlation comparison module is used to compare the real-time behavior with the simulated answering action standard to obtain the correlation degree.

[0018] The information generation module is used to generate an abnormal reminder message when the correlation degree does not meet the preset correlation degree threshold, and send the abnormal reminder message to the proctoring terminal.

[0019] Meanwhile, this application provides an electronic device including a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to run the computer program in the memory to perform the steps in the above-mentioned abnormality alert method.

[0020] In addition, this application also provides a computer-readable storage medium storing a plurality of instructions adapted for loading by a processor to execute the steps in the above-described exception notification method.

[0021] Beneficial Effects: This application provides an anomaly alert method, apparatus, electronic device, and computer-readable storage medium. Specifically, the method is applied to the server of an online examination system, which includes monitoring equipment, a proctoring terminal, an examination terminal, and the server. The server first needs to obtain a preset correlation threshold, a real-time answer page video of the target examinee, and a real-time monitoring video of the target examinee sent by the monitoring equipment. Then, it determines the corresponding simulated answer action standard based on the answer page video and determines the real-time behavior of the target examinee based on the real-time monitoring video. The real-time behavior is then compared with the simulated answer action standard to obtain a correlation degree. Finally, when the correlation degree does not meet the preset correlation threshold, an anomaly alert message is generated and sent to the monitoring terminal. Compared to existing technologies that only monitor real-time video of examinee behavior, this method combines real-time video of the examinee's answer page with real-time video of the examination to comprehensively monitor the examinee's examination process, improving the monitoring effect. Based on the comparison of the corresponding behaviors in the real-time answer page video and real-time video of the examination, a correlation degree is obtained. Based on the correlation degree, abnormalities in the examination are promptly and effectively alerted, so that invigilators can identify abnormal behavior of examinees in a timely manner through the monitoring screen, and also reduce the workload of invigilators. Attached Figure Description

[0022] The technical solution and other beneficial effects of this application will become apparent from the following detailed description of specific embodiments in conjunction with the accompanying drawings.

[0023] Figure 1 This is a system architecture diagram of the online examination system provided in the embodiments of this application.

[0024] Figure 2 This is a flowchart illustrating the exception notification method provided in the embodiments of this application.

[0025] Figure 3 This is a schematic diagram of a standard for real-time behavioral actions and simulated answering actions provided in an embodiment of this application.

[0026] Figure 4 This is a schematic diagram of the structure of the abnormality alert device provided in the embodiments of this application.

[0027] Figure 5 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0028] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0029] The terms "comprising" and "having" and any variations thereof in the specification and claims of this application are intended to cover non-exclusive inclusion; the division of modules appearing in this application is merely a logical division, and in actual application, there may be other division methods, such as multiple modules being combined into or integrated into another system, or some features being ignored or not executed.

[0030] This application provides an anomaly alert method, device, electronic device, and storage medium. The anomaly alert device can be integrated into an electronic device, which can be a server or a terminal, etc.

[0031] Please see Figure 1 , Figure 1 This is a system architecture diagram of the online examination system provided in the embodiments of this application, such as... Figure 1 As shown, this scenario may include terminals and devices. Terminals, devices, and terminals and devices are connected and communicate via the Internet composed of various gateways. The online examination system includes at least a server 101, a monitoring terminal 102, an examination terminal 103, and a monitoring device 104.

[0032] Server 101 can be a standalone server, a server network, or a server cluster. For example, the server described in this application includes, but is not limited to, computers, network hosts, database servers, application servers, or cloud servers composed of multiple servers, wherein the cloud server is composed of a large number of computers or network servers based on cloud computing. Server 101 is mainly used to store and process data sent by examination terminal 103 and monitoring equipment 104, as well as to send reminder information and other data to monitoring terminal 104.

[0033] The monitoring terminal 102 refers to the terminal operated by the invigilator. This terminal can be a laptop, computer, etc. It is used to display to the invigilator the real-time answer page video from the server 101 and the real-time monitoring video sent by the monitoring device 104, as well as display abnormal reminder information.

[0034] Examination terminal 103 refers to the terminal operated by the examinee, which can be a laptop, computer, etc. It should be noted that, depending on the needs of the examination scenario, examination terminal 103 can be equipped with peripheral accessories such as a keyboard, mouse, and electronic drawing board.

[0035] The monitoring equipment 104 includes cameras installed in the examination room, cameras built into the examination terminals, and cameras built into mobile phones. It should be noted that the real-time monitoring video of the target examinee collected by the monitoring equipment 104 includes both overall video and close-up video of the target examinee.

[0036] A communication link is provided between server 101, monitoring terminal 102, examination terminal 103, and monitoring equipment 104 to enable information exchange; the type of communication link may include wired, wireless communication links, or fiber optic cables, etc., which are not limited herein.

[0037] The examination terminal 103 acquires images of the answer page according to a preset frame rate and preset bit rate, and sends them to the server 101. The server synthesizes the answer page images according to a preset encapsulation format to obtain a real-time answer page video. The monitoring device 104 sends the real-time monitoring of the target examinee it acquires to the server 101. The server 101 determines the corresponding simulated answer action standard based on the real-time answer page video, and at the same time determines the real-time behavior of the target examinee based on the real-time monitoring video. Then, the server 101 compares the real-time behavior with the simulated answer action standard to obtain the correlation degree. Finally, when the correlation degree does not meet the preset correlation degree threshold, the server 101 generates an abnormal reminder message and sends the abnormal reminder message to the proctoring terminal 102 to remind the proctor to check the monitoring image of the target examinee in a timely manner based on the abnormal reminder message, and further determine whether the examinee has abnormal behavior.

[0038] In the above process, this method combines the real-time answer page video and real-time monitoring video of the target candidate to comprehensively monitor the candidate's examination process, thereby improving the monitoring effect. Based on the comparison of the corresponding behaviors in the real-time answer page video and real-time monitoring video, the correlation degree is obtained. Based on the correlation degree, abnormalities in the examination are promptly and effectively alerted, so that invigilators can identify abnormal behaviors of candidates in a timely manner through the monitoring screen, thereby reducing the workload of invigilators.

[0039] It should be noted that, Figure 1The system architecture diagram shown is merely an example. The terminals, devices, and scenarios described in this application are for the purpose of more clearly illustrating the technical solutions of this application and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the evolution of systems and the emergence of new business scenarios, the technical solutions provided in this application are also applicable to similar technical problems. Detailed descriptions are provided below. It should be noted that the order of description of the following embodiments is not intended to limit the preferred order of the embodiments.

[0040] In the embodiments of this application, please refer to Figure 2 As shown, Figure 2 This is a flowchart illustrating an anomaly alert method provided in an embodiment of this application. This anomaly alert method is applied to the server of an online examination system, which includes monitoring equipment, proctoring terminals, examination terminals, and the server. The method includes at least the following steps:

[0041] S201: Obtain the preset correlation threshold, the real-time answer page video of the target candidate, and the real-time monitoring video of the target candidate sent by the monitoring equipment.

[0042] The preset correlation threshold refers to the range of correlation between actual actions and simulated answering actions when the latter are performed normally according to the standard. Monitoring equipment may include wide-range video cameras to overcome distortion issues. It should be noted that the timeline of the real-time answering page video and the timeline of the real-time monitoring video are in one-to-one correspondence.

[0043] In one embodiment, S201 includes: acquiring answer page images at various times sent by the examination terminal, wherein the answer page images are collected by the examination terminal according to a preset frame rate and a preset bit rate; and synthesizing a real-time answer page video of the target examinee according to a preset encapsulation format and the answer page images at various times. Here, the answer page image refers to the page image of the examination terminal operated by the examinee. The answer page image can display specific examination questions. In addition, the answer page image can also display the examinee's answer record and answering process, for example, for the multiple-choice question answer area, after the examinee selects the corresponding option, the page will also display the corresponding option in real time by checking the corresponding option. That is, whatever operation the examinee performs, the page of the corresponding examination terminal will respond in real time and display the corresponding operation. It should be noted that this is an example of multiple-choice questions; the answer page image can also be a text representation of short-answer questions, or... This includes the calculation process for calculation problems, etc.; bitrate, also known as bit depth, refers to the number of bits transmitted or decoded per unit time. The higher the bitrate, the higher the quality and clarity of the resulting composite video; frame rate refers to how many frames are moved per second during playback. The higher the frame rate, the smoother the composite video; the preset container format can be FLV format, etc. Using FLV format for video composite can effectively save resource space and significantly improve video loading speed, effectively avoiding situations such as excessively long response times or stuttering, thereby enabling real-time monitoring of the target examinee's examination terminal operation.

[0044] Unlike webcam recording of computer desktop video, this embodiment synthesizes real-time answer page video from images of the answer page sent from the examination terminal to the server. Specifically, the examination terminal captures images of its own page at a preset frame rate and bitrate, obtaining answer page images at various times. These images are then sent to the server's temporary storage space for temporary storage. The server processes and synthesizes these answer page images according to a preset synthesis rule to obtain the real-time answer page video of the target examinee. It should be noted that this embodiment does not synthesize all the answer page images collected during the entire examination process at the end of the examination. Instead, it periodically synthesizes the collected answer page images and stores the synthesized video in the temporary storage space. After one segment of video is synthesized and stored, the next segment's image acquisition and subsequent operations begin, while previously obtained videos are retained, ensuring a continuous and complete video. Furthermore, the answer page images stored in the temporary storage space are cleared after the server synthesizes the real-time answer page video, thus saving storage resources.

[0045] Optionally, before taking the online exam, it is necessary to obtain the candidate's identity information (which may include the admission ticket number, ID card number, fingerprint information, image information, and voice information, etc.) to complete the candidate's identity authentication. Only after the identity authentication is passed can the candidate be allowed to enter their respective exam stage through the exam terminal.

[0046] Furthermore, candidate identity verification can be achieved by comparing the image and voice information collected on-site by monitoring equipment (such as the front-facing camera built into the examination terminal) with historical data in a preset candidate identity information database.

[0047] S202: Determine the corresponding simulated answering action standards based on the real-time answer page video.

[0048] In one embodiment, S202 includes: determining the real-time answering process based on the real-time answering page video; and determining the corresponding simulated answering action standard based on the answering action simulation model and the real-time answering process. The answering action simulation model is trained on a training set formed by the examinee's answering actions and answering process.

[0049] As mentioned earlier, the real-time answer page video is synthesized from images of the answer page at various times. These images display the examinee's answer records and answering process. Therefore, based on the real-time answer page video, the real-time answering process of the target examinee can be determined. Examples include checking an option in a multiple-choice question, editing text in the answer area of ​​a word question, or drawing in the answer area of ​​a drawing question. This real-time answering process is used as model input into the answering action simulation model to obtain the corresponding simulated answering action standard. For example, checking an option in a multiple-choice question corresponds to the simulated answering action standard of the examinee being in front of the examination terminal, with their eyes focused on the page displaying the multiple-choice question, and their hand movement being clicking the mouse. Similarly, editing text in the answer area of ​​a word question corresponds to the simulated answering action standard of the examinee being in front of the examination terminal, with their eyes focused on the question stem and answer area displayed on the examination terminal, and their hand movement being typing on the keyboard.

[0050] S203: Determine the real-time behavior and actions of the target examinee based on real-time monitoring video.

[0051] Among them, real-time actions refer to the actions of candidates at the current moment, which change irregularly over time.

[0052] In one embodiment, S203 includes: correcting the real-time monitoring video using a trained image correction model to obtain a corrected real-time monitoring video; wherein the image correction model is obtained by knowledge distillation of the student image correction model based on the teacher image correction model, the teacher image correction model is trained based on an image sample set constructed from the intrinsic parameter data and distortion coefficients of the monitoring device, and the student image correction model is constructed based on the teacher image correction model; and identifying human images in the corrected real-time monitoring video using a trained behavior recognition model to obtain the real-time behavior of the target examinee.

[0053] Because surveillance equipment inevitably produces distortions during video capture, to minimize the impact of these distortions on subsequent processes, the real-time surveillance video must first be corrected using a trained image correction model. This corrected video is then used as input to a trained behavior recognition model to obtain the real-time actions of the target examinee. The training process of the image correction model includes: acquiring a trained generative adversarial network (GAN); this GAN includes an image discrimination model and a teacher image correction model; constructing a student image correction model based on the teacher image correction model; and performing knowledge distillation on the student image correction model based on the teacher image correction model to obtain the final image correction model. The GAN is pre-trained using image samples and is used to correct distorted images. The discriminator in the pre-trained GAN is the image discrimination model, and the generator is the teacher image correction model, which can be used to correct distorted images. However, due to the large number of parameters in the teacher image correction model, it is impossible to perform independent and real-time correction operations on terminals such as mobile phones and USB cameras. Therefore, after obtaining a pre-trained generative adversarial network, the teacher image correction model in the generative adversarial network is used as the teacher model to construct its corresponding student model, thus obtaining the student image correction model.

[0054] Furthermore, the training process of the aforementioned generative adversarial network includes: constructing an image sample set based on the intrinsic parameter data and distortion coefficients of the monitoring device; and training the generative adversarial network using this image sample set to obtain the trained generative adversarial network. The image sample set is the image dataset used to train the generative adversarial network, and it consists of distorted images and undistorted images. The distorted images are the undistorted images after distortion. The distorted images can be obtained by processing the undistorted images based on the intrinsic parameter data and distortion coefficients of the video acquisition device.

[0055] Furthermore, in the process of constructing a student image correction model based on the teacher image correction model, the teacher image correction model can be pruned and parameters shared to construct the student image correction model. The teacher image correction model can then be used to assist the student image correction model in training, thereby transferring the knowledge learned by the teacher image correction model to the student image correction model. This results in the distilled student image correction model (i.e., the image correction model in this application) having the same image correction function as the teacher image correction model.

[0056] It should be noted that, in addition to correcting real-time monitoring video, the server can also perform 3D fusion processing on it.

[0057] S204: Compare real-time behavioral actions with simulated answering action standards to obtain the correlation degree.

[0058] The simulated answering action standards determined by the real-time answering page video are consistent with the real-time behavioral actions determined by the real-time monitoring video at the same time points.

[0059] In one embodiment, S204 may include: determining the actual eye focus range of the target examinee based on real-time behavioral actions; determining the simulated eye focus range of the target examinee based on simulated answering action standards; and comparing the actual eye focus range with the simulated eye focus range to obtain a correlation degree. The real-time behavioral actions include various actions of the examinee during the examination, such as eye focus range, head movements, hand movements, and overall movements.

[0060] Specifically, by analyzing the candidate's real-time behavior and actions, the actual eye focus range of the target candidate can be determined (e.g., Figure 3 Taking the fan-shaped area marked as the actual eye focus range as an example), and based on the simulated answering action standards determined earlier, the simulated eye focus range of the target candidate is determined (e.g., ...). Figure 3 Taking the fan-shaped area marked as the simulated eye focus range as an example, the two are compared to obtain the correlation range between the actual eye focus range and the simulated eye focus range. Furthermore, based on the correlation range and the simulated eye focus range, the correlation degree between the actual behavioral action and the simulated answering action standard can be calculated. In simpler terms, by comparing the range that the candidate's eyes should focus on during the answering process shown in the video on the answer page with the candidate's actual eye focus range, the size of the range that the candidate is currently focusing on can be determined.

[0061] In one embodiment, S204 may include: determining the actual hand movements of the target examinee based on real-time behavioral actions; determining the simulated hand movements of the target examinee based on simulated answering action standards; and comparing the actual hand movements with the simulated hand movements to obtain a correlation degree. The hand movements include using a mouse, typing on a keyboard, drawing images on a drawing board, touching the face, turning pages, etc.

[0062] Specifically, by analyzing the candidate's real-time actions, the actual hand movements of the target candidate (e.g., touching their face) can be determined. Simultaneously, based on the simulated answering action standards established earlier, the simulated hand movements of the target candidate (typing on a keyboard) can be determined. Finally, the two are compared to obtain the similarity between the actual and simulated hand movements; this similarity is the correlation coefficient. In simpler terms, the correlation coefficient is determined by comparing the similarity between the simulated and actual hand movements of the candidate during the answering process shown in the video on the answer page.

[0063] In one embodiment, S204 may further include: determining the actual location of the target examinee based on real-time monitoring video; determining the simulated location of the target examinee based on real-time answer page video; and comparing the actual location with the simulated location to obtain a correlation degree. Here, location refers to the examinee's relative position in the current examination room.

[0064] Specifically, by analyzing real-time monitoring video, the actual position of the target candidate at a certain moment can be determined (e.g., diagonally behind the examination terminal). At the same time, based on the real-time answer page video, the simulated position of the target candidate at the same moment can be determined (e.g., within a certain range directly in front of the examination terminal display page). Finally, the two positions are compared to obtain the correlation between the actual position and the simulated position.

[0065] S205: When the correlation degree does not meet the preset correlation degree threshold, generate an abnormal reminder message and send the abnormal reminder message to the proctoring terminal.

[0066] The abnormal alert information may include images of abnormal behavior of the examinee (images of the answer page in the real-time answer page video and images in the real-time monitoring video at the same time), the time point of the abnormal behavior, the duration of the abnormal behavior, and the degree of correlation between the two.

[0067] As mentioned above, the preset correlation threshold refers to the range of correlation between actual behavior and the simulated answering action standard when the behavior is normal. Therefore, if the correlation determined in step S204 does not meet the preset correlation threshold, it indicates that the target candidate's actual behavior is abnormal compared to the simulated answering action standard. In this case, the server will send an abnormality alert to the proctoring terminal to remind the proctoring staff to check the monitoring screen of the target candidate in a timely manner to further determine whether there is indeed abnormal behavior.

[0068] For example, based on the aforementioned steps, it is determined that text is being edited at time A, corresponding to the simulated hand action of typing on a keyboard. However, the actual hand action determined from the real-time monitoring video at time A is that the examinee is touching their face. The resulting correlation is 0, which does not meet the preset correlation threshold (60%-100%). Therefore, it can be determined that the target examinee exhibited abnormal behavior at time A. In other words, at time A, the answer page indicates that the examinee should be typing on a keyboard, but the actual monitoring video shows the examinee touching their face, indicating that the two actions are out of sync. The typing action was not triggered by the examinee themselves, thus constituting an anomaly.

[0069] Furthermore, after the server sends the abnormality alert information to the proctoring terminal, the proctoring terminal can display the examinee's examination image on the proctoring page, including images of abnormal behavior (images of the answer page in the real-time answer page video and images in the real-time monitoring video at the same time). The abnormal behavior can also be marked in the image and accompanied by text prompts, such as "At time A, the actual hand movements and the answer process are inconsistent, lasting for 30 seconds" or "At time A, the state is abnormal, lasting for 30 seconds," etc. After seeing the prompt information on the proctoring page, the proctor can click on the target examinee's real-time answer page video and real-time monitoring video to view them.

[0070] Optionally, during the entire examination, in the absence of any abnormal alerts, invigilators may proactively play relevant monitoring videos of any examinee (real-time answer page video and real-time monitoring video) for random checks; after the examination, invigilators may also randomly retrieve monitoring videos of some examinees for review, thereby improving the flexibility and accuracy of the invigilation process.

[0071] In addition to displaying real-time monitoring footage of each examinee and anomaly alerts on the proctoring terminal's monitoring page, the system also displays various exam-related information. Furthermore, proctors and administrators can perform exam configuration and other related operations. For example, the settings and configuration options on the proctoring page allow personnel to complete pre-exam configurations such as scheduling, photo verification, and subject allocation; in-exam monitoring options include accessing exam monitoring footage and examinee answer details at any time; and post-exam tasks such as grading. Thus, a unified proctoring system enables the completion of pre-exam, exam, and post-exam tasks, facilitating remote, home-based proctoring and grading work for proctors.

[0072] As can be seen from the above embodiments, compared with the existing technology that only monitors the behavior of examinees in real-time video, the abnormality reminder method of this application combines the real-time answer page video of the target examinee and the real-time monitoring video to comprehensively monitor the examinee's examination process, thereby improving the monitoring effect. Based on the comparison of the corresponding behaviors in the real-time answer page video and the real-time monitoring video, a correlation degree is obtained. Based on the correlation degree, abnormalities in the examination are promptly and effectively reminded, so that invigilators can promptly identify abnormal behaviors of examinees through the monitoring screen, and also reduce the workload of invigilators.

[0073] Based on the above embodiments, this application provides an anomaly alert device. This device is applied to the server of an online examination system, which includes monitoring equipment, proctoring terminals, examination terminals, and a server. For details, please refer to... Figure 4 The device includes:

[0074] The first acquisition module 401 is used to acquire a preset correlation threshold, the real-time answer page video of the target candidate, and the real-time monitoring video of the target candidate sent by the monitoring device.

[0075] The standard action determination module 402 is used to determine the corresponding simulated answer action standard based on the real-time answer page video;

[0076] The real-time action determination module 403 is used to determine the real-time behavior of the target examinee based on the real-time monitoring video.

[0077] The correlation comparison module 404 is used to compare the real-time behavior action with the simulated answer action standard to obtain the correlation degree.

[0078] The information generation module 405 is used to generate an abnormal reminder message when the comparison correlation does not meet the preset correlation threshold, and send the abnormal reminder message to the proctoring terminal.

[0079] In one embodiment, the first acquisition module 401 includes:

[0080] The second acquisition module is used to acquire the answer page images at various times sent by the examination terminal, wherein the answer page images are collected by the examination terminal according to a preset frame rate and a preset bit rate;

[0081] The video synthesis module is used to synthesize the real-time answer page video of the target examinee based on the preset encapsulation format and the answer page images at each time point.

[0082] In one embodiment, the standard action determination module 402 includes:

[0083] The process determination module is used to determine the real-time answering process based on the video of the real-time answering page;

[0084] The first determining module is used to determine the corresponding simulated answering action standard based on the answering action simulation model and the real-time answering process.

[0085] In one embodiment, the real-time motion determination module 403 includes:

[0086] The video correction module is used to correct the real-time monitoring video using a trained image correction model to obtain a corrected real-time monitoring video. The image correction model is obtained by knowledge distillation of the student image correction model based on the teacher image correction model. The teacher image correction model is trained on an image sample set constructed based on the intrinsic parameter data and distortion coefficients of the monitoring device. The student image correction model is constructed based on the teacher image correction model.

[0087] The second determining module is used to identify human images in the corrected real-time monitoring video through the trained behavior recognition model, and to obtain the real-time behavior of the target examinee.

[0088] In one embodiment, the association comparison module 404 includes:

[0089] The first range determination module is used to determine the actual eye focus range of the target examinee based on the real-time behavioral actions.

[0090] The second range determination module is used to determine the simulated eye focus range of the target examinee based on the simulated answering action standards.

[0091] The first comparison module is used to compare the actual eyeball focusing range with the simulated eyeball focusing range to obtain the comparison correlation degree.

[0092] In one embodiment, the association comparison module 404 includes:

[0093] The first action determination module is used to determine the actual hand movements of the target examinee based on the real-time behavioral actions.

[0094] The second action determination module is used to determine the simulated hand movements of the target examinee based on the simulated answering action standards.

[0095] The second comparison module is used to compare the actual hand movements with the simulated hand movements to obtain the comparison correlation.

[0096] In one embodiment, the association comparison module 404 includes:

[0097] The first location determination module is used to determine the actual location of the target examinee based on the real-time monitoring video.

[0098] The second location determination module is used to determine the simulated location of the target candidate based on the real-time answer page video;

[0099] The third comparison module is used to compare the actual position with the simulated position to obtain the comparison correlation degree.

[0100] Unlike current technologies, the anomaly alert device provided in this application includes a first acquisition module, a correlation comparison module, and an information generation module. The first acquisition module acquires the real-time answer page video and real-time monitoring video of the target examinee, thereby comprehensively monitoring the examinee's examination process and improving the monitoring effect. The correlation comparison module obtains the correlation degree based on the comparison of the corresponding behaviors in the real-time answer page video and the real-time monitoring video. Furthermore, the information generation module provides timely and effective alerts for anomalies during the examination based on the correlation degree, enabling invigilators to promptly identify abnormal behaviors of examinees through the monitoring screen and reducing the workload of invigilators.

[0101] Accordingly, embodiments of this application also provide an electronic device, such as... Figure 5 As shown, the electronic device may include a processor 501 with one or more processing cores, a wireless (WiFi, Wireless Fidelity) module 502, a memory 503 with one or more computer-readable storage media, an audio circuit 504, a display unit 505, an input unit 506, a sensor 507, a power supply 508, and a radio frequency (RF) circuit 509, etc. Those skilled in the art will understand that... Figure 5 The structure of the electronic device shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0102] The processor 501 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs and / or modules stored in the memory 503, and by calling data stored in the memory 503, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. In one embodiment, the processor 501 may include one or more processing cores; preferably, the processor 501 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 501.

[0103] WiFi is a short-range wireless transmission technology. Electronic devices, through the wireless module 502, can help users send and receive emails, browse web pages, and access streaming media, providing users with wireless broadband internet access. Although Figure 5 The wireless module 502 is shown, but it is understood that it is not an essential component of the terminal and can be omitted as needed without changing the nature of the invention.

[0104] The memory 503 can be used to store software programs and modules. The processor 501 executes various functional applications and data processing by running the computer programs and modules stored in the memory 503. The memory 503 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the terminal (such as audio data, phone book, etc.). In addition, the memory 503 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 503 may also include a memory controller to provide access to the memory 503 for the processor 501 and the input unit 506.

[0105] The audio circuit 504 includes a speaker, which provides an audio interface between the user and the electronic device. The audio circuit 504 can transmit the electrical signal converted from received audio data to the speaker, where the speaker converts it into a sound signal for output.

[0106] Display unit 505 can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the terminal. These graphical user interfaces can be composed of graphics, text, icons, video, and any combination thereof. Display unit 505 may include a display panel, which in one embodiment can be configured as a liquid crystal display (LCD), organic light-emitting diode (OLED), or similar form. Furthermore, a touch-sensitive surface can cover the display panel. When the touch-sensitive surface detects a touch operation on or near it, it transmits the information to processor 501 to determine the type of touch event. Subsequently, processor 501 provides corresponding visual output on the display panel according to the type of touch event. Although in Figure 5 In this context, the touch-sensitive surface and the display panel are two separate components for implementing input and output functions. However, in some embodiments, the touch-sensitive surface and the display panel can be integrated to achieve input and output functions.

[0107] Input unit 506 can be used to receive input numerical or character information, and to generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function control. Specifically, in one embodiment, input unit 506 may include a touch-sensitive surface and other input devices. A touch-sensitive surface, also known as a touch display or touchpad, can collect user touch operations on or near it (e.g., user operations using fingers, styluses, or any suitable object or accessory on or near the touch-sensitive surface) and drive corresponding connection devices according to a pre-set program. In one embodiment, the touch-sensitive surface may include a touch detection device and a touch controller. The touch detection device detects the user's touch orientation and the signal generated by the touch operation, transmitting the signal to the touch controller; the touch controller receives touch information from the touch detection device, converts it into touch point coordinates, sends it to the processor 501, and can receive and execute commands from the processor 501. Furthermore, various types of touch-sensitive surfaces, such as resistive, capacitive, infrared, and surface acoustic wave, can be used. In addition to the touch-sensitive surface, input unit 506 may also include other input devices. Specifically, other input devices may include, but are not limited to, one or more of the following: physical keyboard, function keys (such as volume control buttons, power buttons, etc.), trackball, mouse, joystick, etc.

[0108] The electronic device may also include at least one sensor 507, such as a light sensor. Specifically, the light sensor may include an ambient light sensor and a distance sensor. The ambient light sensor can adjust the brightness of the display panel according to the brightness of the ambient light. Other sensors that may be configured in the electronic device, such as gyroscopes, barometers, hygrometers, thermometers, and infrared sensors, will not be described in detail here.

[0109] The electronic device also includes a power supply 508 (such as a battery) that supplies power to various components. Preferably, the power supply can be logically connected to the processor 501 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 508 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0110] The radio frequency (RF) circuit 509 can be used for receiving and transmitting signals during information transmission or calls. Specifically, it receives downlink information from the base station and hands it over to one or more processors 501 for processing; additionally, it transmits uplink data to the base station. Typically, the RF circuit 509 includes, but is not limited to, an antenna, at least one amplifier, a tuner, one or more oscillators, a Subscriber Identity Module (SIM) card, a transceiver, a coupler, a low-noise amplifier (LNA), a duplexer, etc. Furthermore, the RF circuit 509 can also communicate wirelessly with networks and other devices. Wireless communication can use any communication standard or protocol, including but not limited to GSM, GPRS, CDMA, WCDMA, LTE, email, and SMS.

[0111] Although not shown, the electronic device may also include a camera, Bluetooth module, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 501 in the electronic device loads the executable files corresponding to the processes of one or more applications into the memory 503 according to the following instructions, and the processor 501 runs the applications stored in the memory 503 to achieve the following functions:

[0112] The system acquires a preset correlation threshold, a real-time video of the target candidate's answer page, and a real-time monitoring video of the target candidate sent by the monitoring device.

[0113] Based on the real-time answer page video, determine the corresponding simulated answer action standards;

[0114] Based on the real-time monitoring video, determine the real-time behavior and actions of the target examinee;

[0115] The real-time behavioral actions are compared with the simulated answering action standards to obtain the correlation degree.

[0116] When the correlation degree does not meet the preset correlation degree threshold, an abnormal reminder message is generated and sent to the proctoring terminal.

[0117] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0118] Therefore, embodiments of this application provide a computer-readable storage medium storing multiple instructions that can be loaded by a processor to implement the functions of the above-described exception alert method.

[0119] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0120] The above provides a detailed description of the anomaly alert method, apparatus, electronic device, and computer-readable storage medium provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. 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 application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An anomaly alert method, characterized in that, A server used in an online examination system, the online examination system including monitoring equipment, proctoring terminals, examination terminals, and the server; the method includes: The system acquires a preset correlation threshold, a real-time video of the target candidate's answer page, and a real-time monitoring video of the target candidate sent by the monitoring device. Based on the real-time answer page video, determine the corresponding simulated answer action standards; Based on the real-time monitoring video, determine the real-time behavior and actions of the target examinee; The real-time behavioral actions are compared with the simulated answering action standards to obtain the correlation degree. When the correlation degree does not meet the preset correlation degree threshold, an abnormal reminder message is generated and sent to the proctoring terminal; The step of determining the corresponding simulated answering action standard based on the real-time answering page video includes: Based on the video of the real-time answer page, the real-time answering process is determined; the real-time answering process includes checking the options of a multiple-choice question, editing text in the answer area of ​​a text question, or drawing in the answer area of ​​a drawing question; Based on the answering action simulation model and the real-time answering process, the corresponding simulated answering action standards are determined; the simulated answering action standards include the target candidate being in front of the examination terminal, with their eyes focused on the multiple-choice question page displayed on the examination terminal, and their hand action being clicking the mouse, or the target candidate being in front of the examination terminal, with their eyes focused on the question stem and answer area of ​​the text question displayed on the examination terminal, and their hand action being typing on the keyboard.

2. The anomaly alert method according to claim 1, characterized in that, The step of obtaining the preset correlation threshold, the real-time answer page video of the target examinee, and the real-time monitoring video of the target examinee sent by the monitoring device includes: The examination terminal sends answer page images at various times, wherein the answer page images are collected by the examination terminal according to a preset frame rate and a preset bit rate. Based on the preset encapsulation format and the answer page images at each time point, a real-time answer page video of the target examinee is synthesized.

3. The anomaly alert method according to claim 1, characterized in that, The step of determining the real-time behavior of the target examinee based on the real-time monitoring video includes: The real-time monitoring video is corrected by the trained image correction model to obtain the corrected real-time monitoring video; wherein, the image correction model is obtained by knowledge distillation of the student image correction model based on the teacher image correction model, the teacher image correction model is trained on the image sample set constructed based on the intrinsic parameter data and distortion coefficients of the monitoring device, and the student image correction model is constructed based on the teacher image correction model; The trained behavior recognition model identifies human images in the corrected real-time monitoring video to obtain the real-time behavior of the target examinee.

4. The anomaly alert method according to claim 1, characterized in that, The step of comparing the real-time behavioral actions with the simulated answering action standards to obtain the correlation degree includes: Based on the real-time behavioral actions, determine the actual eye focus range of the target examinee; Based on the simulated answering action standards, determine the simulated eye focus range of the target candidate; The actual eyeball focusing range is compared with the simulated eyeball focusing range to obtain the correlation degree.

5. The anomaly alert method according to claim 1, characterized in that, The step of comparing the real-time behavioral actions with the simulated answering action standards to obtain the correlation degree includes: Based on the real-time behavioral actions, determine the actual hand movements of the target examinee; Based on the simulated answering action standards, determine the simulated hand movements of the target examinee; The actual hand movements are compared with the simulated hand movements to obtain the correlation degree.

6. The anomaly alert method according to claim 1, characterized in that, Also includes: The actual location of the target examinee is determined based on the real-time monitoring video. Based on the real-time answer page video, determine the simulated location of the target candidate; The actual location is compared with the simulated location to obtain the correlation degree.

7. An anomaly alert device, characterized in that, A server for use in an online examination system, the online examination system including monitoring equipment, proctoring terminals, examination terminals, and the server; the device includes: The first acquisition module is used to acquire a preset correlation threshold, a real-time answer page video of the target candidate, and a real-time monitoring video of the target candidate sent by the monitoring device. The standard action determination module is used to determine the corresponding simulated answer action standard based on the real-time answer page video; The real-time action determination module is used to determine the real-time behavior of the target examinee based on the real-time monitoring video. The correlation comparison module is used to compare the real-time behavior with the simulated answering action standard to obtain the correlation degree. The information generation module is used to generate an abnormal reminder message when the correlation degree does not meet the preset correlation degree threshold, and send the abnormal reminder message to the proctoring terminal. The standard action determination module is further used to determine the real-time answering process based on the real-time answer page video. The real-time answering process includes checking the options of a multiple-choice question, editing text in the answer area of ​​a text question, or drawing in the answer area of ​​a drawing question. Based on the answering action simulation model and the real-time answering process, the corresponding simulated answering action standard is determined. The simulated answering action standard includes the target candidate being in front of the examination terminal with their eyes focused on the multiple-choice question page displayed on the examination terminal and their hand action being clicking the mouse, or the target candidate being in front of the examination terminal with their eyes focused on the question stem and answer area of ​​the text question displayed on the examination terminal and their hand action being typing on the keyboard.

8. An electronic device, characterized in that, It includes a processor and a memory, the memory being used to store a computer program, and the processor being used to run the computer program in the memory to perform the steps in the abnormality alert method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions adapted for loading by a processor to perform the steps of the exception notification method according to any one of claims 1 to 6.

Images