KR20220142681A - System and method for monitoring cheating in online test - Google Patents

Abstract

An online test cheating monitoring system according to an embodiment of the present invention comprises: a cheating monitoring device as a server providing online test questions; and a plurality of user terminals installed with an online test application provided by the cheating monitoring device and connected to the cheating monitoring device through a network.Each of the user terminals has first and second sensors and transmits test environment information of test-takers sensed through the first and second sensors to the cheating monitoring device. The cheating monitoring device comprises: a cheating behavior detection unit for receiving image information of a test taker sensed through the first sensor installed in the user terminals and analyzing the image information to detect cheating behaviors of the test takers; and a cheating voice detection unit for receiving audio information of the test taker sensed through the second sensor installed in the user terminals and analyzing the audio information to detect cheating voices of the test takers. Accordingly, an online test environment secured with reliability and fairness can be provided.

Description

Online exam cheating monitoring system and method {SYSTEM AND METHOD FOR MONITORING CHEATING IN ONLINE TEST}

An embodiment of the present invention relates to a test cheating monitoring system and method, and more particularly, to an online exam cheating monitoring system and method using video and audio information about a user (examinee) taking an exam.

In the conventional case, it is common for exams at school and workplaces, and certain certification exams, to be taken by presenting questions through a paper exam paper in the presence of a supervisor who oversees the exam, and having them write their answers on a paper answer sheet.

However, due to the recent development of mobile terminals and networks, an online test rather than the existing offline test, that is, a method in which an examinee connects to a network through their terminal and performs a predetermined test, is also being used. Such an online test method is expected to be used more frequently due to convenience or remoteness.

However, when the online test method is introduced, due to online characteristics such as mobility and portability, a different form of cheating may occur than the existing offline method, which may reduce the reliability of the test result.

An embodiment of the present invention uses a sensor (eg, camera, microphone, etc.) installed in a user terminal to receive test environment information including video and audio information of a test taker taking an online test and analyze it to suspect that the test taker is cheating An object of the present invention is to provide an online test cheating monitoring system and method capable of providing an online test environment in which reliability and fairness are secured by detecting the behavior that is being taken.

In order to achieve the above object, an online exam cheating monitoring system according to an embodiment of the present invention includes: a cheating monitoring device as a server for providing online exam questions; An online test application provided by the cheating monitoring device is installed, and a plurality of user terminals connected to the cheating monitoring device through a network, each of the user terminals having a first and a second sensor, The test environment information of the test takers sensed through the first and second sensors is transmitted to the cheating monitoring device, and the cheating monitoring device is configured to transmit the test environment information of the examinees sensed through the first sensor installed in the user terminals. a cheating behavior detection unit that receives the image information and analyzes it to detect cheating behaviors of the test takers; and a fraudulent voice detector configured to receive the test taker's audio information sensed through the second sensor installed in the user terminals, and analyze it to detect the test takers' cheating voices.

The first sensor may include a camera as an image sensor, and the second sensor may include a microphone as an audio sensor.

The cheating monitoring device may include: a memory for storing test environment information of the examinees sensed by the first and second sensors; and a database for separately extracting, storing and maintaining the detected cheating behavior and/or cheating voice information of the test takers from among the test environment information of the test takers stored in the memory.

The database may store and maintain unique information of user terminals connected to the cheating monitoring device and the online test questions.

The cheating behavior detection unit extracts a single frame image or a continuous section frame image from the received image information of the examinee, and recognizes the behavior of an object included in the extracted at least one frame image. By using the technique, it is determined whether the behavior of the object is a cheating behavior, and a frame image including a suspected cheating behavior among the extracted frame images can be stored.

The behavior recognition technique defines the types of cheating, collects images for each defined type of cheating, selects the patterns of cheating in advance through action recognizer model learning, and selects the pre-selected By comparing the cheating patterns with the behavior of the object included in the frame image, it can be determined whether the examinee's behavior substantially matches the cheating patterns.

The cheating voice detector may include: a sound generation section detecting section for detecting and storing a section in which a sound is generated from among sequential audio frames including audio information of the examinee sensed by the second sensor; a speaker detection unit that separates voices for each speaker from the stored sound generation section data and stores the separated voice sections for each speaker; and a voice recognition unit configured to compare and analyze each speaker's voice and the examinee's voice to detect the examinee's cheating voice.

The speaker detector may determine whether human voice data is included in the generated sound data in the stored sound generation section data.

When it is determined that the voice data is included, the speaker detection unit may separate the voice data for each speaker and store the separated voice data section for each speaker when a plurality of speakers are recognized through conversation.

The voice recognition unit separates and extracts the voice data for each speaker by using a voice recognition function, converts it into text and stores it, and compares the converted text with the actually submitted online test text so that the converted text is actually It is determined whether or not it substantially matches the test question text, and when it is determined that the converted text is substantially consistent with the actual test question text, the voice data of the suspected cheating section corresponding to the extracted voice data for each speaker and its conversion You can store the text information.

In the online test cheating monitoring method according to another embodiment of the present invention, a cheating monitoring device as a server providing online test questions and an online test application provided by the cheating monitoring device are installed, and the cheating monitoring device In an online test cheating monitoring method including a plurality of user terminals connected to a network, the image information of the examinee sensed through a first sensor installed in the user terminals is received, and the image information is analyzed to analyze the cheating behavior of the examinees. a fraudulent behavior detection step of detecting and a fraudulent voice detection step of receiving the examinee's audio information sensed through the second sensor installed in the user terminals, and analyzing it to detect the test takers' cheating voices.

storing the test environment information of the test takers sensed by the first and second sensors in a memory; and separately extracting the detected cheating behavior and/or cheating voice information of the test takers from among the test environment information of the test takers stored in the memory, and storing and maintaining it in a database.

The detecting of the cheating behavior may include: extracting a single frame image or a continuous section frame image from the received image information of the examinee; determining whether the action of the object is a cheating action by using an action recognition technique with respect to the action of the object included in the extracted at least one frame image; and storing a frame image including a behavior suspected of cheating among the extracted frame images.

The behavior recognition technique defines the types of cheating, collects images for each defined type of cheating, selects the patterns of cheating in advance through action recognizer model learning, and selects the pre-selected By comparing the cheating patterns with the behavior of the object included in the frame image, it can be determined whether the examinee's behavior substantially matches the cheating patterns.

The fraudulent voice detection step may include: a sound generation section detecting step of detecting and storing a section in which a sound is generated from among sequential audio frames including audio information of the examinee sensed through the second sensor; a speaker detection step of separating a voice for each speaker from the stored sound generation section data and storing the separated voice section for each speaker; and comparing and analyzing the voice of each speaker with the voice of the test taker to detect the test taker's cheating voice.

The step of detecting the speaker may include determining whether human voice data is included in the generated sound data in the stored sound generation section data.

In the speaker detection step, when it is determined that the voice data is included, if a plurality of speakers are recognized through conversation, the voice data is separated for each speaker, and the separated voice data section for each speaker is stored. may include

The voice recognition step may include: using a voice recognition function to separate and extract the voice data for each speaker, convert it into text, and store it; comparing the converted text with the actual submitted online test question text to determine whether the converted text substantially matches the actual test question text; and if it is determined that the converted text is substantially consistent with the actual test question text, storing the voice data of the suspected cheating section corresponding to the extracted voice data for each speaker and the converted text information. .

According to this embodiment of the present invention, the test environment information including the video and audio information of the test taker who is taking the online test is received by using a sensor (eg, camera, microphone, etc.) installed in the user terminal, and the test environment information is analyzed and analyzed. It is possible to provide an online test environment in which reliability and fairness are secured by detecting acts suspected of cheating.

1 is a block diagram showing the overall configuration of an online test cheating monitoring system according to an embodiment of the present invention.
FIG. 2 is a block diagram showing the configuration of the user terminal shown in FIG. 1;
3 is a block diagram showing the configuration of a fraud monitoring apparatus according to an embodiment of the present invention.
Fig. 4 is a block diagram showing an embodiment of an internal configuration of the processor shown in Fig. 3;
5 is a diagram schematically illustrating an operation of detecting cheating by an online test cheating monitoring system according to an embodiment of the present invention.
6 is a flowchart illustrating a method of detecting a cheating behavior using the image information shown in FIG. 5 .
7 is a flowchart illustrating a method of detecting a cheating voice using the audio information shown in FIG. 5 .

The contents described in the technical field of the background of the present invention are only for helping the understanding of the background of the technical idea of the present invention, and therefore it can be understood as content corresponding to the prior art known to those skilled in the art of the present invention. none.

In the following description, for purposes of explanation, numerous specific details are set forth to aid in understanding various embodiments. It will be evident, however, that various embodiments may be practiced without these specific details or in one or more equivalent manners. In other instances, well-known structures and devices are shown in block diagram form in order to avoid unnecessarily obscuring the various embodiments.

Each block of the accompanying block diagram may be executed by computer program instructions (execution engine), which may be mounted on a processor of a general-purpose computer, special-purpose computer, or other programmable data processing equipment, so that the computer or The instructions, executed by the processor of the other programmable data processing equipment, will create means for performing the functions described in each block of the block diagram.

These computer program instructions may also be stored in a computer-usable or computer-readable memory that may direct a computer or other programmable data processing equipment to implement a function in a particular manner, and thus the computer-usable or computer-readable memory. It is also possible to produce an article of manufacture containing instruction means for performing the functions described in each block of the block diagram, the instructions stored in the block diagram.

And, since the computer program instructions may be mounted on a computer or other programmable data processing equipment, a series of operating steps are performed on the computer or other programmable data processing equipment to create a computer-executed process to create a computer or other program It is also possible that instructions for performing the possible data processing equipment provide functionality for performing the functions described in each block of the block diagram.

In addition, each block may represent a module, segment, or portion of code that includes one or more executable instructions for executing specified logical functions, and in some alternative embodiments, the functions recited in the blocks or steps are ordered It is also possible to occur outside of

That is, the two illustrated blocks may be substantially simultaneously performed, and also, the blocks may be performed in the reverse order of corresponding functions, if necessary.

The terminology used herein is for the purpose of describing particular embodiments and not for the purpose of limitation. Throughout the specification, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless otherwise stated. Unless otherwise defined, terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

Hereinafter, embodiments of the present invention will be described in more detail with reference to the accompanying drawings.

1 is a block diagram showing the overall configuration of an online test cheating monitoring system according to an embodiment of the present invention.

Referring to FIG. 1 , the online test cheating monitoring system according to an embodiment of the present invention includes a plurality of user terminals 110_1, 110_2, ?, 110_n, a network 120, and the It is configured to include a cheating monitoring device 130 that monitors cheating of online exams performed by test takers.

In the user terminals 110, an online test application provided by the cheating monitoring device 130, as an example of a server for submitting online test questions, may be installed, through which the examinee accesses the network through his/her terminal. Certain tests may be performed. If the online test application is not installed in the user's terminal, the user may download the application from the cheating monitoring device 130 and install it in the user's terminal.

The user terminals 110 include mobile devices that a user can carry and carry, and may be smartphones, tablets, laptops, etc., as well as devices that can be used by being connected to a network, such as a desktop computer. can be implemented

The network 120 may be implemented as a wireless communication network or a wired communication network. In this case, the communication method is not limited, and not only a communication method using a communication network (eg, a mobile communication network, a wireless LAN network, a wired Internet, a broadcasting network) that the network 120 may include, but also short-range wireless communication between devices may be included.

The cheating monitoring device 130 performs an operation as a server for submitting online test questions, and uses sensors (eg, cameras, microphones, etc.) installed in each of the user terminals 110 to prevent the test takers taking the online test. It is possible to receive test environment information including video and audio information, analyze it, and perform an operation to detect an act suspected of cheating by the examinee. That is, the cheating monitoring device 130 may be composed of at least one server device.

The fraud monitoring apparatus 130 according to an embodiment of the present invention can detect whether the examinee has cheated by analyzing the image information of the examinee received through an image sensor (eg, a camera) installed in the user terminals 110 . In addition, by analyzing the examinee's audio information received through an audio sensor (eg, a microphone) installed in the user terminals 110 to detect whether the examinee has cheated, more reliability and fairness are secured on-line test environment It is characterized in that it can provide. A detailed configuration and operation of the fraud monitoring apparatus 130 according to an embodiment of the present invention will be described in more detail later with reference to FIGS. 3 to 7 .

FIG. 2 is a block diagram illustrating the configuration of the user terminal shown in FIG. 1 .

Referring to FIG. 2 , each user terminal 110 includes a communication module 111 , a memory 112 , a display module 113 , an audio output module 114 , a control unit 116 , and sensors 115a and 115b ). may include. In this case, the sensors 115a and 115b may include a camera 115a as an image sensor and a microphone 115b as an audio sensor as illustrated.

The communication module 111 may communicate with any internal component or at least one external device through the network 120 . In this case, any external device may include, for example, the cheating monitoring device 130 shown in FIG. 1 . Here, as wireless Internet technologies, wireless LAN (WLAN), DLNA (Digital Living Network Alliance), WiBro (Wireless Broadband: Wibro), Wimax (World Interoperability for Microwave Access: Wimax), HSDPA (High Speed Downlink Packet Access) ), High Speed Uplink Packet Access (HSUPA), IEEE 802.16, Long Term Evolution (LTE), Long Term Evolution-Advanced (LTE-A), Wireless Mobile Broadband Service (WMBS), etc. In this case, the communication module 111 transmits and receives data according to at least one wireless Internet technology within a range including Internet technologies not listed above.

In addition, short-range communication technologies include Bluetooth, Radio Frequency Identification (RFID), Infrared Data Association (IrDA), Ultra Wideband (UWB), ZigBee, and Near Field Communication (NFC). , Ultra Sound Communication (USC), Visible Light Communication (VLC), Wi-Fi (Wi-Fi), Wi-Fi Direct (Wi-Fi Direct), etc. may be included.

The memory 112 stores various user interfaces (UIs), graphic user interfaces (GUIs), and the like. In addition, the memory 112 stores data and programs necessary for the user terminal 110 to operate. That is, the memory 112 may store a plurality of application programs (hereinafter, 'applications') driven in the user terminal 110 , data for operation of the user terminal 110 , and commands. That is, at least one or more applications 112a are stored in the memory 112 of the user terminal 110 according to the embodiment of the present invention. At least some of these applications 112a may be downloaded from an external server through wireless communication. An embodiment of the present invention is characterized in that it includes an online test application provided by the cheating monitoring device 130 shown in FIG. 1 .

In addition, the applications 112a are automatic speech recognition (ASR) applications, map applications, media applications (eg, QuickTime, MobileMusic.app, or MobileVideo.app), social networking applications (eg, For example, Facebook, Twitter, etc.), may be implemented as an Internet browsing application. Meanwhile, the applications 112a may be stored in the memory 112 , installed in the user terminal 110 , and driven to perform an operation (or function) of the user terminal 110 by the controller 116 . .

In addition, the memory 112 is a flash memory type (Flash Memory Type), a hard disk type (Hard Disk Type), a multimedia card micro type (Multimedia Card Micro Type), a card type memory (eg, SD or XD memory, etc.) ), magnetic memory, magnetic disk, optical disk, RAM (Random Access Memory: RAM), SRAM (Static Random Access Memory), ROM (Read-Only Memory: ROM), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM ( Programmable Read-Only Memory) may include at least one storage medium.

The display module 113 displays various contents such as various menu screens using a user interface and/or a graphic user interface stored in the memory 112 under the control of the control unit 116, and is controlled by the control unit 116. An execution result screen of the application executed by the user can be displayed. For example, when the online test application is executed, the display module 113 may provide an interface for solving online test questions provided by the server. The display module 113 is a liquid crystal display (LCD), an organic light-emitting diode (OLED), a flexible display (Flexible Display), a three-dimensional display (3D display), an electronic ink display (e- ink display), LED (Light Emitting Diode), and the like.

The audio output module 114 may output voice information included in a predetermined signal-processed signal under the control of the controller 116 .

An embodiment of the present invention is a system for monitoring whether or not the test takers are cheating on an online test through a user terminal. In order to implement this, predetermined sensors 115a and 115b are installed in each of the user terminals 100 . It is characterized in that the test environment information of each examinee sensed by the sensors is used as information for detecting whether or not cheating occurs.

More specifically, the sensors 115a and 115b may include a camera 115a as an image sensor and a microphone 115b as an audio sensor, and the fraud monitoring device 130 is a camera installed in the user terminals 110 . It is possible to receive the image information of the examinee sensed through 115a and analyze it to detect whether the examinee has cheated, as well as detect the examinee's image information sensed through the microphone 115b installed in the user terminals 110 . Audio information may also be received and analyzed to detect whether the examinee has cheated.

3 is a block diagram showing the configuration of a fraud monitoring apparatus according to an embodiment of the present invention.

Referring to FIG. 3 , the fraud monitoring device 130 may be implemented as a server including a memory 132 , a database 133 , a processor 134 , a communication module 136 and an input/output interface 138 , This not only serves to provide online test questions by being connected to the user terminals 110 in which the online test application is installed, but also for each examinee sensed from the sensors 115a and 115b of the user terminals 110 as mentioned above. An operation of detecting whether the test takers are cheating may be performed using the test environment information.

The memory 132 is a computer-readable recording medium and may include a non-volatile mass storage device such as a RAM, a ROM, and a disk drive. Also, an operating system and at least one program code may be stored in the memory 132 . These software components may be loaded from a computer-readable recording medium separate from the memory 132 using a drive mechanism. The separate computer-readable recording medium may include a recording medium such as a floppy drive, a disk, a tape, a DVD/CD-ROM drive, and a memory card. In addition, the software components may be loaded into the memory 132 through the communication module 136 . In the embodiment of the present invention, test environment information of each test taker sensed by the sensors 115a and 115b of the user terminals 110 may be stored in the memory 132 .

The database 133 may store and maintain some of the information transmitted and received from the user terminals 110 connected to the fraud monitoring apparatus 130 according to an embodiment of the present invention. As an example, when it is determined that the test environment information of each examinee sensed from the sensors 115a and 115b of the user terminals 110 stored in the memory 132 is determined to be a suspicious act as a result of analysis, the corresponding The image information and/or audio information of the examinee may be separately extracted and stored and maintained in the database 133 .

In addition, the database 133 serves as a member information DB to store various information such as unique information and online test questions corresponding to the user terminals 110 connected to the fraud monitoring device 130 . can be

The processor 134 may be configured to process instructions of a computer program by performing basic arithmetic, logic, and input/output operations. The instructions may be provided to the processor 134 by the memory 132 or the communication module 136 . As an example, the processor 134 may be configured to execute a received instruction according to a program code stored in a recording device such as the memory 132 . That is, the processor 134 may include a program module that is implemented by software such as C, C++, Java, Visual Basic, Visual C, and the like and performs various functions.

The processor 134 of the cheating monitoring device 130 according to the embodiment of the present invention receives the image information of the examinee sensed through the camera 115a of the user terminal 110, and analyzes it to analyze the examinee's cheating. Receives the examinee's audio information sensed through the cheating behavior detection unit ( 410 in FIG. 4 ) and the microphone 115b of the user terminal 110 that can detect the behavior, and analyzes it to detect the examinee's cheating voice It may include a cheating voice detection unit ( 420 in FIG. 4 ) that can do it. At this time, the fraudulent voice detection unit 420 may include a sound generation section detection unit (422 in FIG. 4), a speaker detection unit (424 in FIG. 4), and a voice recognition unit (426 in FIG. 4), and Various functions implemented through the processor 134 of the fraud monitoring apparatus 130 according to the embodiment will be described in more detail below with reference to FIGS. 4 to 7 .

The communication module 136 may provide a function for the user terminals 110 and the fraud monitoring apparatus 130 to communicate with each other through the network 120 . For example, a control signal, command, content, file, etc. provided under the control of the processor 134 of the fraud monitoring device 130 is transmitted through the network 120 through the communication module 136 to the user terminals 110 . can be transmitted to

The input/output interface 138 may serve to provide an interface between an input device implemented as a keyboard or mouse and an output device such as a display.

4 is a block diagram showing an embodiment of the internal configuration of the processor shown in FIG. 3, and FIG. 5 is a diagram schematically illustrating an operation of detecting cheating by the online test cheating monitoring system according to an embodiment of the present invention. to be.

Also, FIG. 6 is a flowchart illustrating a method for detecting cheating using the image information shown in FIG. 5 , and FIG. 7 is a flowchart illustrating a method for detecting cheating using the audio information shown in FIG. 5 .

Hereinafter, each functional block of the cheating monitoring device according to an embodiment of the present invention and the cheating detection operation in the online test performed by the cheating monitoring device according thereto will be described in more detail with reference to FIGS. 4 to 7 . do.

Referring first to FIG. 4 , the processor 134 of the cheating monitoring device 130 receives the image information of the examinee sensed through the camera 115a of the user terminal 110 , and analyzes it to analyze the examinee’s cheating behavior. A cheating behavior detection unit 410 capable of detecting a behavior, receiving audio information of an examinee sensed through the microphone 115b of the user terminal 110, and analyzing it to detect a cheating voice of the examinee The cheating voice detection unit 420 may be included, and the cheating voice detection unit 420 may include a sound generation section detection unit 422 , a speaker detection unit 424 , and a voice recognition unit 426 in detail.

In addition, in identifying the user terminal 110 of the test takers connected to the cheating monitoring device 130 , the processor 134 may be linked with the database 133 for storing information provided to the user terminal 110 . The processor 134 performs the steps of detecting the cheating behavior using the image information of the examinee (FIG. 6) and the examinee's audio information The cheating monitoring device 130 may be controlled to perform the steps (FIG. 7) of detecting a cheating voice using the .

Here, the processor 134 may be implemented to execute an instruction according to an operating system code and at least one program code included in the memory 132 . At this time, the components in the processor 134, that is, the fraudulent behavior detection unit 410, the sound generation section detection unit 422 as the cheating voice detection unit 420, the speaker detection unit 424, and the voice recognition unit 426 are It may be understood that different functions performed by the processor 134 according to a control command provided by the program code stored in the fraud monitoring device 130 are distinguished and expressed.

The cheating behavior detection unit 410 receives the image information of the examinee sensed through the camera 115a of the user terminal 110 and analyzes it to detect the cheating behavior of the examinee. For example, in the case of an embodiment of the present invention, in analyzing the received image information of the examinee, it is continuously checked whether the test taker included in the image information is the previously registered examinee, and the image information of the examinee is included in the image information. It can be checked whether other people are detected, and based on this, it is possible to detect cheating behavior of the examinee.

Referring to FIG. 5 , the examinee 200 may perform an online test using the user terminal 110 through the online test environment provided by the online test cheating system as shown in FIG. 1 above. Accordingly, for the examinee 200 performing the online test, the image information of the examinee sensed through the camera 115a of the user terminal 110 passes through the network (120 in FIG. 1 ) to the cheating monitoring device (130 in FIG. 1 ). ) can be transmitted. In addition, the cheating behavior detection unit 410, which is a functional block in the processor of the cheating monitoring device, extracts a single frame image or a continuous section frame image from the received image information of the examinee, and analyzes it to analyze the cheating behavior of the examinee. It is possible to perform an action to detect

A detailed operation of the cheating behavior detection unit 410 will be described in more detail with reference to the flowchart shown in FIG. 6 . 6 is a flowchart illustrating steps of detecting a cheating behavior using image information of an examinee during the operation of the cheating monitoring device.

Referring to FIG. 6 , first, a single frame image or a continuous section frame image is extracted from the image information of the examinee sensed through the camera 115a of the user terminal 110 (ST600). Thereafter, the behavior of the object included in the extracted at least one frame image is compared with a previously selected and stored cheating pattern (ST610).

In an embodiment of the present invention, it is characterized in that the behavior of the object included in the frame image is determined by the examinee using an action recognition technology. As an example, briefly describing the cheating detection process using the behavior recognition technique, first, the types of cheating (eg, smartphone use behavior, book usage behavior, seat shift behavior, etc.) are defined, and As a method of learning an action recognizer model by collecting raw images, this means that the cheating behavior detection unit 410 can be learned by a machine learning technique.

In addition, as mentioned above, for example, in determining the misconduct, whether the test taker included in the received image information of the examinee is the previously registered examinee, and whether the image information contains information other than the examinee. Whether or not a person is detected may be additionally considered.

Accordingly, the predetermined cheating pattern is continuously updated, and such information may be stored in the database 133 of the cheating monitoring device 130 .

By comparing the behavior of the object included in the extracted at least one frame image with the previously selected and stored cheating patterns, the behavior of the object is a pre-selected type of cheating patterns, for example, smartphone usage behavior, book usage behavior , it is determined whether the behavior of the test taker is a cheating behavior, that is, it is determined whether the behavior of the examinee is a cheating behavior (ST 620), and when it is determined that the behavior of the object conforms to the cheating patterns, among the extracted frame images A frame image including a behavior suspected of cheating is stored in the database 133 (ST 620).

The cheating monitoring apparatus 130 according to the embodiment of the present invention analyzes the examinee's image information received through the image sensor (eg, camera) installed in the user terminals 110 as described with reference to FIG. In addition to being able to detect whether or not cheating has occurred, audio information of an examinee received through an audio sensor (eg, a microphone) installed in the user terminals 110 is also analyzed to detect whether the examinee has cheated, making it more reliable and It is characterized in that it can provide an online test environment in which fairness is secured.

Accordingly, the cheating voice detection unit 410 receives the examinee's audio information sensed through the microphone 115b of the user terminal 110, analyzes it, and detects the examinee's cheating voice. The cheating voice detection unit 420 may include a sound generation section detection unit 422 , a speaker detection unit 424 , and a voice recognition unit 426 .

Referring to FIG. 5 , audio information of the examinee sensed through the microphone 115b of the user terminal 110 for the examinee 200 performing the online test is transmitted through the network ( 120 in FIG. 1 ) to the cheating monitoring device. (130 in FIG. 1) may be transmitted. In addition, the cheating voice detection unit 420, which is a functional block in the processor of the cheating monitoring device, sequentially extracts audio frames from the received audio information of the examinee and analyzes them to detect the cheating voice of the examinee. can be done

More specifically, the sound generation section detector 422 of the cheating voice detector 420 captures and stores the section in which a sound is generated in the extracted audio frame, and then the speaker detector 424 performs the The voice is separated for each speaker from the stored sound generation section data, the operation of storing the separated voice section for each speaker is performed, and finally, the voice recognition unit 426 compares and analyzes the voice of each speaker with the voice of the examinee. It is possible to perform an operation to detect the test taker's cheating voice.

The detailed operation of the cheating voice detection unit 420 will be described in more detail through the flowchart shown in FIG. 7 . 7 is a flowchart illustrating steps of detecting a cheating voice using audio information of an examinee during the operation of the cheating monitoring device.

Referring to FIG. 7, first, the sound generation section detecting unit 422 detects a section in which a sound is generated among sequential audio frames including audio information of the examinee sensed through the microphone 115b of the user terminal 110. Save (ST700). Thereafter, the speaker detection unit 424 determines whether human voice data is included among the generated sound data in the stored sound generation section data (ST 710). When is recognized, the voice data is separated for each speaker, and the separated voice data section for each speaker is stored (ST720). Then, the voice recognition unit 426 separates and extracts the voice data for each speaker by using the voice recognition function, converts it into text and stores it (ST720), and compares the converted text with the actually submitted online test text. do (ST730).

Thereafter, by comparing the converted text with the actual submitted online test question text, it is determined whether the converted text substantially matches the actual test question text, that is, the converted text corresponding to the examinee's voice corresponds to cheating. it is determined (ST 750), and when it is determined that the converted text is substantially consistent with the actual test question text, the voice data of the suspected cheating section corresponding to the extracted voice data for each speaker and the converted text information is stored in the database 133 (ST 620).

In another embodiment, in the step of determining whether or not there is cheating (ST 750), when a plurality of speaker-specific voices are detected in the detected sound, it is considered that several people gather and talk and conduct an online test, and this behavior is regarded as cheating. can be judged as

In addition, the step (ST 750) of determining whether or not cheating is performed is to take the online test while making a phone call, etc., if the examinee's own voice continues for more than a preset time (eg, about 20 seconds or more) to the detected sound. These actions may be regarded as misconduct.

According to the online test cheating monitoring system according to the embodiment of the present invention, test environment information including video and audio information of an examinee taking an online test using a sensor (eg, camera, microphone, etc.) installed in a user terminal It is possible to provide an online test environment in which reliability and fairness are secured by receiving and analyzing the test taker to detect suspected misconduct.

As described above, the present invention has been described with specific matters such as specific components and limited embodiments and drawings, but these are only provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. , various modifications and variations are possible from these descriptions by those of ordinary skill in the art to which the present invention pertains.

Therefore, the spirit of the present invention should not be limited to the described embodiments, and not only the claims to be described later, but also all those with equivalent or equivalent modifications to the claims will be said to belong to the scope of the spirit of the present invention. .

Claims (20)

a cheating monitor as a server that provides online exam questions;
The online test application provided by the fraud monitoring device is installed, and includes a plurality of user terminals connected to the cheating monitoring device through a network,
Each of the user terminals includes first and second sensors, and transmits test environment information of test takers sensed through the first and second sensors to the cheating monitoring device,
The fraud monitoring device,
a cheating behavior detection unit receiving the image information of the examinee sensed through the first sensor installed in the user terminals, and analyzing the image information to detect the cheating behavior of the test takers; and
and a cheating voice detector configured to receive the test taker's audio information sensed through the second sensor installed in the user terminals, and analyze the test taker's cheating voice to detect the test takers' cheating voice. The method of claim 1,
The first sensor includes a camera as an image sensor, and the second sensor includes a microphone as an audio sensor. The method of claim 1,
The fraud monitoring device,
a memory for storing test environment information of the test takers sensed by the first and second sensors; and
The online test cheating monitoring system further comprising a database for separately extracting, storing and maintaining the detected cheating behavior and/or cheating voice information of the test takers from among the test environment information of the test takers stored in the memory. 4. The method of claim 3,
The database is an online test cheating monitoring system for storing and maintaining unique information of user terminals connected to the cheating monitoring device and the online exam questions. The method of claim 1,
The cheating behavior detection unit,
Extracting a single frame image or a continuous section frame image from the received image information of the examinee, and using an action recognition technique for the action of the object included in the extracted at least one frame image An online test cheating monitoring system for determining whether a behavior is a cheating behavior, and storing a frame image including a behavior suspected of cheating among the extracted frame images. 6. The method of claim 5,
The behavior recognition technique is
The types of cheating are defined, and images are collected for each defined type of cheating, the patterns of cheating are selected in advance through action recognizer model learning, and the predetermined cheating patterns and the An online test cheating monitoring system that compares the behavior of the object included in the frame image to determine whether the examinee's behavior substantially matches the cheating patterns. The method of claim 1,
The fraudulent voice detection unit,
a sound generation section detecting unit for detecting and storing a section in which a sound is generated from among sequential audio frames including audio information of the examinee sensed by the second sensor;
a speaker detection unit that separates voices for each speaker from the stored sound generation section data and stores the separated voice sections for each speaker; and
and a voice recognition unit configured to compare and analyze each speaker's voice and the examinee's voice to detect the test taker's cheating voice. 8. The method of claim 7,
The speaker detection unit is an online test cheating monitoring system for determining whether human voice data is included in the generated sound data in the stored sound generation section data. 9. The method of claim 8,
When it is determined that the voice data is included, the speaker detection unit separates the voice data for each speaker and stores the separated voice data section for each speaker when multiple speakers are recognized through conversation surveillance system. 10. The method of claim 9,
The voice recognition unit separates and extracts the voice data for each speaker by utilizing a voice recognition function, converts it into text and stores it, and compares the converted text with the actually submitted online test text to ensure that the converted text is actually It is determined whether or not it substantially matches the test question text, and when it is determined that the converted text is substantially consistent with the actual test question text, the voice data of the suspected cheating section corresponding to the extracted voice data for each speaker and its conversion Online exam cheating monitoring system that stores text information. A cheating monitoring device as a server that provides online test questions and an online testing application provided by the cheating monitoring device are installed, and online exam cheating monitoring including a plurality of user terminals connected to the cheating monitoring device through a network In the method,
a cheating behavior detection step of receiving the image information of the examinee sensed through the first sensor installed in the user terminals, and analyzing the image information to detect the cheating behavior of the test takers; and
and a fraudulent voice detection step of receiving the test taker's audio information sensed through the second sensor installed in the user terminals, and analyzing the test taker's cheating voice to detect the test taker's cheating voice. 12. The method of claim 11,
The first sensor includes a camera as an image sensor, and the second sensor includes a microphone as an audio sensor. 12. The method of claim 11,
storing the test environment information of the test takers sensed by the first and second sensors in a memory; and
The online test cheating monitoring method further comprising the step of separately extracting the detected cheating behavior and/or cheating voice information of the test takers from among the test environment information of the test takers stored in the memory, storing and maintaining in a database. 14. The method of claim 13,
The database is an online test cheating monitoring method for storing and maintaining unique information of user terminals connected to the cheating monitoring device and the online exam questions. 12. The method of claim 11,
The fraudulent behavior detection step includes:
extracting a single frame image or a continuous section frame image from the received image information of the examinee;
determining whether the action of the object is a cheating action by using an action recognition technique with respect to the action of the object included in the extracted at least one frame image; and
and storing a frame image including a behavior suspected of cheating among the extracted frame images. 16. The method of claim 15,
The behavior recognition technique is
The types of cheating are defined, and images are collected for each defined type of cheating, the patterns of cheating are selected in advance through action recognizer model learning, and the predetermined cheating patterns and the An online test cheating monitoring method for comparing the behavior of the object included in the frame image to determine whether the behavior of the examinee substantially matches the cheating patterns. 12. The method of claim 11,
The fraudulent voice detection step includes:
a sound generation section detecting step of detecting and storing a section in which a sound is generated from among sequential audio frames including audio information of the examinee sensed through the second sensor;
a speaker detection step of separating a voice for each speaker from the stored sound generation section data and storing the separated voice section for each speaker; and
and a voice recognition step of detecting the test taker's cheating voice by comparing and analyzing the speaker's voice and the test taker's voice. 18. The method of claim 17,
The speaker detection step is
and determining whether human voice data is included in the generated sound data in the stored sound generation section data. 19. The method of claim 18,
The speaker detection step is
When it is determined that the voice data is included, if a plurality of people talk and a plurality of speakers are recognized, separating the voice data for each speaker, and storing the separated voice data section for each speaker monitoring method. 20. The method of claim 19,
The voice recognition step is
Separating and extracting the voice data for each speaker using a voice recognition function, converting it into text and storing the same;
comparing the converted text with the actual submitted online test question text to determine whether the converted text substantially matches the actual test question text; and
When it is determined that the converted text is substantially consistent with the actual test question text, online test fraud comprising the step of storing the voice data of the suspected cheating section corresponding to the extracted speaker-specific voice data and the converted text information How to monitor behavior.

Images