KR102593624B1 - Online Test System using face contour recognition AI to prevent the cheating behaviour and method thereof - Google Patents

Abstract

An online examination system and method using facial contour recognition artificial intelligence to prevent cheating are disclosed. The system includes an online test server that registers test taker information and frontal face photos, issues a QR code, and provides a test program for conducting an online test; It is connected to the online test server through a wired or wireless communication network, and recognizes the artificial intelligence facial recognition module of the face captured by the front camera of the test taker's terminal and the facial behavior pattern captured by the front camera, and recognizes the outline of the face, eyes 2, nose, and ears. 2. A test taker terminal equipped with a 5-point scale of facial features and an anti-cheating module; and a proctor terminal that receives facial photos and cheating information of one or more test taker terminals through the online test server during an online test, and transmits a cheating prevention alarm or warning message to the test taker terminal.
The test taker terminal is equipped with the facial recognition module and the cheating prevention module, and the cheating prevention module detects visual cheating in the video data of the camera of the test taker terminal in real time, such as the facial outline and eyes2/nose/ It is further equipped with a 5-point scale fraud prevention module for ear 2's facial features and a voice recognition module for detecting auditory fraud.

Description

Online test system using face contour recognition AI to prevent the cheating behavior and method thereof}

The present invention relates to an online examination system and method using artificial intelligence to prevent cheating, and more specifically, to tablet PCs and smartphones equipped with an artificial intelligence facial recognition module and a 5-point facial feature point scale cheat prevention module. , In PC-based online exams and UBT exams, AI-based facial outline recognition technology (posenet algorithm) is used to extract the object of the test taker's face from the test taker's terminal and detect the facial outline and eyes2 to prevent proxy testing or cheating. /nose/ear2 feature points are extracted, and the anti-fraud module detects and tracks the head movement to the right/left depending on whether the eye/nose 3 points get closer to the ear 2 points at either end, and detects facial behavior patterns using an artificial device. It is equipped with an intelligent facial recognition module and a 5-point facial feature point scale fraud prevention module. When facial outlines are not recognized during face recognition, or when the tablet PC's camera image deviates from the test screen where it is recorded (eyes/nose 3 points are positive) (If the distance in the relevant direction, such as eye/ear distance, nose/ear distance, etc. exceeds a certain value depending on whether it is approaching the end ear 2 point), a warning message or alarm is output or saved in the information device of the test taker's terminal at the end of the test. It relates to a tablet PC, smartphone, and PC-based online test system and method using facial contour recognition artificial intelligence to prevent cheating, transmitting to the test server and providing scoring results to test takers.

Face recognition technology mainly uses appearance-based matching method and feature-based face recognition. Face recognition is recognized differently depending on the camera's shooting angle, direction of lighting, posture, changes in expression, and changes in the face over time.

Feature-based face recognition uses a detection method using Haar-like features and a detection method using MCT (Modified Census Transform) images from image data captured with a digital camera, IoT device camera, or smartphone camera. . Using a face and eye detector learned with Haar-like features from the input image of the smartphone camera, the outline of the face and forehead/eyes/nose/mouth are detected, and the region of interest (ROI) is used to detect circular pupils. Interest) is converted to grayscale, and the histogram of the eye image (pixel value of each pixel on the x-axis, After calculating the number of corresponding pixel values on the y-axis and binarizing the eye image, preprocessing of the photo in the eye area is performed through histogram equalization, and in the face area, eyebrows, eyes, nose, mouth, The face is recognized by detecting the face data of the facial features of the outline, extracting the texture features and shape features, and comparing the similarity with the feature points of the face photo stored in the face recognition DB.

The feature values of the eyebrows, eyes, nose, mouth, and chin in the face area are expressed as the difference between the sum of pixels included in the white area of the Haar-like feature and the sum of the pixels included in the black area.

For example, the distance from the eye area detected in a face area photo of standard width and height size to both endpoints of the right and left eyes, and the size of the pupil (iris) extracted using the Hough circle transform algorithm. The value is used as the feature value.

As prior art 1 related to this, Patent Publication No. 10-2017-0050465 discloses “Face recognition device and method.”

According to this embodiment, when recognizing faces from input images using machine learning, the face recognition rate is improved by normalizing face pose and perspective, and face learning data is acquired by automatically generating virtual face images as face learning data. Provides a facial recognition device and method that saves cost and time.

1 is a configuration diagram of an existing face recognition device.

The face recognition device 100 is an image display device, an image capture device, a face recognition server, a tablet PC, a laptop, a personal PC, a smartphone, a personal digital assistant (PDA), a mobile communication terminal, and It can be any one of the intelligent robots.

The face recognition device 100 includes an input image acquisition unit 112 that acquires an input image from a camera; A front pose image is generated by detecting the face area in the input image and normalizing the face pose, and the perspective of the front pose image is used to remove perspective distortion according to the distance between the camera and the subject. ) a normalization unit 114 that normalizes and generates a normalized image; a feature vector extraction unit 116 that extracts a feature vector representing the face of the subject from the normalized image; and a face recognition unit 118 that recognizes the face of the subject included in the input image by applying the feature vector to a previously learned classification model.

The input image acquisition unit 112 acquires the input image input from the camera. The cameras may be depth perception cameras, stereo cameras, and color cameras (eg, Kinect cameras, etc.). Additionally, the input image is an image containing the face of a subject to be recognized and includes a two-dimensional still image and a moving image. Input images may include color images, depth images, and color-depth (RGB-D) images.

The normalization unit 114 detects the face area from the input image and normalizes the face pose and perspective to generate a normalized image. If there is a change in the face pose, the gray scale, shape, and location of feature points change, so the face recognition rate deteriorates. Additionally, if the distance between the camera and the subject changes, even if it is the same subject, different perspective distortions occur at each photographed location, making it appear as if a different subject was photographed. Therefore, to improve the face recognition rate, it is necessary to normalize the facial pose and perspective of the input image.

The normalization unit 114 inputs facial images for learning in various poses to the input layer of the first artificial neural network, and trains the first artificial neural network so that facial images for learning in the frontal pose are output from the output layer of the first artificial neural network. Face pose normalization learning unit; And inputting data output from the output layer of the first artificial neural network to the input layer of the second artificial neural network, and training the second artificial neural network so that a facial image for learning without perspective distortion is output from the output layer of the second artificial neural network. Includes a perspective normalization learning unit.

The normalization unit inputs learning face images of various poses with various perspective distortions to the input layer of an integrated artificial neural network that integrates the first artificial neural network and the second artificial neural network for which learning has been completed, and has no perspective distortion of the frontal pose. The integrated artificial neural network is trained so that learning face images are output from the output layer of the integrated artificial neural network.

The feature vector extraction unit 116 is determined by machine learning and extracts a feature vector representing the subject's face from the normalized image.

A feature vector is a vector whose elements include feature values used for face recognition. Filters used to extract feature vectors include Gabor filter, Haar filter, LBP (Local Binary Pattern) - including DLBP (Discriminative LBP), ULBP (Uniform LBP), NLBP (Number LBP) - etc., but are necessarily limited to these. does not work and other filters can be used.

The face recognition unit 118 recognizes the face of the subject included in the input image by applying the feature vector extracted by the feature vector extractor 116 to the previously learned classification model. Pre-trained classification models may include, but are not necessarily limited to, Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), and Softmax.

The virtual face image generator 124 may generate a plurality of virtual face images that are used for learning by the normalizer 114, the feature vector extractor 116, and the face recognition unit 118.

The plurality of virtual face images refers to face images generated by the virtual face image generator 124 transforming a 3D face model synthesized using one or more 2D reference images obtained from a camera.

* Features of the facial recognition system

Recognition rate varies depending on lighting or distance from the camera.

The shape of the face continues to change depending on the angle, expression, and age of the face, and may be perceived differently depending on changes in appearance such as eyebrows, glasses, and makeup.

* Configuration steps of face recognition system

- Image acquisition: Image acquisition from CCD camera or CMOS Image Sensor camera

- Preprocessing: Noise removal image processing and segmentation

- Face detection: Detect face area in ROI area from input image

- Face normalization: standard size, feature extraction, brightness and geometric normalization

- Face recognition: Comparison and recognition of detected images and database images

* Geometrical methods

- Features: Recognition by extracting geometric feature points of the face and determining whether they match

- Performance: Because the face is three-dimensional and can be rotated, there are limitations to its application.

* Eigenfaces: Developed by Pentland in 1991

- Features: PCA (Principal Component Analysis) is applied to extract feature points of the face, and Euclidean distance is applied to measure similarity.

- Performance: A representative face recognition method that responds sensitively to changes in lighting or environment, but is widely used and has proven performance.

* Fisherfaces

- Features: A face recognition method that applies FLD (Fisher Linear Discriminant) as a classification algorithm.

- Performance: An algorithm that is more accurate and insensitive to environmental changes by learning the characteristics of each person, and can be applied in real time on-line.

* Method based on SVM (Support Vector Machine)

- Features: Algorithm using PCA and SVM (Support Vector Machine).

- Performance: An algorithm that achieves a relatively high recognition rate, but consumes a lot of time and memory to apply to multi-class applications such as face recognition. A representative face recognition algorithm currently being studied.

* Neural network

- Features: Retrieve learned facial shapes using a sliding window in black and white still images

- Performance: Extraction of two or more faces is possible, but the speed is slow and learning is difficult.

* Fuzzy + Neural Network

- Feature: Uses a fuzzy membership function instead of the brightness value of the pixel as input to the neural network circuit.

- Performance: Performance is improved compared to the method using only neural networks, but processing speed is lower.

* Wavelet + Elastic Matching

- Features: Uses frequency conversion and is effective in processing changes in posture and facial expression.

- Performance: The amount of calculation is large compared to the recognition rate.

As prior art related to this, Patent Registration No. 10-1765770 registers "A system and method for detecting cheating in an online test using image processing and a recording medium storing a program for implementing the same."

The online test cheating detection system and method uses image processing to detect suspected cheating by test takers during online tests, thereby supervising online test cheating and ensuring the reliability and authenticity of online tests. Provides a system and method for detecting cheating in online exams using image processing, and a recording medium storing a program for implementing the same.

Figure 2 is a configuration diagram of an online evaluation system including a conventional online test cheating detection system.

The online evaluation system includes a test taker terminal 110 for the test taker to take the online test; A camera 120 attached to the examinee's terminal or provided in a smart device to photograph the front of the examinee; a cheating detection server 100 that receives video information of the test taker from the camera in units of frames and detects cheating through image processing; When the cheating detection server determines that the test taker is suspected of cheating, a proctor terminal 130 for notifying the proctor of this; and an online evaluation server 140 for presenting questions for the online exam to be taken by the test taker.

The cheating detection server 100 receives the examinee's video information through a communication network, and when it determines that cheating is suspected, it sends a cheating detection message to the supervisor and sends it to the online evaluation server 140. A transmitting and receiving unit 101 for informing to stop the online test; a storage unit 102 for storing the examinee's image information and coordinate information received through the transceiver 101; a face recognition unit 103 for defining a rectangular area including a face area from the received image information of the examinee; a pupil coordinate acquisition unit 104 that sets pupil coordinates of both eyes by image processing the shape and brightness of both eyes in a coordinate system based on the rectangular area; An intermediate coordinate acquisition unit 105 that determines the intermediate coordinates of the pupil coordinates of both eyes, a background recognition unit 106 that sets all areas except the examinee's body and face areas as the background, and the face recognition unit ( Whether or not there are two or more face areas acquired in 103), the reference pupil coordinate value received from the pupil coordinate acquisition unit 104 is compared with the average real-time pupil coordinate value, and the real-time pupil coordinate value is determined based on the reference pupil coordinate value. Whether the shape balance value is outside the preset error range, whether the moving distance of the intermediate coordinates received from the intermediate coordinate acquisition unit 105 is outside the preset error range, and the background received from the background recognition unit 106 It includes a fraud determination unit 107 to determine whether a change in the area is detected.

In the embodiment, the rectangular area including the face area was defined based on the content disclosed in the paper "Robust Real-Time Face Detection Using Face Certainty Map" (Volume 4642 of the series Lecture Notes in Computer Science pp 29-38).

The fraud detection server 100,

A rectangular area containing the examinee's face area is defined from the frame-by-frame image information in which the examinee is captured, and the shape and contrast of both eyes are image processed in a coordinate system based on the rectangular area, and both end points of both eyes (310 , 320) and set the center point (350, 360) of the black area (330, 340) located between them as the pupil coordinates of both eyes, and calculate the average value of the examinee's pupil coordinates for a certain period of time before the start of the test. The reference pupil coordinate value is obtained, and when the test starts, the average value of the pupil coordinates of both eyes of the test taker is calculated at regular intervals to obtain the real-time pupil coordinate average value, and the real-time pupil coordinate average value is based on the reference pupil coordinate value. If it is outside the set error range, this is notified to the supervisor terminal,

Determine the center point 370 of the pupil coordinates of both eyes as the intermediate coordinate,

The movement direction and movement distance of the intermediate coordinates are tracked on a frame-by-frame basis, and when the movement distance of the intermediate coordinates is outside a preset error range, the examinee's head rotation direction or movement direction is informed, and the examinee's frame-by-frame image information is provided. If there are two or more face areas obtained from , this is notified to the supervisor terminal,

All areas except the examinee's body and face are set as background, and when a change in the background is detected during the test time, this is notified to the proctor terminal,

The fraud detection server 100,

The examinee's video information is received through a communication network, and if it is determined that cheating is suspected, a cheating detection message is sent to the proctor terminal 130, and this is sent to the test taker's online exam question. Notifies the online evaluation server 140 that presents the questions to stop the online test.

When the online evaluation server 140 receives a message notifying that a suspected cheating behavior has been detected from the cheating detection server 100 while posing a question for an online test to be taken by the examinee, the online evaluation server 140 determines the examinee's online test. After stopping the question asking and receiving a message from the proctor terminal 130 or the cheating detection server 100 that the question may be restarted, the online test question is restarted.

Recently, face recognition technology uses head/eye/nose/mouth features using HOG, LBP, and Gabor feature extraction algorithms, which are previously mainstream hand-crafted features, and uses deep learning-based features for face detection and CNN. Using facial landmark detection and facial feature extraction and classification technology using the (Convolutional Neural Network) algorithm, it provides results similar to the human recognition level of 97.53%.

In 2014, artificial intelligence facial recognition technologies such as Facebook's DeepFace were released. In the case of DeepID1 ["Deep Learning Face Representation from Predicting 10,000 Classes," CVPR 2014 ] presented at CVPR, it uses 39x31 pixels, which is smaller than DeepFace's input resolution of 152x152 pixels.

In addition, AlexNet-based face detector, AlexNet, which was pre-trained with ImageNet training data, is fine-tuned with face images, so the detection performance is not high yet, and faces are detected using an SVM classifier in the final stage. .

Patent Publication No. 10-2017-0050465 (publication date May 11, 2017), “Face recognition device and method,” SK Telecom Co., Ltd. Patent registration number 10-1765770 (registration date August 1, 2017), “A system and method for detecting cheating in online exams using image processing and a recording medium storing a program for implementing the same”, Bu-kyung Son

The purpose of the present invention to solve the above problems is to provide a tablet PC equipped with an artificial intelligence facial recognition module and an eye2/nose/ear2 facial feature point 5-point scale anti-cheating module on the test taker's terminal during online and UBT tests. In smartphone and PC-based online tests and UBT tests, AI-based facial outline recognition technology (posenet algorithm) is used to detect objects on the test taker's face and outline and Facial feature points of eyes 2/nose/ear 2 are extracted, and the anti-fraud module detects and tracks head movement to the right/left depending on whether the 3 eyes/nose points get closer to the 2 ear points at either end, thereby detecting and tracking the facial behavior. It is equipped with an artificial intelligence facial recognition module that detects patterns and a fraud prevention module with a 5-point scale of facial features. If the facial outline is not recognized during face recognition, or if the tablet PC's camera image deviates from the test screen where the image is captured (eye / Depending on whether the 3 points of the nose are close to the 2 points of the ears at either end, if the distance between the eyes and ears or the distance in the corresponding direction of the distance between the nose and ears exceeds a certain value, a warning message or alarm is output, or an alarm is displayed on the information device of the candidate's terminal. We provide an online test system based on tablet PCs, smartphones, and PCs that uses facial contour recognition artificial intelligence to prevent cheating, which saves and transmits to the test server at the end of the test and provides scoring results to test takers.

Another object of the present invention is to provide a method of providing an online exam based on a tablet PC, smartphone, or PC using facial contour recognition artificial intelligence.

In order to achieve the purpose of the present invention, an online testing system using facial contour recognition artificial intelligence to prevent cheating registers test taker information and frontal face photos, issues a QR code attached to the test taker's test ticket, and provides online testing. An online test server that provides a test program for conducting; It is connected to the online test server through a wired or wireless communication network, and recognizes the artificial intelligence facial recognition module of the face captured through the front camera of the test taker's terminal and the facial behavior pattern captured by the front camera to determine the outline of the face and eyes 2/nose/ A test taker terminal equipped with a 5-point scale of facial features of ear 2 and an anti-cheating module; And during an online test or UBT test, a proctor terminal receives the face photo and cheating information of one or more test taker terminals through the online test server and sends an anti-cheating alarm or warning message to the test taker terminal. ,
The online test server further includes a greeting service (TTS) that provides a greeting through TTS conversion technology after recognizing the QR code attached to the test taker's application form,

The candidate terminal uses any one of a tablet PC, smartphone, or PC on which a test program (App) is installed. The test program is installed, and the artificial intelligence facial recognition module and the face of eyes 2/nose/ears 2 are used. It is equipped with a 5-point feature point scale fraud prevention module, and through facial feature extraction and classification, the facial feature points of eyes 2/nose/ear 2 are similar to the center point (pupil) of the left/right ear and left/right eye, respectively. Calculate the Clydian distance (d) and similarity, check whether a surrogate test is performed by comparing it with data from the standard-sized frontal face photo DB of the online test server, and use the data captured with the camera of the candidate's terminal during the online test. If the facial outline is not recognized when recognizing a face by recognizing facial behavior patterns, or if the face is off the test screen, the distance between the right eye and right ear and the distance between the left eye and left ear vary depending on the angle at which the front face is turned to the left or right, so the distance between the eyes and left ear varies. If the distance in the corresponding direction, such as the distance between the ears and the distance between the nose and the ears, exceeds a certain standard, it is judged to be an act of misconduct.

The AI-based facial contour recognition technology used in the facial recognition module uses the posenet algorithm to measure the distance between the eyes and ears and the distance between the nose and ears.
If the facial behavior pattern captured by the camera of the test taker's terminal is recognized and judged to be cheating, it is transmitted to the online test server, and after confirmation by the proctor's terminal linked to the online test server, an alarm or warning message is sent to the test taker's terminal. occurs.

In order to achieve another purpose of the present invention, a method of providing an online test using facial contour recognition artificial intelligence to prevent cheating involves registering member information, and after logging in/user authentication, applicant information and a standard-sized frontal face photo are provided online. Registering and saving as a test server; The online test server issuing a QR code attached to the test taker's test ticket corresponding to the test taker information and frontal face photo; The online test server notifies the test schedule and location according to the administered test; The test taker terminal equipped with a facial recognition module using AI-based facial contour recognition technology uses a frontal face recognition algorithm to receive the recognition results of the frontal face photo from the camera to the online test server and size-corrected/rotated/angle-corrected standards. Generating a full-size frontal face photo and comparing the proctor terminal with the applicant information already stored in the online test server and the standard-size frontal face photo to confirm the applicant's identity, preventing proxy testing, and determining whether to take the test; The facial recognition module and the 5-point scale cheating prevention module of the candidate terminal recognize the outline of the candidate's face and the facial feature points on the eye 2/nose/ear 2 5-point scale, and the facial feature points of eye 2/nose/ear 2 are recognized. Calculate the Euclidean distance (d) and similarity between the left/right ear and the center point (pupil) of the left/right eye, respectively, and determine whether the 3 points of the eye/nose are closer to the 2 points of the ears at either end. It detects abnormal facial behavior patterns by detecting head movement, and if the facial outline is not recognized during face recognition, if the test taker moves away from the test screen on the test taker's terminal, the 3 points of the eyes/nose are close to the 2 points of the ears at both ends. If the distance in the corresponding direction (eye-to-ear distance or nose-to-ear distance) exceeds a certain standard, a photo of the face or the candidate's voice data related to cheating is transmitted to the online test server and output to the proctor's terminal, and the online test Transmitting a warning message or alarm from the server to the candidate's terminal; And it provides a test program (App) and online test papers to the candidate terminal and the proctor terminal, and stores and manages the candidate information, the candidate's on-site face photo, and the proctor information in the database of the online test server, and manages them during the online test. A step of storing the answers written on the test paper on each test taker's terminal during a certain test time, then transmitting and storing them from each test taker's terminal to the online test server at the end of the test, and providing the scoring results of the answers written on the test taker's test paper to the test taker's terminal. Includes,
The facial contour recognition technology used in the facial recognition module uses the posenet algorithm to measure the distance between the eyes and ears and the distance between the nose and ears.

The candidate terminal uses any one of a tablet PC, smartphone, or PC on which a test program (App) is installed, and the test program is installed, the facial recognition module, the outline of the face, and eyes 2/nose/ears. 2. Facial feature points on a 5-point scale Equipped with an anti-cheating module, when facial outlines are not recognized during face recognition, when the test taker's terminal is out of the test screen, the distance between the eyes and ears, and the distance between the nose and ears in the corresponding direction. If it exceeds a certain standard, it is judged to be an act of fraud.

delete

The online system and method using facial contour recognition artificial intelligence to prevent cheating of the present invention is a tablet PC and smart phone equipped with an artificial intelligence facial recognition module and a 5-point scale cheat prevention module for eye2/nose/ear2 facial feature points. In phone and PC-based online tests and UBT tests, face recognition and facial behavior patterns are detected in non-face-to-face tests to prevent proxy testing or cheating.

A proxy test prevention and cheating prevention technique during the test was implemented in the test taker's terminal, combining an artificial intelligence facial recognition module and a cheating prevention module that utilizes the outline of the face and eye2/nose/ear2 facial feature points on a 5-point scale.

The AI-based facial contour recognition technology used in the facial recognition module of the test taker's terminal and the 5-point scale of eye2/nose/ear2 facial features and anti-cheating module uses the posenet algorithm to recognize the facial contour of the camera image of the tablet PC. . When taking an online test or UBT test, AI-based facial recognition technology is used to verify the tablet PC candidate's identity through a proctor terminal linked to the test server and prevent proxy testing. It has the effect of preventing proxy testing or cheating.

1 is a configuration diagram of an existing face recognition device.
Figure 2 is a configuration diagram of an online evaluation system including a conventional online test cheating detection system.
Figure 3a is a diagram showing the concept of a facial contour recognition artificial intelligence platform equipped with a tablet PC, smartphone, and PC-based test online test/UBT test server during language and health education online tests/UBT tests.
Figure 3b is a diagram showing the face recognition function on the Ubiquitous Based Learning (UBL) and Ubiquitous Based Testing (UBT) NSDAI platform.
Figures 3c and 3d show the goals and User Experience (1. User registration -> 2. Learning -> 3. QR code) of the UBI cloud app and NS face platform used on tablet PCs, smartphones, and PCs during online exams/UBT exams. Creation -> 4. UBT App/Web login (QR code/passwd, ID/passwd) -> 5. Supervisor confirmation and test -> 6.7 UBT App/Web camera facial contour recognition 8. End of test) am.
Figure 4 shows the process of the test taker department from candidate registration, 1) candidate registration, 2) learning, 3) issuance of a QR code attached to the test taker's test ticket, 4) QR code and face recognition, 5) supervisor confirmation (candidate's face/candidate's face) information), 6) Taking the test - This is a picture showing the process.
Figure 5 is a screen of the candidate-face matching code issuance unit - candidate registration/learning/QR code issuance during the online test/UBT test based on the machine learning and learning results of the registered candidate's face.
Figure 6 is a screen of the test taker confirmation section in the UBT system that measures similarity through AI-based facial outline recognition of the test taker's face.
Figure 7 shows the rule user definition section and program test screen when judging candidates.
Figure 8 shows the test server connection/login during an online test or UBT test/left-QR code recognition/greeting heard when QR code is recognized (TTS)/right-face recognition (smartphone/tablet PC front camera)/face in the background This is a demonstration screen that includes the area recognition start/recognition rate display process - the UBT authentication process customization section utilizing the generated user code.
Figure 9 is a diagram illustrating the configuration of a tablet PC, smartphone, and PC-based online test system using facial contour recognition artificial intelligence to prevent cheating according to the present invention.
Figure 10 is a diagram explaining the functions of the AI-based facial recognition module and cheating prevention module during an online test or UBT test.
Figure 11 is a flowchart showing a method of providing an online exam based on a tablet PC, smartphone, and PC using facial contour recognition artificial intelligence to prevent cheating according to the present invention.

Hereinafter, the configuration and operation of the preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the present invention, if it is determined that a detailed description of a related known technology or configuration may unnecessarily obscure the gist of the present invention, the detailed description will be omitted. In addition, the attached drawing numbers are assigned the same drawing numbers in other drawings when indicating the same configuration.

Figure 9 is a diagram illustrating the configuration of a tablet PC, smartphone, and PC-based online test system using facial contour recognition artificial intelligence to prevent cheating according to the present invention.

Figure 10 is a diagram explaining the functions of the AI-based facial recognition module and cheating prevention module during an online test or UBT test.

The UBT system and method using facial contour recognition artificial intelligence of the present invention is a tablet PC and smartphone equipped with an artificial intelligence facial recognition module and a 5-point scale fraud prevention module for facial contour and eye2/nose/ear2 facial feature points. , In PC-based online exams and UBT exams, a front-facing camera is installed on the test taker's terminal using AI-based facial contour recognition technology (posenet algorithm, machine learning model which allows for real-time face pose estimation) to prevent proxy testing or cheating. It extracts the object of the test taker's face that is photographed, extracts the facial outline and facial feature points of eyes 2, nose, and ears 2, and extracts facial behavior patterns in real time to prevent cheating. The anti-fraud module uses an artificial intelligence facial recognition module and eye 2 that detect and track head movement to the right/left according to whether the three eye/nose points on the face get closer to the two ear points at either end, and detect the behavioral pattern of the face. /Nose/Ears2 A 5-point scale of facial features is equipped with an anti-cheating module, and if the facial outline is not recognized during face recognition through FACE RECOGNITION/FACE MOTION RECOGNITION/RESULT ANLAYSIS, a tablet PC camera video is recorded. If the distance in the relevant direction of the eye/ear distance or nose-ear distance exceeds a certain value depending on whether the 3 points of the eyes/nose are close to the 2 points of the ears at either end, the test taker's terminal will be sent to prevent cheating. A warning message or alarm is output, or answers written on the test paper are stored on the test taker's terminal and sent to the test server at the end of the test, and the test server provides scoring results to the test takers.

A proxy test prevention and cheating prevention technique was implemented in the test taker's terminal that combines an artificial intelligence facial recognition module and a cheating prevention module using the facial outline and eye2/nose/ear2 facial feature points on a 5-point scale.

The AI-based facial contour recognition technology used in the facial recognition module of the test taker's terminal and the face outline and eye2/nose/ear2 facial feature point 5-point scale in the anti-cheating module uses posenet to recognize the facial contour of the camera image of the tablet PC. An algorithm was used. When taking an online test or UBT test, AI-based facial recognition technology is used to verify the tablet PC candidate's identity through a proctor terminal linked to the test server and prevent proxy testing. In the online test or UBT test, the candidate's face is recognized through Prevent proxy testing or cheating.

The test paper questions in the test program (App) include not only text and images for each question, but also multimedia test questions that include text, images, VR/AR content, voice, and video, and are displayed on the test taker's terminal through the test server.

1. Artificial Intelligence Hub Platform

ㆍArtificial intelligence hub platform based on NSD-AI (NSDevil's Artificial intelligence) platform

ㆍ Instructor grading guide through phrase recognition in medical education (joint research with Inje University School of Medicine and others)

ㆍ Experiential learning support system through object recognition (Fun Edu Lab, Korean Folk Village, etc.)

ㆍ Disease analysis function in the dental education sector through tooth image analysis (joint research by Yonsei University College of Dentistry and 3 other countries)

ㆍ Based on UBT connect platform

ㆍTablet PC-based evaluation platform

ㆍ Language assessment section (speaking/listening/reading/writing)

ㆍHealth and medical evaluation sector (listening/reading/writing)

ㆍOther job evaluation categories (reading/writing)

The tablet PC, smartphone, and PC-based online test system using artificial intelligence facial contour recognition uses an online test server (100), a proctor terminal (200), and a PC and smartphone or tablet PC. It includes test taker terminals (300, 310, 311) equipped with a cheat prevention module that utilizes the outline and eyes2/nose/ear2 facial features on a 5-point scale.

The online test system using facial contour recognition artificial intelligence to prevent cheating of the present invention is

Test taker information and frontal face photos are registered and stored in the test server, a QR code attached to the test taker's test ticket corresponding to the test taker information is issued, and the QR code attached to each test taker's test ticket is recognized and TTS conversion technology is used. An online test or UBT test server 100 that provides greetings and provides online test papers and a test program for conducting an online test or UBT test;

It is connected to the online test or UBT test server 100 through a wired or wireless communication network (LAN, Wi-Fi, LTE 4G/5G), and includes an artificial intelligence facial recognition module of the face captured through the front camera of the candidate's terminal, and a front camera. A test taker terminal (300, 310, 311) equipped with a cheat prevention module that recognizes the facial behavior pattern captured by the facial outline and facial feature points of eyes 2, nose, and ears 2 and a 5-point scale; and

It is connected to the online test or UBT test server 100 through a wired or wireless communication network, and during the online test or UBT test, facial photos and cheating information of one or more test taker terminals are provided through the online test or UBT test server 100. Includes a supervisor terminal 200 that receives and transmits an anti-cheating alarm or warning message to the candidate's terminal,
The online or UBT test server 100 further includes a greeting service unit (TTS) that provides a greeting through TTS conversion technology after recognizing the QR code attached to the test taker's test ticket,

The candidate terminals (300, 310, 311) use any one of a tablet PC, smartphone, or PC on which a test program (App) is installed, and the test program is installed, the artificial intelligence facial recognition module, the outline of the face, and the eyes2 Equipped with a 5-point scale fraud prevention module for the facial feature points of /nose/ear2, and extracting and classifying facial features to identify the left/right ear and left/right eye of the facial feature points of eye2/nose/ear2, respectively. Calculate the Euclidean distance (d) and similarity with the center point (pupil), and compare it with the data in the standard-sized frontal facial photo DB of the online test server to check whether it is a proxy test. When taking an online test, By recognizing the facial behavior pattern captured by the camera of the test taker's terminal, if the facial outline is not recognized during face recognition, if it deviates from the test screen, the distance between the right eye and the right ear, and the distance between the left eye and the left eye, depending on the angle at which the front face is turned to the left or right, Since the distance between the ears varies, if the distance in the corresponding direction of the eye-to-ear distance or the nose-to-ear distance exceeds a certain standard, it is judged to be cheating.

The AI-based facial contour recognition technology used in the facial recognition module uses the posenet algorithm to measure the distance between the eyes and ears and the distance between the nose and ears.
If the facial behavior pattern captured by the camera of the test taker's terminal is recognized and judged to be cheating, it is transmitted to the online test server, and after confirmation by the proctor's terminal linked to the online test server, an alarm or warning message is sent to the test taker's terminal. occurs.

The candidate terminals (300, 310, and 311) use any one of a tablet PC, smartphone, or PC on which a test program (App) is installed, and the test program is installed, an artificial intelligence facial recognition module, and the outline and eyes of the face2/ Nose/Ears2 Equipped with a 5-point scale for facial features and an anti-cheating module.

The test taker's terminal is equipped with a facial recognition module and a cheating prevention module. The cheating prevention module detects visual cheating in the camera vision image data of the test taker's terminal in real time, including the face outline and eyes 2/nose/ear 2. It is equipped with a 5-point scale fraud prevention module and a voice recognition module that detects auditory fraud.

The anti-cheating module is equipped with a voice recognition module in the test taker's terminal that recognizes the test taker's speech sounds in order to prevent auditory cheating by the test taker, and detects the test taker's speech sounds related to cheating during an online test or UBT test. Immediately transmit and store the cheating face photo and sound waveform data to the online test or UBT test server (100), and transmit them to the proctor terminal (200) to prevent the proctor from cheating through the test taker's terminal (100). It generates an alarm or sends a warning message to the test taker's terminal, thereby preventing cheating.

The AI-based facial contour recognition technology used in the facial recognition module uses a posenet algorithm (machine learning model which allows for real-time face pose estimation).

In the embodiment, during the online test or UBT test, a tablet PC was used as the test taker's terminal.

The test taker terminals (300, 310, and 311) are equipped with a facial recognition module, eyes2/nose/ear2 facial feature point 5-point scale fraud prevention module, and voice recognition module.

The candidate terminals (300, 310, 311) detect the ROI (Region of Interest) of the facial image captured in real time with the front camera of the candidate terminal during an online test or UBT test, and use a standard size frontal face with size/rotation/angle correction. Facial objects are extracted from photos using a face recognition algorithm, facial behavior patterns are recognized (pattern recognition), the outline of the applicant's face and facial feature points of eyes 2, nose, and ears 2 are extracted, and feature extraction and classification are performed. Calculate the Euclidean distance (d) and similarity of the facial feature points of Eye2/Nose/Ear2 with the center point (pupil) of the left/right ear and left/right eye, respectively, and perform an online test or UBT. Comparison with the facial photo DB of the test server 100 confirms whether a proxy test is performed, and facial behavior patterns captured by the camera of the candidate's terminal are recognized during an online test or UBT test, and the right eye is adjusted according to the angle at which the front face is turned left and right. Since the distance between the right ear and the left eye and the left ear are different, cheating is determined and transmitted to the test server 100, and confirmed by a proctor terminal linked to the test server 100, and the test server 100 An alarm or warning message is generated through the test taker's terminal.

The candidate terminals (300, 310, 311) use AI-based facial outline recognition technology (posenet algorithm) on camera images through a facial recognition module and a 5-point scale fraud prevention module for face outline and eye2/nose/ear2 facial feature points. Detects head movement to the right/left depending on whether the 3 points of the eyes/nose are close to the 2 points of the ears at either end, detects abnormal facial behavior patterns related to cheating, and detects facial contours when recognition of the face is not possible. If the test screen deviates from the test screen (if the eye-to-ear distance or nose-to-ear distance in the corresponding direction exceeds a certain standard depending on whether the 3 points of the eye/nose are close to the 2 points of the ears at either end), the test taker's terminal is used to prevent cheating. A warning message or alarm is output.

(Example)

We used TensorFlow to estimate the pose of a human face on Android on a tablet PC using the PoseNet model. For reference, the PoseNet algorithm is a vision model that detects the positions of major human body parts and estimates a person's pose from images or videos. TensorFlow Lite provides an Android sample application that detects and displays major human body parts in real time using the camera of the candidate's device. This model can estimate the position of a person's elbow and/or knee in an image, and in the embodiment, a facial outline pose estimation model was implemented by identifying eyes2/nose/ear2 in real time when recognizing a face.

The posenet algorithm used in the artificial intelligence facial recognition module uses a 5-point scale fraud prevention module for left eye (leftEye), right eye (rightEye), nose, left ear (leftEar), and right ear (rightEar). Implemented,

After finding the x,y coordinates of leftEar and the x,y coordinates of rightEar, you can find the diameter of the circle based on this and create an angel ring with ellipse(). The diameter can be easily obtained using dist() in p5js without having to use the Pythagorean theorem.

<Connecting ml5js and webcam>

<script src="https://unpkg.com/ml5@0.3.1/dist/ml5.min.js"></script>

let video;

let poseNet;

function setup() {

createCanvas(400, 400);

video = createCapture(VIDEO);

//Use to prevent the video from appearing twice

video.hide();

poseNet = ml5.poseNet(video,modelLoaded);

poseNet.on('pose',gotPoses);

console.log(ml5);

}

// Pose-related callbacks are entered

function gotPoses(poses) {

}

// To check if loading was successful

function modelLoaded() {

console.log('Model Loaded');

}

function draw() {

//Draw the webcam at position 0,0.

image(video,0,0);

}

The face area is extracted in real time from the front camera face image of the test taker's terminal, adjusted to the standard size of the frontal face photo through size correction/rotation/angle correction, and this is stored in the standard size frontal face photo database previously stored in the test server (100). Comparison with learning data -> Learning model of face photo -> Extract and classify face outline and facial feature points of eyes 2/nose/ear 2, and facial outline of face recognition data, eyes 2/nose/ear 2 Group clustering of feature points (cluster center k-means algorithm) density estimation, and the distance (Euclidean distance) and similarity of facial feature points are calculated to determine the standard stored in the face photo DB of the online test or UBT test server. Proxy testing is prevented by comparing the Euclidean distance and similarity of the facial feature points of the applicant's face photo to confirm that the applicant's photo is the applicant's photo, and cheating is judged during the online test or UBT test.

For example, when the direction of the face captured by the front camera of the test taker's terminal (tablet PC) changes or the test taker turns his or her face, the outline of the face in the face photo and the face of eyes 2/nose/ears 2 are displayed. If the characteristic points are not recognized or the angle facing the test program changes within the effective distance (20~30cm) of the user's face and the smartphone/tablet PC, it is recognized as cheating, and the cheating detection result is sent to the online test or It is transmitted to the UBT test server (100), transmitted to the proctor terminal (200) through the test server (100), and the proctor sends an alarm to the candidate's terminal through the test server (100) to prevent the candidate's terminal from cheating. generates or sends a warning message, thereby preventing fraudulent activity.

Alternatively, [Text A] data reference. The moment the warning threshold set in the predefined test taker program becomes 0, a warning message can be displayed on the test taker's terminal without a server connection and transmitted back to the proctor/cheating detection server/other test server. Or, after the test is over, the RAW DATA is sent to the test server/other test management server that detects cheating for post-processing.

In addition, the candidate terminal is equipped with a facial recognition module and a cheating prevention module, and the cheating prevention module further includes a voice recognition module that recognizes the candidate's speech sound to prevent auditory cheating of the candidate, During an online test or UBT test, if the candidate's speech related to cheating input through the microphone of the test taker's terminal is heard, the voice is recognized and immediately transmitted to the online test or UBT test server (100) and stored, and the test server (100) This is transmitted to the proctor's terminal (200) and the proctor generates an alarm or sends a warning message to the test taker's terminal through the test server (100) to prevent the test taker's terminal from cheating. do. Alternatively, the program of the test taker's terminal may display an alarm or warning message directly on the test taker's terminal without a command from the test server 100, referring to the predefined distance thresholds of the distance between the test taker's eyes and ears and the distance between the nose and ears. and transmitted to the proctor terminal 200, a cheating detection server, or other test management server.

The online test or UBT test server 100 includes a WWW server 101, a control unit 102, a member register 103, a user authentication unit 104, a test notification unit 105, a QR code management unit 106, and a greeting system. Study (TTS) (107), supervisor confirmation unit (108), face recognition unit (109), cheating processing unit (110), test management unit (111), test information DB (120), candidate DB (121), and face Includes photo DB (123).

The online test or UBT test server 100 is

A WWW server 101 connected to a test taker terminal and a supervisor terminal using any one of a tablet PC, smartphone, and PC through wired and wireless communication;

A control unit 102 that controls the provision of online or ubiquitous-based learning (UBL) and ubiquitous-based testing (UBT) application services to test taker terminals and proctor terminals of tablet PCs, smartphones, and PCs through a wired or wireless communication network;

A member registration unit (103) connected to the control unit (102), which receives the candidate's membership information and stores and manages the ID/Passwd;

A user authentication unit 104 connected to the control unit 102 and authenticating the user using QR code/Passwd or ID/Passwd;

A test notice unit ( 105);

A QR code management unit 106 connected to the control unit 102 and issuing QR codes attached to test takers corresponding to test taker information to test takers and managing the QR codes;

A greeting service unit (TTS) 107 that provides a greeting through TTS conversion technology after recognizing the QR code attached to the test taker's application form;

It is connected to the control unit 102, and receives the candidate's frontal face photo from the camera from the test taker terminal, generates a standard size frontal face photo corrected for size/rotation/angle to the standard size of the frontal face photo, and tests it. The proctor compares the candidate information previously stored in the database of the server 100 with the frontal face photo of a standard size and checks the proctor at the proctor terminal 200 to check whether the test is conducted by proxy. Proctor verification unit (108), which verifies the candidate's eligibility to take the test;

It is connected to the control unit 102, and uses a facial recognition module in the test taker's terminal to recognize the outline of the test taker's face and the characteristic points of the eye 2/nose/ear 2 face, and receives the face recognition result from the test taker terminal. Wealth (109);

It is connected to the control unit 102, and the facial recognition module and anti-cheating module in the test taker's terminal use facial outline recognition technology (posenet algorithm) on the front camera image so that 3 points of eyes/nose are close to 2 points of ears at both ends. It detects head movement to the right/left depending on whether the face is lost, detects abnormal facial behavior patterns related to cheating by looking at cheat sheets, and detects facial contours that are not recognized when recognizing a face or moving away from the test screen (eyes/nose 3 Depending on whether the dot approaches the two end points of the ears, if the distance in the corresponding direction of the eye-to-ear distance or nose-to-ear distance exceeds a certain standard value, a warning message is sent from the test server 100 to the test taker's terminal to prevent cheating. A cheating processing unit 110 that outputs an alarm or, conversely, receives a warning message or alarm from the test taker's terminal to the test server 100;

It is connected to the control unit 102, provides a test program (App) and test papers to the candidate terminal and the proctor terminal, manages candidate information, on-site facial photos of the test taker, and proctor information, and manages the test taker information and the test taker's on-site face photo and proctor information during the online test or UBT test. The test paper answers are stored on each test taker's terminal within a certain test time and are sent to the test server at the end of the test. The test management department (111) stores and manages the test takers' test paper answers, scoring results, supervisor information, and test taker status information. ;

Test information DB (120) that stores test takers' test papers, written answers, and scoring results; It includes a candidate DB (121) and a face DB (123) that store candidate information and standard-sized frontal face photos.

Candidate terminals include not only smartphones and tablet PCs for online tests (Internet Based Test (IBT)) or ubiquitous-based tests (UBT), but also mobile communication terminals with Internet access, or laptops.

The test taker's terminal uses any one of tablet PCs, smartphones, and PCs on which the test program (App) is installed. A 5-point scale of facial features is provided with a cheat prevention module, and an online test or test program (App) downloaded from the UBT test server (100) is installed.

The AI-based facial outline recognition technology used in the facial recognition module uses the posenet algorithm.

The test taker's terminal is equipped with a facial recognition module and a cheating prevention module. The cheating prevention module detects visual cheating in the real-time visual camera vision image of the test taker's terminal, including facial contours and eye2/nose/ear2 face. It is equipped with a 5-point scale fraud prevention module and a voice recognition module that detects auditory fraud.

The online test or UBT test server (100) provides university exams such as medical/dental/pharmacy/engineering colleges, TOEIC/TOEFL, language tests, civil service exams, certification exams, and health and medical education study and test papers. The bank's various official certification tests or non-official tests are notified of the test schedule and location, and the test program in the database of the test server 100 is used to provide test questions to candidate terminals through a wired or wireless communication network, and test questions stored in online tests or offline. We provide a UBT test using any type of PC/smart device (tablet PC) that can be used to take the test.

Once the test schedule and location are confirmed, the control unit 102 appoints a proctor, provides the proctor with information on the proctor appointment, announces the exam schedule and location to test takers via text message/web page, and appoints a proctor at the test location on the day of the exam. The test is progressed and managed by transmitting and receiving test information and test paper information to the terminal 200, and the scoring results from the results of the automatic scoring result book 124 and the inspection management department 125 are provided to the relevant candidate terminals.

The test information stored in the test information database 120 includes test title, test schedule and location, test time, location information of the test location, information on the proctor for each assigned exam room, list of candidates, number of seats for each exam room, proctor information, A/ B Includes question information, answer information, scoring information, and supervisor information for each type of test paper.

And, the candidate database 121 contains candidate information including name, resident registration number, home address, mobile phone number, email, etc., candidate photo (standard-size frontal face photo), QR code/passwd, ID/passwd identification information, and test Relevant information and candidate application status are stored.

Figure 11 is a flowchart showing a method of providing an online exam based on a tablet PC, smartphone, and PC using facial contour recognition artificial intelligence to prevent cheating according to the present invention.

The method of providing online exams based on tablet PCs, smartphones, and PCs using facial contour recognition artificial intelligence to prevent cheating according to the present invention is

A step (S10) of registering member information and, after logging in/user authentication, the online test or UBT test server 100 registering and storing the candidate information and a standard-sized frontal face photo;

A step of issuing a QR code attached to the test taker's test ticket corresponding to the test taker information and frontal face photo (S20);

The online test or UBT test server 100 announces the test schedule and location according to the subjective test (S30);

It is activated by scanning the QR code attached to the test taker's test ticket with a smartphone and placing it on the recognition screen of the test taker's terminal (PC, tablet PC, etc.). After the QR code is recognized, a greeting is provided through the TTS (Text To Speech) function. Step (S40);

A candidate terminal equipped with a face recognition module using AI-based facial outline recognition technology (posenet algorithm) uses a frontal face recognition algorithm to transmit the recognition results of the camera's frontal face photo to the online test or UBT test server (100). , a standard size frontal face photo with size correction/rotation/angle correction is generated, and the proctor terminal 200 compares the applicant information with the standard size frontal face photo and its characteristic points to confirm the applicant's identity and conducts a proxy test. preventing and deciding whether to take the test (S50);

Facial behavior patterns captured by the front camera of the candidate's terminal are recognized by the facial recognition module of the test taker's device to extract facial objects, and eyes 2/nose of a standard-sized frontal face photo created by size correction/rotation/angle correction. /Extracts the facial feature points of Ear2, recognizes the outline of the candidate's face in a standard size and the facial feature points on a 5-point scale of Eye2/Nose/Ear2, and recognizes each of the facial feature points of Eye2/Nose/Ear2. Calculate the Euclidean distance (d) and similarity between the center point (pupil) of the left/right ear and the left/right eye, and move to the right/left depending on whether the 3 points of the eye/nose are closer to the 2 points of the ears at either end. By detecting head movement, abnormal facial behavior patterns related to cheating are detected, and if the facial outline is not recognized during face recognition or the test screen of the test taker's terminal is off (the 3 points of the eyes/nose are close to the 2 points of the ears at both ends), (If the distance in the corresponding direction (eye-to-ear distance or nose-to-ear distance exceeds a certain threshold), facial photos related to visual cheating and/or voice data of the test taker related to auditory cheating to prevent cheating) Transmitting to the online test or UBT test server 100, output to the proctor terminal through the test server 100, and receiving a warning message or alarm from the test server 100 to the candidate's terminal (S60) );

A test program (App) and online test papers are provided to the candidate's terminal and the proctor's terminal, and the candidate's information, the candidate's on-site face photo, and the proctor's information are stored and managed in the database of the test server (100), and are managed during the online test or UBT test. The answers written on the test paper are stored in each test taker's terminal during a certain test time, and at the end of the test, they are transmitted from the test taker's terminal to the test server 100 and stored (S70), and the test server 100 records the grading results of the answers written on the test paper by the test takers. It includes a step (S80) of providing to the candidate's terminal,
The facial contour recognition technology used in the facial recognition module uses the posenet algorithm to measure the distance between eyes and ears and the distance between nose and ears,

The candidate terminal uses any one of tablet PCs, smartphones, and PCs on which the test program (App) is installed. The test program is installed, and the artificial intelligence facial recognition module and the outline of the face and eyes 2/nose are used. / Equipped with a cheat prevention module on a 5-point scale for the facial features of Ear 2, if the facial outline is not recognized during face recognition, if it deviates from the test screen of the test taker's terminal, the distance between the eyes and ears, and the direction of the distance between the nose and ears If the distance exceeds a certain standard, it is judged as cheating.

delete

The candidate's terminal uses artificial intelligence (AI)-based facial outline recognition technology to use the posenet algorithm to recognize the facial outline of the tablet PC's camera image. During an online test or UBT test, AI-based facial recognition technology is used to verify the tablet PC candidate's identity through a proctor terminal linked to the test server (100), prevent proxy testing, and identify the test taker's face at the online or UBT test site. It can be used as a facial recognition access control system and prevents proxy testing or cheating in online exams and UBT exams.

The candidate terminal uses any one of a tablet PC, smartphone, or PC on which a test program (App) is installed, and the test program is installed, the artificial intelligence facial recognition module, the outline of the face, and eyes 2/nose/ It is equipped with a fraud prevention module that measures the facial features of ear 2 on a 5-point scale.

The AI-based facial outline recognition technology used in the facial recognition module uses the posenet algorithm.

The test taker terminals (300, 310, and 311) are equipped with a facial recognition module and a cheat prevention module with a 5-point scale for facial outlines and eye2/nose/ear2 facial feature points,

During an online test or UBT test, the ROI of a face image captured in real time with the front camera of the candidate's terminal is detected and a facial behavior pattern is used on a standard size frontal face photo that has been size corrected/rotated/angle corrected using a frontal face recognition algorithm. By recognizing the outline of the applicant's front face and the facial feature points of eyes 2/nose/ear 2, the facial feature points of eyes 2/nose/ear 2 are extracted, and through facial feature extraction and classification, the left/right side of the facial feature points of eye 2/nose/ear 2 is determined on a 5-point scale. Calculate the Euclidean distance (d) and similarity between the right ear and the center point (pupil) of the left/right eye, and calculate this as standard-sized face data from the online test or the frontal facial photo DB of the UBT test server (100). It checks whether the test is a proxy by comparing it with the online test or UBT test. Since the distance between the right eye and right ear and the distance between the left eye and left ear vary depending on the angle at which the front face is turned left and right during the online test or UBT test, cheating is judged and the test server It is transmitted to (100), the proctor terminal (200) linked to the test server (100) confirms it, and an alarm or warning message is sent to the corresponding candidate's terminal to be output.

Candidate terminals (300, 310, 311) are connected to the online test or UBT test server (100) through a wired or wireless communication network (LAN, Wi-Fi, LTE 4G/5G), and the artificial intelligence facial recognition module and the facial outline that constitutes the facial feature points It is equipped with a cheat prevention module on a 5-point scale for the facial features of eyes2/nose/ears2, and ROI (Region of Interest) of facial images captured in real time with the front camera of the test taker's terminal during online or UBT tests. Detects and creates a standard-size frontal face photo with size/rotation/angle correction based on the vertex of the nose, and facial objects corrected to standard size, facial outlines, and facial feature points of eyes 2/nose/ears 2. Extract and obtain the facial outline and feature vector Calculate the Euclidean distance (d) and similarity of feature points and compare them with data from an online test or a standard-sized frontal facial photo DB stored in the UBT test server 100 to confirm whether a surrogate test is performed, During an online test or UBT test, cheating is determined based on the angle at which the front face captured by the camera of the candidate's terminal is turned left and right, and the cheating information of the face photo is transmitted to the test server 100, and the proctor's terminal detects the cheating. When the information is confirmed, an anti-cheating alarm or warning message is output from the test server 100 to the test taker's terminal.

The candidate terminal uses facial outline recognition technology (posenet algorithm) for camera images through a facial recognition module and an anti-fraud module to move the head to the right/left depending on whether the 3 points of the eyes/nose are close to the 2 points of the ears at either end. Detects abnormal behavior patterns on the face related to cheating, and detects facial contours that are not recognized during face recognition and when the face deviates from the test screen (eyes/nose 3 points are close to the 2 ears at both ends) If the distance in the relevant direction (distance between nose and ear, or distance between nose and ear) exceeds a certain standard (numerical value), a warning message or alarm is output to the test taker's terminal to prevent cheating.

The method operates by scanning the QR code attached to the test taker's test taker with a smartphone and placing it on the recognition screen of the test taker's terminal camera. After recognizing the QR code, it further includes the step of providing a greeting through the TTS (Text To Speech) function. Includes.

The test paper questions in the test program include not only text and images for each question, but also multimedia test questions that include text, images, VR/AR content, voice, and video, and are displayed on the test taker's terminal through the test server.

The method of providing online tests using facial contour recognition artificial intelligence is:

The candidate terminal uses any one of a tablet PC, smartphone, or PC on which a test program (App) is installed. The test program is installed, and the artificial intelligence facial recognition module and the outline of the face and eyes 2/nose are used. / Equipped with a fraud prevention module based on a 5-point scale of facial features of ear 2.

The candidate terminal is equipped with the facial recognition module and the cheating prevention module, and the cheating prevention module detects visual cheating in the camera vision image data of the candidate terminal in real time, such as the facial outline and eyes2/nose/ Ear 2 is equipped with a 5-point facial feature point scale to prevent fraud and a voice recognition module to detect auditory fraud.

The voice recognition module recognizes the candidate's speech sound in order to prevent auditory cheating by the candidate in the test taker's terminal, and when the test taker's voice related to cheating is heard during an online test or UBT test, it turns on the microphone of the test taker's terminal. The candidate's voice data entered through the test is immediately transmitted to the online test or UBT test server (100) and stored, and it is transmitted to the proctor's terminal (200) linked to the test server (100) so that the proctor can check the candidate's terminal for cheating. To prevent this, the step of generating an alarm or transmitting a warning message to the test taker's terminal through the test server 100 is further included.

The candidate's terminal detects the ROI of the face image captured in real time with the front camera of the candidate's terminal during a non-face-to-face online test or UBT test, and uses a face recognition algorithm for a standard size frontal face photo that has been size-corrected/rotated/angle-corrected. By extracting facial objects and recognizing facial behavior patterns, the outline of the applicant's face and facial feature points of eyes 2/nose/ear 2 are extracted, and the facial outlines and eye 2/nose are extracted through facial feature extraction and classification. / Calculate the Euclidean distance (d) and similarity between the facial feature points of Ear 2 and the center point (pupil) of the left/right ear and left/right eye, respectively, and calculate these in the online test or face photo on the UBT test server. We check whether the test is a proxy test by comparing it with the standard size frontal face data in the DB, and by recognizing the facial behavior pattern captured by the camera of the test taker's terminal during the online test or UBT test. If the facial outline is not recognized during face recognition, the test taker When leaving the test screen of the terminal, the distance between the right eye and right ear and the distance between the left eye and left ear vary depending on the angle at which the front face is turned left and right, so a fraudulent act is determined and transmitted to the test server 100, and the test is performed. The proctor terminal 200 linked to the server 100 confirms this and generates an alarm or warning message to the candidate's terminal through the test server 100.

In addition, the method stores the user's facial movements in the camera image of the candidate's terminal in frame units as shown in [Text A], subtracts a defined frame, time, or number of times, and sends a warning message to the candidate's terminal when it reaches 0. After expressing or transmitting the proctor device or fraud detection image or sound data to the test server 100, the step includes initializing the number of warning triggers and starting the deduction again from the next movement.

[Text A] - Facial behavior pattern information stored on the tablet PC of the test taker's terminal (e.g. facial pattern data of a test taker at Seoul National University Dental School): Information such as date, time, angle/distance of movement, and phase (left and right) It is saved every frame or hourly, and based on this, a warning message is output to the test taker's terminal, and the supervisor confirms and detects the test taker's face recognition cheating - RAW data is sent to the system of the online test or UBT test server and the main online test or UBT server. Send and post-process.

200519094106,0.02,R

200519094111,0.03,R

200519094120,0.42,R

200519094148,0.01,R

200519094150,0.21,R

200519094155,0.22,R

200519094158,0.09,R

200519094209,0.11,L

200519094214,0.09,R

200519094216,0.10,R

200519094219,0.09,R

200519094224,0.01,L

In the embodiment, when the test taker's terminal connects to the test server 100 and applies for the UBT test online, the test application reception process includes sending the test taker's test application information to the test server 100 to register the test taker; Once the exam schedule, time, and location are determined, an exam information announcement process that announces exam information including the exam schedule, time, and location information to test takers; Once the test location and time are set, a proctor is set for each test room at the set date and time of the test, and proctor appointment information is provided to the proctor, informing the proctor of the appointment. Appointment confirmation information is received from the proctor to the test server to complete the proctor setting process. Complete the proctor setup process; On the day of the test, wait for the test site setting information to be input from the test site supervisor, and when the test site setting information is input from the test taker, the test program (app) is provided to the test taker's terminal and can be used on the test taker's smartphone/tablet PC (administrator). A test site setting process that requires installation on a test taker (e.g., tablet PC) and requests test access confirmation from test takers connected through the test program installed on the test taker's smartphone/tablet PC; Once the test location setting is completed, refer to the set test question type and question selection setting information (difficulty level) to select one of the registered and saved list of questions to be administered, and then take the test by communicating with the test program on the installed test taker terminals. A test progress process in which questions are provided, and the test taker's terminal receives the answer information of the test paper to the test server through the test program and stores it for each test taker; A test completion process where all test questions are presented and the test is completed, waiting for the proctor to enter termination confirmation information, and automatically deleting the applications installed on the test taker's tablet PC when the proctor's termination confirmation information is entered; The exam presentation process consists of comparing each test taker's written answers received and stored on the test server with the scoring answer information and grading them, and announcing the grading result information.

The appointment of a supervisor is selected from pre-registered supervisors, and these supervisors are pre-contracted supervisors and are registered and managed in the test information database of the test server.

Once a supervisor is appointed, the supervisor appointment information is provided through the test server, and appointment confirmation information is received from the supervisor to complete the supervisor setup process.

Proctor appointment information includes test location, test time, and number of test takers.

Candidate information includes a full face photo of each candidate, personal information including name, gender, address, and contact information.

The test setting process is a process to check the parts required for the test and set up the equipment to conduct the test before the pre-set preparation time on the test day.

When a supervisor checks the applicant's information and face photo and enters the test location setting information, which is confirmation information, the test server 100 provides the test program (App) required for the test to the test taker's terminal. This is a process in which test takers install the test program on their terminal (tablet PC), send and receive data through the test program, receive the identification information entered by the test takers when entering application information, and complete test access confirmation.

The proctor's test setting information includes information on the number of test takers participating in the test, and the test takers' identification information consists of one of the personal information such as name and resident registration number or member ID entered by the test takers when entering application information.

The test program (App) consists of accessing the system of the test server 100 and entering identification information to download and install the test program on the test taker's terminal, or when the proctor enters test setting information in the system, the proctor's terminal may include a process of providing the directory path of the test program and installing the test program on the test taker's terminal according to the directory path.

Figure 3a is a diagram showing the concept of a facial contour recognition artificial intelligence platform equipped with a tablet PC, smartphone, and PC-based test online test/UBT test server during language and health education online tests/UBT tests.

Figure 3b is a diagram showing the face recognition function on the Ubiquitous Based Learning (UBL) and Ubiquitous Based Testing (UBT) NSDAI platform.

Figures 3c and 3d show the goals and User Experience (1. User registration -> 2. Learning -> 3. QR code) of the UBI cloud app and NS facere platform used on tablet PCs, smartphones, and PCs during online exams/UBT exams. Creation -> 4. UBT App/Web login (QR code/passwd, ID/passwd) -> 5. Supervisor confirmation and test -> 6.7 UBT App/Web camera recognition of facial contour 8. End of test) am.

Figure 4 shows the process of the test taker department from candidate registration, 1) candidate registration, 2) learning, 3) issuance of a QR code attached to the test taker's test ticket, 4) QR code and face recognition, 5) supervisor confirmation (candidate's face/candidate's face) information), 6) Taking the test - This is a picture showing the process.

Figure 5 is a screen of the candidate-face matching code issuance unit - candidate registration/learning/QR code issuance during the online test/UBT test based on the machine learning and learning results of the registered candidate's face.

This is the candidate registration/study/QR code issuance screen during the online test/UBT test.

Figure 6 is a screen of the test taker confirmation section in the UBT system that measures similarity through AI-based facial outline recognition of the test taker's face.

Figure 7 shows the rule user definition section and program test screen when judging candidates.

Figure 8 shows the test server connection/login during an online test or UBT test/left-QR code recognition/greeting heard when QR code is recognized (TTS)/right-face recognition (smartphone/tablet PC front camera)/face in the background This is a demonstration screen that includes the area recognition start/recognition rate display process - the UBT authentication process customization section utilizing the generated user code.

(1) The test taker's terminal is equipped with a proxy test prevention and test cheat prevention method that combines an artificial intelligence facial recognition module and a 5-point scale cheating prevention module for eye2/nose/ear2 facial features.

Artificial intelligence that has learned the candidate's registered photo generates a recognition code (QR code) for each user, and the test taker scans the test ticket with the corresponding recognition code (QR code) attached to the camera of the test site PC or tablet PC, and then recognizes the user. The AI-based facial recognition module compares the candidate information previously stored in the test server 100 with the result model learned from a standard-sized frontal face photo, calculates the distance (d) and similarity of feature points for facial recognition, and sets a constant If the number exceeds the number, it is recognized as you, and you move to the test screen and take the test. In the case of falling short of a certain level (completely falling short, judgment withheld), the information is transmitted to the proctor's terminal and the test is conducted after the proctor directly verifies personal information. During the test, the artificial intelligence of the facial recognition module or facial contour/eye2/nose/ear2 facial feature points on a 5-point scale and the anti-cheating module exist in the background or in another layer below the test screen or at the same coordinate value location as the test screen. The camera attached to or built into the test taker's terminal is used to recognize the user's face/voice until the end of the test. If an abnormal pattern is recognized, a warning is given to the test taker or the supervisor checks and takes action according to the designated pattern to stop the test. Alternatively, without providing any information to the test taker, the abnormal phenomenon and normal application status information are saved as images or text/numerical information on the test taker's terminal and sent to the test server after the test is completed, so that the test supervisor / test taker can determine whether the test is a proxy test or cheating. Assists test administrators in making decisions regarding misconduct. When a prior warning is given about a candidate's visual/auditory misconduct, it is provided in the form of a warning message or alarm to the terminal of the test proctor or test director through the test server 100, and can be notified in various ways, such as sending SMS. there is.

Online testing, or UBT testing, can prevent proxy testing, detect visual and auditory cheating, and reduce the number of proctors, dramatically lowering test costs while ensuring a minimum level of test reliability even in online assessments without proctors. .

<System configuration>

1) When using online: User program (PC/smart phone/tablet PC) > Wired and wireless communication network (WAN, LTE 4G/5G) and internal network (LAN, Wi-Fi) > Server (exam information/candidate information) and recognition information is exchanged to check the necessary information, and if correct, it returns True/if incorrect, False, and the user program executes the specified action according to the returned value. The applicant's recognition information and results are also transmitted to the proctor program (PC/smartphone/tablet PC) through the test server. (Sent via program or via SNS/SMS/email)

2) When used offline: The artificial intelligence learning module is installed in the user program (PC/smartphone/tablet PC), recognizes the facial contour of the candidate's face, and only the final facial recognition result is transmitted to the test server (100). When transmitting the final facial recognition result from the candidate's terminal to the test server (100), the information is displayed on the server program (dashboard) and transmitted to the proctor or test administrator terminal through the corresponding information program or via SNS/SMS/email. .

<Some APKs of actual app prototypes under development - attached for Android>

https://we.tl/t-wFdXexsors File download password: nsdevil

You can directly check some of the above functions by installing the app on your Android smartphone or tablet PC.

<Web version function address>

1) Face recognition module: https://facere.nsdai.org (id: nsdevil, passwd: nsdevil) -> Take a photo of the QR code attached to the test taker's test taker with a smartphone and place it on the PC/tablet PC camera for recognition. being.

2) Face contour and 5-point scale of eyes2/nose/ear2 Fraud prevention module - Facial contour recognition module: https://headpos.ublrandd.com.np (id: nsdevil, passwd: nsdevil), face contour and Eyes2/Nose/Ears2 5-point scale of facial features Anti-cheating module

In online tests and UBT tests based on tablet PCs, smartphones, and PCs equipped with an artificial intelligence facial recognition module, facial outline, eyes2/nose/ears2, and a 5-point scale for facial features, anti-cheating modules can be used to prevent proxy testing or fraud. To prevent this, AI-based facial contour recognition technology (posenet algorithm) is used to recognize the outline of the test taker's face and eye 2/nose/ear 2 feature points on the test taker's terminal, and the eye/nose 3 points are close to the ear 2 points at both ends. The facial behavior pattern is detected by detecting and tracking the movement of the head to the right/left depending on whether the face is recognized. If the facial outline is not recognized when recognizing the face, or if the camera image of the test taker's terminal (tablet PC) deviates from the test screen where it is recorded. (Depending on whether the 3 points of the eyes/nose are close to the 2 points of the ears at either end, if the distance between the eyes and ears or the distance in the corresponding direction of the distance between the nose and ears exceeds a certain value), a warning message or alarm is output to the test taker's terminal, or After storing the answers written in the test paper on the candidate's terminal, they are transmitted from each candidate's terminal to the test server (100) at the end of the test, and the test server (100) provides the scoring results to the test takers and acts as their representative in the online test and UBT test. Prevent testing or cheating.

Embodiments according to the present invention are implemented in the form of program instructions that can be executed through various computer means and can be recorded on a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, and data structures alone or in combination. Computer-readable recording media include storage, magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. -Hardware devices configured to store and execute program instructions in optical media and storage media such as ROM, RAM, flash memory, etc. may be included. Examples of program instructions may include those produced by compilers, machine language code, as well as high-level language code that can be executed by a computer using an interpreter. The hardware device may be configured to operate as one or more software modules to perform the operations of the present invention.

As described above, the method of the present invention is implemented as a program and can be stored on a recording medium (CD-ROM, RAM, ROM, memory card, hard disk, magneto-optical disk, storage device, etc.) in a form that can be read using computer software. ) can be stored in .

Although the present invention has been described with reference to specific embodiments of the present invention, the present invention is not limited to the same configuration and operation as the specific embodiments to illustrate the technical idea as described above, and is not limited to the technical idea and scope of the present invention. It can be implemented with various modifications, and the scope of the present invention should be determined by the claims described later.

100: Online exam or UBT exam server
101: WWW server 102: Control unit
103: Member register 104: User authentication book
105: Test notice department 106: QR code management department
107: Greeting Provider (TTS) 108: Supervisor Verification Department
109: Face recognition unit 110: Fraud processing unit
111: Test management department 120: Test information DB
121: Candidate DB 123: Face photo DB
200: Supervisor terminal 300,310,311: Candidate terminal

Claims (12)

An online test server that registers test taker information and a standard-sized frontal face photo, issues a QR code attached to the test taker's application form, and provides a test program for conducting online tests;
It is connected to the online test server through a wired or wireless communication network, and recognizes the facial recognition module of the face captured through the front camera of the test taker's terminal and the facial behavior pattern captured by the front camera to determine the outline of the face and eyes2/nose/ears2. A test taker terminal equipped with a 5-point scale of facial features and a cheating prevention module; and
During an online test, it includes a proctor terminal that receives face photos and cheating information of one or more test taker terminals through the online test server and transmits an anti-cheating alarm or warning message to the test taker terminal,
The online test server further includes a greeting service (TTS) that provides a greeting through TTS conversion technology after recognizing the QR code attached to the test taker's application form,
The candidate terminal uses any one of a tablet PC, smartphone, or PC on which a test program (App) is installed. The test program is installed, and the facial recognition module and the outline of the face and eyes 2/nose/ears 2 are used. A standard-sized frontal face photo with a 5-point scale of facial features, equipped with an anti-cheating module, and size/rotation/angle corrected by detecting the ROI of the face image captured in real time with the front camera of the test taker's terminal during the online test. Using a frontal face recognition algorithm, facial objects are extracted and facial behavior patterns are recognized, and the outline of the applicant's face and facial features are extracted and classified to the left and right sides of the facial feature points of eyes 2, nose, and ears 2, respectively. Calculate the Euclidean distance (d) and similarity between the center point (pupil) of the ear and the left/right eye, and compare it with the data in the standard-sized frontal facial photo DB of the online test server to check whether a surrogate test is performed. During the online test, the facial behavior pattern captured by the camera of the candidate's terminal is recognized. If the facial outline is not recognized during face recognition, if the face moves away from the test screen, the right eye and right ear are affected depending on the angle at which the front face is turned to the left and right. Since the distance and the distance between the left eye and the left ear are different, if the distance in the corresponding direction of the eye-to-ear distance or the nose-to-ear distance exceeds a certain standard, it is judged to be an act of cheating.
The facial contour recognition technology used in the facial recognition module uses the posenet algorithm to measure the distance between the eyes and ears and the distance between the nose and ears.
If the facial behavior pattern captured by the camera of the test taker's terminal is recognized and judged to be cheating, it is transmitted to the online test server, and after confirmation by the proctor's terminal linked to the online test server, an alarm or warning message is sent to the test taker's terminal. An online exam system using facial contour recognition artificial intelligence that prevents cheating. delete According to paragraph 1,
The candidate terminal is equipped with the facial recognition module and the cheating prevention module, and the cheating prevention module detects visual cheating in the camera vision image data of the candidate terminal in real time, such as facial outline and eyes2/nose. /Ear 2 Equipped with a fraud prevention module that uses a 5-point scale for facial features and a voice recognition module that detects auditory fraud,
The voice recognition module recognizes the candidate's speech sound in order to prevent the candidate from cheating on the candidate terminal, and when the candidate's speech related to cheating is heard during an online test, the voice recognition module inputs the candidate's voice through the microphone of the candidate terminal. Immediately transmit and store voice data to the online test server, and transmit it to the proctor terminal linked to the online test server to generate an alarm to the test taker's terminal through the online test server to prevent cheating. An online exam system using facial contour recognition artificial intelligence that sends warning messages and prevents cheating. delete According to paragraph 1,
The online test server is
A WWW server connected to the candidate terminal and the supervisor terminal through wired or wireless communication;
A control unit that controls to provide online or ubiquitous based learning (UBL) and testing (UBT) to the candidate terminal and the proctor terminal through a wired or wireless communication network;
A member registration unit connected to the control unit, which receives applicant information and stores and manages ID/Passwd;
A user authentication unit connected to the control unit and authenticating the user using a QR code/Passwd or ID/Passwd;
A test notice unit connected to the control unit and providing schedules and locations for tablet PC, smartphone, and PC-based online learning and testing or ubiquitous online learning (UBL) and testing (UBT);
A QR code management unit connected to the control unit and issuing to test takers a QR code attached to the test taker's test ticket corresponding to the test taker information and managing the QR code;
It is connected to the control unit, receives the front face photo of the candidate taken by the camera from the test taker terminal, creates a standard size front face photo with size correction/rotation/angle correction, and creates the candidate information previously stored in the database of the online test server. A proctor confirmation unit where the proctor compares the frontal face photo of a standard size and verifies it at the proctor terminal to check whether the proctor is taking the test and confirms eligibility to take the test;
A face recognition unit connected to the control unit and recognizing the outline of the test taker's face and facial feature points of eyes 2/nose/ear 2 using a facial recognition module in the test taker terminal and receiving the face recognition result from the test taker terminal;
The candidate's terminal measures the distance between the eyes/nose and ears using facial contour recognition technology (posenet algorithm) for the camera image, and moves the head to the right/left depending on whether the 3 points of the eyes/nose are close to the 2 points of the ears at either end. Detects abnormal facial behavior patterns related to cheating, and if the facial outline is not recognized during face recognition or if it deviates from the test screen, the distance in the corresponding direction of the distance between the eyes and ears and the distance between the nose and ears exceeds a certain standard. If the number exceeds the limit, a warning message or alarm is output to the test taker's terminal to prevent cheating, or a cheating processing unit receives a warning message or alarm from the test taker's terminal to the online test server;
A test program (App) and test paper are provided to the candidate terminal and the proctor terminal, and the candidate information, the candidate's on-site face photo, and proctor information are managed, and the test paper is written to each candidate terminal within a certain test time during the online test. A test management unit that saves the answers and transmits them to the online test server at the end of the test, and stores and manages the test takers' test answers, scoring results, supervisor information, and test taker status information; and
Test information DB that stores the test papers, written answers, and scoring results of test takers'terminals; Candidate DB and face DB that store candidate information and standard-sized frontal face photos;
An online exam system using facial contour recognition artificial intelligence to prevent cheating, including. delete According to paragraph 1,
The test paper questions of the test program (App) include not only text and images for each question, but also multimedia test questions containing text, images, VR/AR content, voice and video, and are displayed on test taker terminals through the test server. An online exam system using facial contour recognition artificial intelligence to prevent cheating. Registering member information and, after logging in/user authentication, registering and storing candidate information and a standard-sized frontal face photo on an online test server;
The online test server issuing a QR code attached to the test taker's test ticket corresponding to the test taker information and frontal face photo;
The online test server notifies the test schedule and location according to the administered test;
The recognition results of the front face photo from the camera using a frontal face recognition algorithm are received from the test taker's terminal equipped with a facial recognition module using AI-based facial outline recognition technology, and the size-corrected/rotated/angle-corrected standard is received from the online test server. A full-sized frontal face photo is created, and the proctor's terminal compares the applicant information already stored in the online test server with the standard-sized frontal face photo and its characteristic points to confirm the applicant's identity, prevent proxy testing, and determine whether or not to take the test. deciding step;
The face recognition module of the test taker terminal and the eye 2/nose/ear 2 facial feature point 5-point scale fraud prevention module that recognizes the face behavior pattern recognize the facial behavior pattern captured by the front camera of the test taker terminal and identify facial objects. Extract the outline of the applicant's face and the facial feature points on a 5-point scale of eyes2/nose/ear2, and extract the center points of the left/right ear and left/right eye, respectively, of the facial feature points of eye2/nose/ear2 ( Calculate the Euclidean distance (d) and similarity with the pupil, detect abnormal facial behavior patterns by detecting head movement to the right/left, and if facial outline is not recognized during face recognition, the test taker In order to prevent cheating, if the distance in the relevant direction (eye-to-ear distance, nose-to-ear distance) exceeds a certain standard, or the test screen of the terminal is off, a photo of the face or voice data of the test taker related to cheating is recalled from the test taker's terminal. transmitting to an online test server and output to a proctor's terminal through the online test server, and transmitting a warning message or alarm from the online test server to the candidate's terminal; and
A test program (App) and online test papers are provided to the candidate terminal and the proctor terminal, and the candidate information, the candidate's on-site face photo, and the proctor information are stored and managed in the database of the test server, and during the online test, a certain test time is provided. It includes the step of storing the answers written on the test paper on each test taker's terminal, then transmitting and storing them from each test taker's terminal to the online test server at the end of the test, and providing the scoring results of the answers written on the test paper by the test takers to the test taker's terminal,
The facial contour recognition technology used in the facial recognition module uses the posenet algorithm to measure the distance between eyes and ears and the distance between nose and ears,
The candidate terminal uses any one of a tablet PC, smartphone, or PC on which a test program (App) is installed, and the test program is installed, the facial recognition module, the outline of the face, and eyes 2/nose/ears. 2. Facial feature points on a 5-point scale Equipped with an anti-cheating module, when facial outlines are not recognized during face recognition, when the test taker's terminal is out of the test screen, the distance between the eyes and ears, and the distance between the nose and ears in the corresponding direction. A method of providing online exams using facial contour recognition artificial intelligence to prevent cheating by judging it as cheating if it exceeds a certain threshold. delete According to clause 8,
It is operated by taking a photo of the QR code attached to the candidate's test ticket with a smartphone and placing it on the recognition screen of the camera of the test taker's terminal. After recognizing the QR code, it further includes the step of providing a greeting through the TTS (Text To Speech) function. How to deliver online exams using facial contour recognition artificial intelligence to prevent cheating. According to clause 8,
The test paper questions in the test program (App) include not only text and images for each question, but also multimedia test questions that include text, images, VR/AR content, voice, and video, and are displayed on the test taker's terminal through the test server. A method of providing online exams using facial contour recognition artificial intelligence to prevent behavior. According to clause 8,
The candidate terminal is equipped with the facial recognition module and the cheating prevention module, and the cheating prevention module detects visual cheating in the camera vision image data of the candidate terminal in real time, such as the facial outline and eyes2/nose/ It is equipped with a fraud prevention module that measures ear 2's facial features on a 5-point scale and a voice recognition module that detects auditory fraud.
The voice recognition module recognizes the candidate's speech sound in order to prevent the candidate from cheating on the candidate terminal, and when the candidate's speech related to cheating is heard during an online test, the voice recognition module inputs the candidate's voice through the microphone of the candidate terminal. Voice data is immediately transmitted to the online test server and stored, and it is transmitted to the proctor terminal linked to the online test server, and an alarm is sent to the candidate terminal through the online test server to prevent the candidate terminal from cheating. A method of providing an online exam using facial contour recognition artificial intelligence to prevent cheating, further comprising sending a warning message.